5 6242249775165472907 PDF

Research Article
A 46-year Analysis of Gender

Trends in Academic Authorship in
Orthopaedic Sports Medicine
Abstract
Downloaded from https://journals.lww.com/jaaos by VA2NP/EQXy7tVERrPFBmNI720/El0h9aq4v0nmuiUXA4hX9aojlvboi4jKuu7mZhvkCXeM/gYZgbkkjxDDJ5K2pz+RbquskZst6Vc+6Er3OPXEqQXUHNh7517s346CEBlaanNKkDNKUe0ao+O5YPVBTVF7/FgR7/ on 08/14/2019
Chang-Yeon Kim, MD, MS Introduction: Participation of women in medicine has increased

Lakshmanan Sivasundaram, MD markedly in recent decades, but gender disparities still exist,
particularly in academic medicine. To provide insight into the gender
Nikunj N. Trivedi, MD
gap, specifically in academic orthopaedic sports medicine, we
Allison Gilmore, MD investigated the relationship between gender and authorship in
Robert J. Gillespie, MD orthopaedic sports literature from 1972 to 2018.
Michael J. Salata, MD Methods: Information about every original article in four prominent
orthopaedic sports medicine journals between 1972 and 2018 was
Raymond W. Liu, MD
extracted from PubMed. The proportions of female first, second,
James E. Voos, MD middle, and senior authors over time were determined. Gender
influences on level of evidence, academic degrees, and academic
productivity and longevity were also studied. Student t-test, multiple
linear regression, chi-square test, Cochran-Armitage trend test, and
Kaplan-Meier analysis were used to determine significance between
From the Department of Orthopaedic groups.
Surgery, University Hospitals of Results: In our sample, 16.6% of all authors were female. The
Cleveland, Case Western Reserve
University (Dr. Kim,
proportion of female authors increased from 2.6% (1972 to 1979) to
Dr. Sivasundaram, Dr. Trivedi, 14.7% (2010 to 2018). Female authors averaged fewer publications
Dr. Gilmore, Dr. Gillespie, Dr. Salata, (1.9 versus 2.8 articles for male authors) and were more likely to be
Dr. Liu, and Dr. Voos), and the
University Hospitals of Cleveland,
attributed middle authorship (45.9% versus 37.1%) than senior
Sports Medicine Institute, Cleveland, authorship (14.7% versus 22.1%, P , 0.001). Female authors were
OH (Dr. Gillespie, Dr. Salata, and more likely to be full-time research staff, such as a PhD (18.2%
Dr. Voos).
versus 9.0%, P , 0.001), which correlated with a higher level of
Correspondence to evidence (B = 20.162, P , 0.001). Gender differences in academic
Dr. Sivasundaram:
lakshmanan.sivasundaram@ longevity decreased over decades (1972 to 1989, 1990 to 1999,
uhhospitals.org 2000 to 2008), demonstrated by decreasing significance of Kaplan-
None of the following authors or any Meier log-rank tests (,0.01, ,0.01, 0.045).
immediate family member has Conclusion: Female investigators in orthopaedic sports medicine are
received anything of value from or has
stock or stock options held in a
authoring publications at a growing rate, increasing almost sevenfold
commercial company or institution from 1972 to 2018. Although women published two-thirds the volume
related directly or indirectly to the of male investigators overall, and were more likely to be full-time
subject of this article: Dr. Kim,
Dr. Sivasundaram, Dr. Trivedi,
research staff, gender differences in academic productivity and
Dr. Gilmore, Dr. Gillespie, Dr. Salata, longevity have decreased over time.
Dr. Liu, and Dr. Voos. Level of Evidence: Level III, Retrospective Cohort Design,
J Am Acad Orthop Surg 2019;27: Observational Study
493-501
DOI: 10.5435/JAAOS-D-18-00669
Copyright 2019 by the American

Academy of Orthopaedic Surgeons. N otable advancements have been
made in the participation of
women in medicine. Although women
accounted for only nine percent of
accepted medical student applicants
in 1965, their representation was
July 1, 2019, Vol 27, No 13 493
Copyright © the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
Gender Trends in Sports Medicine
Table 1
Study Characteristics
Impact
Year First No. of Issues in No. of Issues in Year First Indexed Factor
Published First Volume 2017 Volume on PubMed (2016)
A. Orthopaedic sports
medicine journal metrics
AJSM 1972 4 14 1978 5.67
Arthroscopy 1985 4 12 1985 4.29
JSES 1993 1 12 1993 2.73
Sports Health 2009 6 6 2009 1.56
B. The number of articles and authors

included in the analysis
Original research articles 18,354
Total number of authors identified 82,559
Unique authors identified 31,796
Matched to gender (%) 30,005 (94.4%)
Female (%) 4,986 (16.6%)
Male (%) 25,019 (83.4%)
AJSM = American Journal of Sports Medicine, JSES = Journal of Shoulder and Elbow Surgery
44% in 1998 to 1999 and 48% in edly underrepresented in academic research longevity (years of active
2008 to 2009.1 Despite the rising publishing and rarely publish as publication).
matriculation of women to medical senior author, which plays a large
schools, there is still a paucity of role in academic mobility.4,6,7,9–14 Methods
female representation in academic Over time, efforts have been made
medicine.2–5 This gender disparity to promote gender diversity in aca- All original articles published
becomes more pronounced with demic promotions and publishing. between the years 1972 and 2018 in
increasing rank; an investigation in However, the progress of female four prominent orthopaedic sports
2011 found that only 19% of tenured representation in the field of ortho- medicine journals (American Journal
professors and 17% of full professors paedic sports medicine research has of Sports Medicine, Arthroscopy,
were women, compared with 41% of not been studied. In addition, most Journal of Shoulder and Elbow
assistant professors.2 When looking previous gender analysis in academic Surgery, Sports Health) (Table 1)
specifically at subspecialties like or- authorship focuses on a subset of ar- were extracted from PubMed. To be
thopaedic surgery, only seven percent ticles from selected years, limiting the included in the study, a journal had
of academic faculty were women in validity of their conclusions over to be based in the United States and
2007, with one female department time.4,11 Finally, these studies only have a reputation for publishing
chair in the United States.2,5–7 identify first and senior authors, work related to orthopaedic sports
Many factors have been attributed which may underrepresent the pro- medicine. Both Arthroscopy and The
to this underrepresentation, includ- portion of women in publications.4,9 Journal of Shoulder and Elbow
ing relative lack of exposure to sur- The purpose of this study was to Surgery investigate surgical man-
gical subspecialties in medical school investigate the gender trends of all agement of sports-related issues.
education, continued misperceptions cited authors from all original articles Sports Health and The American
about orthopaedic surgery, and published in four major orthopaedic Journal of Sports Medicine are
unconscious gender biases.5 Female sports medicine journals from the administered by the American Ortho-
recipients of academic grants and year 1972 to 2018. Our secondary paedic Society of Sports Medicine. In
awards report experiencing gender objectives were to investigate the addition, journals were required to
discrimination as well as sexual gender differences in leadership roles have an orthopaedic surgeon as editor-
harassment.8 Women are also mark- (first and senior authorships) and in-chief, with the assumption that such
494 Journal of the American Academy of Orthopaedic Surgeons
Chang-Yeon Kim, MD, MS, et al
journals would be more applicable combination of the most common 1989, 1990 to 1999, 2000 to 2009,
to orthopaedic surgeons. Author first and last names from Social 2010 to 2018. The proportion of
names, author degree, year of Security and Census Bureau data- female and male authors who pub-
publication, level of evidence (ob- bases.16 For all authors with a lished 1, 2 to 5, 6 to 10, and .10
tained from abstracts using regular common name, all publications were articles during these periods was
expressions, when available), and cross-referenced with place of pub- then assessed. Trends in research
article title were extracted from lication and affiliation of authors to longevity was evaluated using a
each article. Articles that were not confirm whether they were from the subset of authors who had their first
original research studies (eg, com- same or different author. article published between 1972 and
ments, letters, editorials, memo- The proportion of female first 1989, 1990 and 1999, or 2000 and
randa, historical articles, conference/ (publications with one author were 2008. Each author in these three
society proceedings, lectures) were treated as having a first author), sec- periods was followed for 10 years
excluded. ond, middle (defined as all authors from first publication, so that all
For authors with a full name listed, between second and senior, when authors received the same follow-up
gender was determined through a applicable), and senior authors was period regardless of year of first pub-
custom computer algorithm that determined for each decade (1972 to lication. A Kaplan-Meier survival
queried a database containing 1980, 1981 to 1990, 1991 to 2000, analysis was then performed to
1,877,786 names in 178 countries 2001 to 2010, 2011 to 2018). A sec- observe the time between initial and
(gender-api.com) to match first ondary analysis of women in leadership last publication within the 10-year
names with gender. The Gender- roles (first or senior) was conducted for period, and whether this time was
Application Programming Interface every year between 1972 and 2018. affected by gender. Because the large
has been validated to be the most Additional criteria such as total publi- majority of authors published only
accurate online platform for gender cation number per author, authorship once, the survival analysis was limited
determination in a benchmark study roles throughout an author’s publica- to authors who had at least 1 year of
by the Gender Gap in Science tion history, academic degrees (MD, publication history, as a screen for
group.15 In addition to providing a MD/PhD, PhD, other), and level of authors with notable academic careers.
gender assignment, the Application evidence were also stratified by gen- In an effort to correlate the pro-
Programming Interface also provides der. Multiple linear regression analy- portion of publications by female
an accuracy value, ranging between sis was conducted to correlate gender authors over time with the propor-
zero and 100%, indicating the con- of first or senior authors with level of tion of female orthopaedic surgeons
fidence in its gender assignment. Any evidence. and orthopaedic sports medicine
assignment with less than 90% The top 10 most published male providers over time, a separate gen-
confidence was manually reviewed and female authors during the overall der classification was conducted on
by one of the authors in this study, period and by decade were also all orthopaedic surgeons listed in the
including a search of the author’s identified. All authors listed in the top American Board of Orthopaedic
online profile. 10 were verified to have correctly Surgery (ABOS) public directory.
For the minority of authors who assigned gender using online aca- This directory lists the 29,172
listed their middle names as first demic profiles. In addition, the pub- orthopaedic surgeons who have
names (eg, J. Edgar Hoover), gender lishing periods of the top 20 authors passed Part I of the board certifica-
was derived from the middle names, who published for at least 10 years tion examination. The directory
and this was confirmed to be accurate (as a screen for authors with long- provides the dates of certification and
by manually checking online profiles. term publication history) were strat- also lists whether the surgeon has
Authors with identical first and last ified into four consecutive quartiles obtained Subspecialty Certification
names could be differentiated by (for instance, an author who pub- in Orthopaedic Sports Medicine,
middle initial or a suffix (eg, John lished for 20 years would have four which the ABOS started to adminis-
Smith, John A Smith, John B Smith). consecutive publishing quartiles of 5 ter in 2007. Although limited to cur-
In these instances, a manual review years each). The distribution of first, rently active orthopaedic surgeons,
was conducted to confirm that these second, middle, and senior author- the directory made it possible to
were separate individuals. Another ship in each quartile was then ana- analyze the total number of female
review was conducted on any author lyzed to assess changes in authorship and male orthopaedic surgeons cer-
with more than one publication roles through academic careers. tified each year between 1972 and
(36% of identified authors) who Overall research productivity was 2018, as well as the subset of female
had a common name, as defined by a divided into four periods: 1972 to and male orthopaedic surgeons who
July 1, 2019, Vol 27, No 13 495
Figure 1 Results
Analysis of Female
Authorship From 1972 to
2018
A total of 18,354 articles and 82,550
authors were included for analysis.
Of those, 31,796 were unique au-
thors, with 30,005 matched to gender
(94.4%). There were 4,986 female
authors (16.6%) and 25,019 male
authors (83.4%) (Table 1). Overall
proportion of female authors
increased from 2.6% (1972 to 1979)
to 14.7% (2010 to 2018) (Figure 1).
First and senior author representa-
tion increased from 1.9% and 3.5%
Representation of female authors in orthopaedic sports medicine literature from (1972 to 1979) to 13.4% and 9.2%
1972 to 2018. (2010 to 2018), respectively. Second
and middle author representation
increased from 3.5% and 2.6%
received a sports subspecialty certifi- Algorithm development and statisti- (1972 to 1979) to 18% and 17.7%
cation since 2007. cal analysis were performed using (2010 to 2018), respectively. All
Cochran-Armitage trend test was Python programming language (Python linear trends were statistically sig-
conducted to determine significance foundation, www.python.org) and R nificant (P , 0.0001). A year-by-
of gender trends over time. Chi- (R Foundation, Vienna, Austria, year analysis of leadership positions
square analysis was used for cate- www.r-project.org). For all statisti- (first or senior authors) (Figure 2)
gorical variables and Student t-test cal testing, a P value , 0.05 was showed an overall increase in female
was used for continuous variables. considered significant. first authors, with female senior
Figure 2
Proportion of female authors on a yearly basis from 1972 to 2018.
Table 2
Analysis of all Unique Authors
Female Male P value
Distribution of publication count: frequency (%)

1 3,531 (70.9) 15,455 (61.8) ,0.0001
2-10 1,368 (27.5) 8,551 (34.2) —
11-20 56 (1.1) 624 (2.5) —
21-30 16 (0.3) 180 (0.7) —
31-40 6 (0.1) 79 (0.3) —
.40 6 (0.1) 133 (0.5) —
Distribution of authorship type: frequency (%)
First 1,682 (17.6%) 16,096 (22.9%) ,0.0001
Second 2075 (21.7%) 12,540 (17.9%) —
Middle 4,387 (45.9%) 26,034 (37.1%) —
Senior 1,404 (14.7%) 15,502 (22.1%) —
Highest academic degrees: frequency (%)
MD 1815 (36.4%) 16,297 (65.1%) ,0.0001
MD/PhD 214 (4.3%) 2,508 (10.0%) —
PhD 905 (18.2%) 2,264 (9.0%) —
Other (ie, BA, BS, MS) 2048 (41.1%) 3,954 (15.8%) —
authorship reaching a peak in 1983 465) compared with the top male of female authors had three or more
and then stabilizing. authors. Of the top 10 female au- first and senior publications, re-
thors, 3/10 were clinicians (held a spectively, compared with 5.6% and
MD or MD/PhD) compared with 4.7% of male authors. A multiple
Analysis of Unique Authors 9/10 top male authors. For the top 20 linear regression analysis comparing
Of all unique female authors, 70.9% female and male authors who pub- leadership gender with level of evi-
published one article compared with lished for at least 10 years, female dence demonstrated that female first
61.8% of male authors, and 0.5% authors were most commonly attrib- and senior authors were markedly
published more than 20 articles uted middle authorship throughout associated with higher level of evi-
compared with 1.5% of male authors their publication history (39.2% in dence (P , 0.001, Table 3). How-
(P , 0.0001) (Table 2). Female au- the first quartile to 58.8% in the ever, when the academic degrees of
thors published a larger proportion fourth quartile, Figure 3). In contrast, authors were added to the regression
of second (21.7% versus 17.9%, P , the most common role for male au- model, gender was not a notable
0.001) and middle author pub- thors changed from first authorship in covariate for level of evidence com-
lications (45.9% versus 37.1%, P , their first quartile of publication his- pared with author degree, with a
0.001) and a smaller proportion of tory (44.6% of publications in quar- dedicated PhD demonstrating the
first (17.6% versus 22.9%, P , tile) to senior authorship in their third strongest association with higher
0.001) and senior author pub- and fourth quartiles of publication level of evidence (B = 20.491 for first
lications (14.7% versus 22.1%, P , history (48.5% of publications in author, B = 20.361 for senior
0.001). By number of publications, quartile, P , 0.0001). author, P , 0.001).
female authors averaged 1.9 pub- Both female and male authors
lications compared with 2.8 pub- demonstrated an increase in produc-
lications for males (P , 0.0001). A Analysis of Leadership Roles tivity from 1972 to 2018. The pro-
higher proportion of female authors In total, 76.7% and 82% of female portion of female authors publishing
held a PhD without a MD (18.2%) authors did not have a first and senior more than one article increased from
compared with male authors (9.0%). publication, respectively, compared 10.4% in 1972 to 1989 to 29.4% in
The top female authors published with 66.8% and 74.6% of male au- 2010 to 2018, compared with a rise
3.4 times fewer articles (1,581 versus thors (P , 0.0001); 1.9% and 1.6% from 24.2% to 36.3% for male
July 1, 2019, Vol 27, No 13 497
Figure 3 In terms of research longevity,

overall, 74.1% of female authors
published for less than 1 year, com-
pared with 64.8% of male authors.
Among authors who first published
in 1972 to 1989 with a research lon-
gevity of at least 1 year, 45% of
women published at 5 years (versus
70% of men) and 10% published at
10 years (versus 21% of men) after
initial publication (Figure 4). Among
authors who first published in 2000
to 2008 with a research longevity of
at least 1 year, 55% of women
published at 5 years (versus 59.3%
of men) and 14% of women pub-
lished at 10 years (versus 17.2% of
men) after initial publication. The
Proportion of authorship roles throughout the publishing periods of the top 20 gender difference in research lon-
female and male authors. Each author’s full publishing period was divided into gevity up to 10 years became less
four quartiles to represent the changes in authorship order for each researcher
throughout different phases of their academic career. notable in more recent periods,
demonstrated by the decreasing
gap in the survival curves and
authors (Table 4). The greatest articles, which rose fourfold from decreasing significance of the log-rank
increase in productivity occurred for 0.5% in 1972 to 1989 to 2.0% in test (Figure 4, P values displayed on
female authors publishing 6 to 10 2010 to 2018. plots).
Table 3
Multiple Linear Regression Model Correlating Leadership, Gender, and Degree to Level of Evidence (Levels 1, 2, 3, 4, 5)
Standardized 95% CI Lower 95% CI Upper
B Coefficient Bound Bound P value
Model w/gender
First author gender 20.162 20.056 20.25 20.073 ,0.001
Senior author gender 20.199 20.062 20.297 20.102 ,0.001
Model w/gender and degree
Gender
First author gender 20.024 20.008 20.116 0.068 0.608
Senior author gender 20.086 20.027 20.191 0.019 0.108
First author degree
MD Reference — — — —
MD/PhD 20.116 20.048 20.196 20.037 0.004
PhD 20.491 20.13 20.617 20.365 ,0.001
Other 20.066 20.018 20.182 0.05 0.264
Senior author degree
MD Reference — — — —
MD/PhD 20.221 20.097 20.295 20.146 ,0.001
PhD 20.361 20.11 20.473 20.248 ,0.001
Other 20.127 20.029 20.267 0.013 0.076
CI = confidence interval
Because a lower number indicates higher level of evidence, a negative B is associated with a higher level of evidence. For gender analysis, males
were used as the reference, so that a negative coefficient indicates that female authors publish research with high level of evidence.
Table 4
Research Productivity by Decade
1972-1989a 1990-1999a 2000-2009a 2010-2018a
Publication Count Female Male Female Male Female Male Female Male
1 89.6% 75.8% 82.1% 68.8% 75.9% 65.6% 70.6% 63.7%

2-5 9.4% 20.9% 16.5% 27.6% 21.8% 28.6% 25.7% 29.4%
6-10 0.5% 2.1% 0.6% 1.9% 1.6% 3.1% 2.0% 3.3%
.10 0.5% 1.2% 0.8% 1.7% 0.7% 2.6% 1.7% 3.6%
a
P , 0.0001. Percentages, not absolute frequencies, were reported to protect author privacy.
Figure 4
Survival analysis plots of female and male authors with their first publication between 1972 and 2008. The plots reflect the
overall span of continued publication over a 10-year period for each individual. Convergence of the female and male plots is
observed with time.
Analysis of the Total Digital Content 1, http://links.lww.

com/JAAOS/A320). When decade
Discussion
Orthopaedic Surgeon and
Sports Medicine Provider intervals were analyzed, the propor-
Our study shows that over the last 46
tion of female orthopaedic surgeons
Workforce years, female authorship in ortho-
certified between 1972 and 1979 was
The proportion of female orthopaedic paedic sports medicine literature has
1.7% compared with 12% in 2010
surgeons certified in 1972 was 2% to 2018 (P , 0.001, Appendix II, increased markedly (from 2.6% to
compared with 13.8% in 2018 (P , Supplemental Digital Content 1, 14.7%). Although there continues to
0.001, Appendix I, Supplemental http://links.lww.com/JAAOS/A320). be differences in leadership roles and
Digital Content 1, http://links.lww. When analyzed in intervals of 4 publication volume between female
com/JAAOS/A320). The proportion years, the proportion of female and male investigators, the overall
of female orthopaedic sports medicine orthopaedic sports medicine pro- trend leans toward gender parity, as
providers increased as well, from 4. viders certified in 2007 to 2010 was evidenced by decreasing gaps in aca-
2% in 2007 to 10.3% in 2017, 4% compared with 8% in 2015 to demic productivity and longevity in
although the interval analyzed was 2017 (P = 0.018, Appendix III, recent decades. More female inves-
too small to detect statistical signifi- Supplemental Digital Content 1, tigators tended to be full-time research
cance (Appendix I, Supplemental http://links.lww.com/JAAOS/A320). staff with a PhD, which was associated
July 1, 2019, Vol 27, No 13 499
with a higher level of evidence in assumption of nonwork-related re- there should be greater flexibility for
publications. sponsibilities. In surveys of senior tenure-track positions, with protected
Although women are becoming orthopaedic residents and surgeons, research time and accommodations for
more involved in orthopaedic sports markedly more women planned on familial obligations. Lewis et al5 sug-
research, they are still less likely to reducing their work hours or changing gest recruiting female medical students
publish in leadership roles compared to a part-time status than men, despite early on to careers in orthopaedic
with men. In particular, although the no differences in projected retirement surgery, preferably with female ortho-
proportion of female first authors has age.19 The authors theorize that this paedic attendings as role models. This
been rising steadily from 1972 to may due to women taking on a larger has been successfully implemented in
2018, the proportion of female senior share of familial responsibilities the Perry Initiative’s Medical Student
authors has plateaued. This trend is compared with men. 2 Such duties Outreach Program, where the match
consistent with the work by Sing may place a notable burden on rate for program alumnae was twice
et al9 who demonstrated a similar female clinicians trying to pursue that of the percentage of all females in
plateau in senior authorship among research in addition to clinical orthopaedic residency.21 Our decade-
investigators in spine research. Our responsibilities, compared with by-decade analysis of research lon-
results are corroborated by our full-time research staff who can gevity suggests that efforts to promote
analysis of female and male authors devote more time to research. Our gender parity may be having a positive
with mature research careers, which study agrees with these findings, effect, as the gender gap in 10-year
demonstrated that women were most as women were most commonly research longevity has been narrowing
likely to serve as middle authors attributed middle authorship through- over time.
throughout their publication history. out their publication history, An important limitation of our study
In contrast, men transitioned from whereas men transitioned to se- is that we focused on four subspecialty-
writing mostly first-author pub- nior authorship in the third and specific journals, which may not
lications in the first quartile of fourth quartiles of their publica- include all orthopaedic sports-related
publication history to senior-author tion history. publications. Although we initially
publications in their third and fourth Aside from work-life balance, gen- considered including articles from more
quartiles. Furthermore, the top 10 der discrimination, sexual discrimi- general journals such as The Journal of
female authors produced 3.4 times nation, and a lack of female mentors Bone and Joint Surgery and Clinical
fewer publications than male authors may also contribute to this issue. In a Orthopaedics and Related Research,
and were mostly full-time research survey of K08 and K23 career devel- we found it challenging to manually
staff. These findings are consistent with opment award recipients, Jagsi et al8 judge whether such articles were sports
recent reports that women have lower found that 69.6% of female acad- studies. Therefore, the decision was
h-indices among orthopaedic faculty emicians experienced bias, with made to focus on journals that spe-
and are less likely to be appointed to 26.2% of responders reporting they cifically published studies on ortho-
senior leadership positions such as experienced a gender disadvantage paedic sports medicine. However, we
department chairmanship.4,17,18 in career advancement. Concerning chose these journals based on their
Our finding that women in ortho- is that 30% of respondents reported prominence and reputation, and thus
paedic sports-related research were less personally experiencing sexual ha- they are likely good representations of
likely to be full-time clinicians is con- rassment, with 41% of these women the overall field. Second, our analysis
sistent with previous findings in other experiencing unwanted sexual ad- of authorship role is limited to tracking
subspecialties.9 In their investigation of vances.8 Zhuge et al2 state that the author order in publications. Third,
gender in spine-related research, Sing lack of same-sex mentors may make although our algorithm was able to
et al9 hypothesized that the lack of it difficult for the female academic identify academic degrees (eg, MD,
female clinician–scientists may be clinician to find someone who can MD/PhD), we were unable to identify
attributed to higher requirements for empathize with the gender-specific the specialties of each author inves-
in-hospital postoperative care in spine difficulties women face in academia. tigated (ie, orthopaedic surgeon,
surgery than in other subspecialties. Our study comes at a time of more nonsurgical orthopaedic surgeon,
However, our results suggest that this open dialogue regarding the role of physiatrist). Thus, inclusion of clini-
may not be a contributing factor, as gender discrimination in the healthcare cians who are not orthopaedic sur-
sports medicine is predominantly an environment.20 Efforts to correct this geons may affect our conclusions.
outpatient specialty. issue must be conducted at an institu- However, we based our journal
An additional burden on potential tional and personal level. At an insti- inclusion criteria on relevance to
female clinician–scientists may be the tutional level, Zhuge et al2 suggest that orthopaedic surgeons (as discussed in
the Methods section). In addition, 9. Sing DC, Jain D, Ouyang D: Gender trends
during manual review of a random
Conclusion in authorship of spine-related academic
literature-a 39-year perspective. Spine J
sample of authors to assess accuracy 2017;17:1749-1754.
Our bibliometric analysis of 46 years
(for gender probability below 90%),
in orthopaedic sports medicine liter- 10. Filardo G, da Graca B, Sass DM, Pollock
most clinicians were indeed orthopae- BD, Smith EB, Martinez MA: Trends and
ature demonstrates that although comparison of female first authorship in high
dic surgeons. Fourth, our analysis of
women are still underrepresented in impact medical journals: Observational
the ABOS physician directory dem- study (1994-2014). BMJ 2016;352:i847.
research, especially in leadership
onstrates that although increasing in
roles, their productivity has increased 11. Brinker AR, Liao JL, Kraus KR, et al:
number, both general- and sports- Bibliometric analysis of gender authorship
markedly over time. The overarching
certified female orthopaedic surgeons trends and collaboration dynamics over 30
trend toward gender parity in aca- years of spine 1985 to 2015. Spine 2018;
constitute a minority of the total
demic medicine should be com- 43:E849-E854.
workforce. This discrepancy would
mended, but our results highlight the 12. Gu A, Almeida N, Cohen JS, Peck KM,
definitely contribute to the fact that a
need for further investigation into Merrell GA: Progression of authorship of
smaller proportion of orthopaedic scientific articles in the journal of hand
equitable authorship opportunities. surgery, 1985-2015. J Hand Surg 2017;42:
sports research is produced by women.
291.e291-291.e296.
However, most of our analysis exam-
ines relative trends normalized within 13. Mimouni M, Zayit-Soudry S, Segal O, et al:
References Trends in authorship of articles in major
our sample and is not dependent on ophthalmology journals by gender, 2002-2014.
the total orthopaedic surgeon work- References printed in bold type are Ophthalmology 2016;123:1824-1828.
force. For example, the average num- those published within the past 5 years. 14. O’Connor EE, Chen P, Weston B, et al:
ber of publications per author, Gender trends in academic radiology
placement of authors into publication 1. Leadley J: Women in Medicine: Statistics and publication in the United States revisited.
Benchmarking Report 2008-2009, Acad Radiol 2018;25:1062-1069.
count bins (eg, 1, 10, or .40 pub- Washington, DC, Association of American
lications), and distribution of author- Medical Colleges, 2009. 15. Gender Gap In Science: Telling the Gender
from a Name. https://gender-gap-in-science.
ship type (eg, first, second) are 2. Zhuge Y, Kaufman J, Simeone DM, Chen org/2018/07/16/telling-the-gender-from-a-
independent of the total number of H, Velazquez OC: Is there still a glass name/. Accessed December 10, 2018.
ceiling for women in academic surgery?
orthopaedic sports surgeons. Finally, Ann Surg 2011;253:637-643. 16. Five Thirty Eight: Dear Mona, What’s The
we were unable to identify whether Most CommonN In America? https://
3. O’Connor MI: Medical school experiences fivethirtyeight.com/features/whats-the-
authors were members of an academic shape women students’ interest in most-common-name-in-america/. Accessed
department, which would affect ability orthopaedic surgery. Clin Orthop Relat Res December 10, 2018.
to conduct research. 2016;474:1967-1972.
17. Jagsi R, Guancial E, Worobey C, et al: The
Despite these limitations, our study 4. Okike K, Liu B, Lin YB, et al: The orthopedic “gender gap” in authorship of academic
is the most comprehensive analysis of gender gap: Trends in authorship and editorial medical literature: A 35-year perspective. N
board representation over the past 4 decades. Engl J Med 2006;355:281-287.
gender trends in orthopaedic sports Am J Orthop 2012;41:304-310.
medicine literature to date. With over 18. Nkenke E, Seemann R, Vairaktaris E, et al:
5. Lewis VO, Scherl SA, O’Connor MI: Women Gender trends in authorship in oral and
18,000 articles and 30,000 unique in orthopaedics—way behind the number maxillofacial surgery literature: A 30-year
authors identified from 1972 to curve. J Bone Joint Surg Am 2012;94:e30. analysis. J Craniomaxillofac Surg 2015;43:
913-919.
2018, this study accounts for nearly 6. Silvestre J, Wu LC, Lin IC, Serletti JM:
every article published in the four Gender authorship trends of plastic surgery 19. Hariri S, York SC, O’Connor MI, Parsley
research in the United States. Plast Reconstr BS, McCarthy JC: Career plans of current
major sports medicine subspecialty Surg 2016;138:136e-142e. orthopaedic residents with a focus on sex-
journals. Although most previous based and generational differences. J Bone
7. Bhattacharyya N, Shapiro NL: Increased Joint Surg Am 2011;93:e16.
studies only analyzed first and senior female authorship in otolaryngology over
authors, our study accounts for all the past three decades. Laryngoscope 2000; 20. Jagsi R: Sexual harassment in medicine:
cited authors in a given article. 110:358-361. #MeToo. N Engl J Med 2018;378:209-211.
Inclusion of all years, not specific 8. Jagsi R, Griffith KA, Jones R, 21. Lattanza LL, Meszaros-Dearolf L,
years per decade, also makes it pos- Perumalswami CR, Ubel P, Stewart A: O’Connor MI, et al: The Perry initiative’s
Sexual harassment and discrimination medical student outreach program recruits
sible for our study to track all authors experiences of academic medical faculty. women into orthopaedic residency. Clin
continuously over time. JAMA 2016;315:2120-2121. Orthop Relat Res 2016;474:1962-1966.
July 1, 2019, Vol 27, No 13 501
Review Article
Artistic Gymnastics Injuries;

Epidemiology, Evaluation, and
Treatment
Abstract
Natasha Desai, MD Artistic gymnastics is a physically demanding sport that requires

Danica D. Vance, MD flexibility, agility, and extreme upper and lower body strength. The
specific biomechanics of the sport leads to a unique injury profile.
Melvin P. Rosenwasser, MD
Gymnastic skills require intense upper body weight-bearing, placing
Christopher S. Ahmad, MD unusual forces across the upper extremity joints and predisposing
them to injury. In addition, the required body control during air
aerobatics (tumbling, twisting, flipping) necessitates precise landing
techniques to avoid spine and lower extremity injury. Common
gymnastic injuries include those of the spine and upper extremity such
as spondylolysis, shoulder instability, ulnar collateral ligament injuries,
capitellar osteochondritis dissecans, and several wrist pathologies.
Understanding the injury etiology, prevention, and treatment protocols
is important for a successful recovery and return to sport.
A rtistic gymnastics is known for

its high-flying acrobatics and
feats of strength. Competitive gym-
The sport involves rigorous year-
round training that is started as early
as 4 years of age. On average, gym-
nastics is divided into two categories; nasts are reported to train 5.36 days
artistic and rhythmic gymnastics. per week and 5.04 hours a day.1 Elite-
Artistic gymnastics became a part of level gymnasts tend to specialize in
the Olympics in 1896 and consists of their sport by the age of 12 years
four events for women (vault, uneven with peak training intensity occur-
bars, balance beam, and floor exer- ring at 18 years of age.
cise) and six events for men (floor,
pommel horse, still rings, vault, par-
From the Department of Orthopedic
Surgery, Columbia University, New allel bars, and horizontal bar). This Epidemiology of Injuries
York, NY. article will focus on the injury profile
None of the following authors or any
of artistic gymnastics. Each event is Medical professionals are reported to
immediate family member has scored on difficulty and performance treat more than 86,000 gymnastics
received anything of value from or has execution. The men’s events all related injuries every year. The risk of
stock or stock options held in a heavily rely on upper body strength injury rises as the training hours and
commercial company or institution
related directly or indirectly to the
and upper extremity weight-bearing the level of performance increases.
subject of this article: Dr. Desai, and double leg landings. The wom- The National Collegiate Athletic
Dr. Vance, Dr. Rosenwasser, and en’s events entail more skills Association (NCAA) found women’s
Dr. Ahmad. demanding extreme flexibility and artistic gymnastics to have the sec-
J Am Acad Orthop Surg 2019;27: more single leg landings compared ond highest injury rate in practice,
459-467 with men. Both men and women only behind football, and fourth
DOI: 10.5435/JAAOS-D-18-00147 gymnasts do a notable amount of highest when you combine compe-
tumbling, which includes both upper tition and practices. In the 1980s, the
Academy of Orthopaedic Surgeons. and lower extremity rebounding, NCAA gymnastics injury report rate
flipping, twisting, and hard landings. was as high as 100% to 200%,
July 1, 2019, Vol 27, No 13 459
Common Artistic Gymnastics Injuries
Figure 1 mats, spotting coaches during com- sive forces across the wrist reaches
petition, and increased intensity and higher than two times the athlete’s
pressure.5 body weight.10,11 This large force is
Injuries in gymnastics can be distributed across small joint surface
divided into acute and overuse in- areas leading to extreme stress.
juries, which demonstrate similar In addition, the variety of body
incidence. This review will focus positions during gymnastics places
more on the overuse or chronic in- notable stress across the axial spine
juries as their treatment and rehabil- because it repetitively cycles between
itation tend to be more sport specific. extreme hyperextension and hyper-
The biomechanics of gymnastic ma- flexion (Figure 1). Good core control
neuvers leads to unique stresses to the and lower extremity flexibility play a
body not found in other sport, pre- big role in injury prevention
disposing gymnasts to specific types throughout flexion and extension.
of injuries and specific needs in Hard landings during tumbling,
rehabilitation. Gymnastic skills use dismounts, and vaulting also put
Diagram of a gymnast displaying extreme ranges of motion across the high-impact compressive forces on
hyperextension of the lumbar spine joints and involve intense upper body the spine.12 The proper landing
during a floor routine.
weight-bearing that places unusual technique after dismount is impor-
forces across the upper extremity tant to prevent injury. If done
causing many universities to drop joints as well as increased forces incorrectly, a substantial load is
the sport entirely.2 However, in across the spine and lower extremi- placed on the hips and ankles leading
recent years, injury rates have ties during rebounding and hard to lower extremity injury. Skilled
decreased because of a variety of landings.6,7 gymnasts adopt a specific landing
reasons including safer training technique compared with less elite
protocols and injury rehabilitation. gymnasts.13 The correct landing
This, along with the introduction of Biomechanics Overview technique involves controlling knee
Title IX, has helped increase the and hip flexion, thereby reducing the
number of NCAA women’s gym- An appreciation of the biomechanics forces across the lower extremities
nastics teams with currently ap- is critical to understanding the etiol- and preventing injury13 (Figure 2).
proximately 40 teams in 2018 and ogies of many gymnastic injuries.
only 16 NCAA men’s gymnastics Gymnastics is unique in that it re-
teams.3 Recently, a 10-year obser- quires periods of extreme upper body Head
vational study of NCAA Division 1 weight-bearing in combination with
gymnasts found that injury inci- body flexibility and movement. Gymnastics has a lower head injury
dence was 8.78 per 1,000 athlete- Gymnasts must controllably twist, rate than contact and collision sport.
exposures for men and 9.37 per rotate, and swing in the air while Male gymnasts have a head injury
1,000 athlete-exposures for women. maintaining specific body positions.8 rate of 0.44 per 1,000 athletic ex-
Female gymnasts more frequently A lapse in control of their angular posures and female gymnasts have a
sustained major injuries compared momentum can put them at increased head injury rate of 0.19 per 1,000
with men and more frequently risk for orthopaedic injuries. athletic exposures.4
underwent surgery after injury, Several gymnastics skills such as
24.4% of women required surgery landing, tumbling, and rebounding Concussion
versus 9.2% of men.4 The risk of place unusual amounts of stress Concussions are the most common
injury during competition is almost across the body and joints. Specific head injury sustained in gymnas-
two times higher than during prac- events like beam, vault, and floor tics.1,4,14 The diagnosis and manage-
tice sessions.5 The increased inci- exercises require full upper body ment of concussions in gymnastics
dence in competition may be weight-bearing with the elbow or is similar and standard across all
explained by a combination of fac- wrist in extreme extension. Overtime, sports. The diagnosis is clinical,
tors including increased fatigue of high volumes of training can lead to a using a combination of clinical exam-
completing an entire routine versus buildup of microtrauma causing a ination tests including a vestibular-
an individual skill, decreased pro- more severe injury.8,9 For example, oculomotor and balance examination,
tective equipment/techniques such as during back handsprings, compres- and standardized testing such as the
Natasha Desai, MD, et al
Figure 2 Figure 3
Diagram of a gymnast showing biomechanics of landing after dismount,

demonstrating the (A) incorrect and (B) correct landing technique. A,
Demonstrates a gymnast landing with her knees and feet together and hips
tucked under, causing a stiff impact with an upright torso. B, The gymnast lands
with her hips, knees, and toes in line and her feet shoulder-width apart. On the Photograph showing the stork
lateral image, hip flexion is approximately 30° and her knees do not extend past maneuver: The patient stands on
her toes. In addition, her core is engaged and her trunk is parallel with her lower one leg and hyperextends the lumbar
legs. spine to load the posterior lumbar
spine elements. Reproducible pain
on the ipsilateral side of the standing
leg is characteristic of a
Sport Concussion Assessment Tool- spondylolysis lesion.
5th Edition (SCAT5) or MACE.15-17
Spine
Treatment includes standard cognitive
As previously discussed in the bio-
and physical rest until symptom res- Spondylolysis presents with focal
mechanics section, gymnastics places
olution, followed by a gradual return low back pain that occurs with
tremendous force across the spine,
to play. Recovery may be augmented extension and commonly radiates to
which in return, predisposes the athlete
with vestibular therapy, vision ther- the buttock or thigh region. Gym-
to several spine overuse injuries.12
apy, cervical spine physical therapy nasts will present with chronic and
(PT), and/or aerobic training.15,18 It is dull back pain when performing skills
vital to fully rehabilitate any balance Spondylolysis and that require lumbar spine hyperex-
and vestibular-ocular deficits before Spondylolisthesis tension and rotation such as hand-
starting return to sport.16,17 Gymnasts Spondylolysis is an overuse injury of springs, and backward twists, as well
experience high vestibular stressors the pars interarticularis that results as axial loading such as the rebounding
while performing and therefore will from either an initial traumatic during vaulting and tumbling.12 The
need to be adequately tested and microfracture with subsequent pro- athlete should have a normal neuro-
rehabilitated before return to sport. In gressive stress fractures or a fatigue vascular examination. They can have
addition, athletes should be tested in a fracture from repetitive overload.19 pain with extension, although gym-
controlled environment on their abil- This injury can progress to a spondy- nasts will naturally be able to extend
ity to spin, flip, and invert without lolisthesis, defined as the anterior or further than the average population.
symptoms. The athlete should pro- posterior translation of one vertebra The Stork can be performed which
gress slowly through tumbling and on another (listhesis).19 Gymnasts requires the gymnast to stand on one
twisting exercises, starting with more often develop bilateral pars inter- leg and extend. This test places stress
protective equipment. If restricted articularis fractures, termed isthmic on posterior structures and is positive
from sport for a prolonged time, Spondylolisthesis. The radiographic if pain is reproduced. Although often a
vestibular and aerobic therapy that incidence of spondylolysis in gymnas- useful test, it is nonspecific for spon-
incorporates basic gymnastic skills, tics is reported to be as high as 11% to dylolysis12 (Figure 3).
such as inversions and low-impact 16%; however, radiographic findings Diagnostic imaging includes plain
tumbling, can be helpful. do not always correlate clinically.20 radiographs, MRI, single-photon
July 1, 2019, Vol 27, No 13 461
Figure 4 rest of approximately 10 to 12 weeks investigated that will decrease the

and avoidance of extension activi- postsurgical recovery time and allow
ties.12 In younger athletes, bracing for a faster return to sport.24 Criteria
treatment may be used but is pro- to return to play include pain-free end-
vider dependent. Clinically, gym- range of motion (ROM) and baseline
nasts are often able to return to full strength and aerobic fitness.12
sport independent of radiographic
healing. Therefore, for chronic cases
with a low chance of complete bone Upper Extremity
healing, it is recommended that
management strategies target core Gymnastics involves a large amount
control and return to sport be of upper extremity weight-bearing.
directed by patient symptoms. For This occurs in the form of static holds
symptomatic athletes, initial man- and rebounding while tumbling or
agement should include rest followed vaulting. The repetitive high-impact
by aggressive rehabilitation. PT weight-bearing predisposes gymnasts
should target the anterior pelvic tilt to upper extremity injuries. Injuries
that gymnasts commonly have. This vary between sex, training intensity,
involves increasing the flexibility of and age of participation. Common
hip flexors and quadriceps and injuries are highlighted below.
strengthening weak core and ham-
string muscles.12,21 Gymnasts with Shoulder
Lateral radiograph of the lumbar neurologic deficits, worsening spon- Types of shoulder injuries seen in
spine demonstrating an L5 dylolisthesis, and unresolved pain gymnastics include traumatic injuries,
spondylolysis (red arrow) and an L5-
S1 spondylolisthesis (white lines). should be excluded from sport par- rotator cuff injuries, and shoulder
ticipation. Athletes with bilateral instability including superior labral
spondylolysis with or without tears from anterior to posterior. Mul-
emission CT (SPECT) and CT.12 spondylolisthesis should return to tidirectional shoulder instability has a
Standard spine plain radiographs sport with caution and close obser- high incidence among gymnasts.25
with flexion and extension views are vation. In skeletally immature gym- Elite gymnasts often have fundamen-
evaluated for evidence of a spondy- nasts with spondylolisthesis, it is tal laxity that allows them to excel in
lolysis lesion or spondylolisthesis. A recommended that a lateral radio- this sport but predisposes them to
spondylolysis lesion will appear on graph be taken every 6 to 12 months instability issues. The increased
the oblique view as a lucency in the to evaluate for progression.12 shoulder motion initially leads to
region of the pars19 (Figure 4). If initial nonsurgical management subacromial bursitis and/or shoulder
SPECT imaging is more sensitive to fails, then surgical correction is indi- impingement. As the microtrauma
these lesions. Lesions seen on SPECT cated. The Buck procedure was the continues, plastic deformation of the
but not on plain radiographs tend to first surgical technique designed to supporting shoulder ligamentous
be ,1 year old and more responsive repair a pars interarticularis defect. It structures occurs, leading to acquired
to conservative treatment.12 A posi- uses a lag screw across the pars in- increased laxity and instability.25,26
tive SPECT is followed by a CT scan terarticularis fracture to compress the Patients will present with shoulder
of the spine. CT scans are more defect, with or without the addition pain, instability, and/or clicking dur-
specific than SPECT imaging and of bone graft.22 Morscher et al23 ing specific overhead movements
have the potential to be used as a modified the Buck procedure to use when the shoulder is rotated. Gener-
prognostic predictor for successful bone graft to fill the pars defect fol- alized ligamentous laxity should be
bony healing12 (Figure 4). The use of lowed by hook screw fixation. determined on physical examination.
MRI to diagnose spondylolysis re- Alternatively, the Scott technique Signs of global laxity include ability
mains controversial because data involves bone grafting the defect and to touch thumb to forearm and elbow
have shown MRI to be inferior to passing a wire through the transverse and knee hyperextension. Anterior
SPECT in diagnosing lesions. processes of the vertebrae to com- instability physical examinations
Controversy exists for the ideal press the bone graft into the defect, include the apprehension, relocation,
treatment of these injuries. All treat- aiding in stabilization.22 Newer mini- and release test. Posterior instability
ment strategies involve a period of mally invasive techniques are being examinations include the posterior
load and shift, posterior stress, and cally the capitellum, is a common site Figure 5
the Jerk and Kim tests.26 Magnetic for OCD in gymnasts. The radio-
resonance arthrography or MRI is capitellar joint provides axial stability
useful in confirming diagnosis by to the elbow, transferring approxi-
evaluating the integrity of the shoulder mately 60% of compressive forces
capsule and labral tissue (Figure 5). across the capitellum.30
Initial management involves a Patient typically presents with
period of rest and restriction from vague lateral elbow pain and may
overhead activities. A supervised have mechanical symptoms, such as
exercise and PT program is initiated. catching, clicking, or locking. On
PT focuses on strengthening the physical examination, patients have
dynamic stabilizers of the shoulder tenderness over the radiocapitellar
joint and enhancing proprioception. joint and may lack full extension with
If symptoms improve after completion an increased valgus carrying angle. Axial T2-weighted MR arthrogram
of the PT program, patients may return Patients often have crepitus, especially image demonstrating an anterior
labral tear (red arrow).
to sport. If symptoms persist, surgical with pronation and supination. Patients
treatment is considered, which includes will have a positive radiocapitellar
an arthroscopic shoulder débridement compression test, where pronation and
Figure 6
with labral repair and an inferior supination of the forearm with the
capsular shift for the redundant cap- elbow in extension reproduces pain at
sule.26,27 After surgical management, the radiocapitellar joint.30,31
return to sport occurs at 9 to Initial imaging includes three-view
12 months. The athlete should achieve plain radiographs of the elbow:
full painless ROM that compares to extension AP, 45°-flexion AP, and
the contralateral side and full strength lateral30 (Figure 7). Early in this
before returning to competition.28 injury, radiographs may be negative.
Positive findings on plain radio-
Elbow graphs include lucencies, flattening,
The elbow experiences high valgus sclerosis, fragmentation, and intra-
and varus stresses during gymnastic articular loose bodies. These changes
exercises including tumbling, bars, are typically found in the antero-
dismount, and landing. Specifically, lateral aspect of the capitellum. MRI
repetitive valgus stress on the elbow is the best imaging modality for
results in: (1) traction/tensile forces evaluating OCD lesions and can
on the medial structures, (2) com- demonstrate early-stage lesions
pression of the lateral structures (ie, when radiographs appear normal.
the radiocapitellar joint), and (3) Early MRI findings include uniform
Diagram showing biomechanics of
shearing forces in the posteromedial low-signal-intensity changes in the tumbling: large valgus forces are
compartment, causing a sequela of superficial capitellum on T1- generated across the elbow during
weighted images with normal T2 tumbling events resulting in (A)
elbow injuries29 (Figure 6). tension across the ulnar collateral
imaging. As the lesion progresses,
ligament and (B) compression of the
changes are also seen on T2 imag- capitellum.
Osteochondritis Dissecans ing30,31 (Figure 8). Minamie et al
Osteochondritis dissecans (OCD) re- originally classified OCD lesions
fers to an acquired lesion involving based on AP radiographs. Stage I the lesion. Stable lesions are those
varying degrees of resorption, frag- lesions consisted of a translucent with an open capitellar physis, good
mentation, and sclerosis of the articular cystic shadow in the middle/lateral elbow ROM on examination and
cartilage and underlying subchondral capitellum. Stage II lesions demon- localized flattening or radiolucency,
bone. The etiology of OCD is multifac- strated a split line between the lesion which includes most stage I lesions.
torial including chronic repetitive and the normal bone and stage III Nonsurgical treatment of stable le-
trauma and overuse, vascular insuffi- lesions consisted of loose bodies.31 sions includes a period of rest and
ciency, inflammatory changes, and Treatment of OCD of the cap- activity modification for 6 months.
genetic factors.30,31 The elbow, specifi- itellum is dictated by the stability of As symptoms improve, PT and
July 1, 2019, Vol 27, No 13 463
Figure 7 Figure 8 Figure 9
Coronal MRI image demonstrating Arthroscopic image of the elbow

an osteochondritis dissecans lesion demonstrating an osteocondritis
AP radiograph of the elbow of the capitellum (red circle).
demonstrating an osteochondritis dessicans lesion after antegrade
dissecans lesion of the capitellum drilling.
with a residual loose body (red
circle).
common in gymnasts, especially
those with higher carrying angles. ers radiograph series includes AP,
Repetitive valgus stress results in high lateral, internal/external oblique
strengthening exercises are initiated. tensile forces through the UCL. views, and an oblique axial view
Outcomes are favorable, with radio- Chronic repetitive medial tensile with the elbow in 110° of flexion to
graphic healing of 80% of stable le- stress leads to an overuse injury, demonstrate posteromedial olecranon
sions.32 Return to sport is allowed causing a UCL tear or insufficiency. osteophytes. If medial instability is
after radiographic evidence that the Associated medial elbow patholo- suspected, stress AP radiographs can
lesion has healed and the patient is gies included flexor tendinitis or be performed.29 MR arthrogram is the
asymptomatic.31,32 medial epicondylitis and ulnar nerve benchmark to evaluate the UCL. MRI
Surgical treatment is reserved for neuritis.29,33 imaging is also useful and can show
unstable lesions, which include stage Patients present with vague medial focal discontinuity of the ligament
III and some of stage II lesions or elbow pain. On examination, pa- with joint fluid extravasation on T2
those that have failed nonsurgical tients have pain with palpation over images (Figure 11).
management. Unstable lesions en- the UCL, approximately 2 cm distal Initial treatment of UCL injuries
tail the presence of loose bodies, to the medial epicondyle. Specific consist of rest, PT, and NSAIDs.
mechanical symptoms, and a loss physical examinations to test the Partial injuries have been successfully
of ROM on examination. Surgical integrity of the UCL include the treated with appropriate nonsurgical
treatments include arthroscopic removal moving valgus stress test and measures. After a period of at least
of loose bodies, abrasion chondroplasty, the milking maneuver. In addition, 2 months’ rest, athletes begin an
microfracture, antegrade/retrograde the absence of pain with resisted wrist upper extremity strengthening pro-
drilling, fixation of fragments, and os- flexion can help distinguish a UCL gram before full return. Surgical
teochondral autograft/allograft trans- injury from medial epicondylitis.33 treatment is recommended for com-
plantation30,31 (Figures 9 and 10). Several image modalities are used plete UCL tears in competitive
Return to sport is dependent on the to help diagnose UCL injuries. Plain overhead athletes or partial tears that
patient regaining full painless ROM radiographs of the elbow may reveal have failed conservative management.
and baseline strength, approximately avulsion fragments in acute injury Surgical options include UCL recon-
3 months after surgery.31 or signs of chronic UCL injury struction with autograft/allograft tissue
including loose bodies, ossifications or primary UCL repair (Figure 12).
Ulnar Collateral Ligament of the ligament, subchondral sclero- Return to sport occurs approximately
Ulnar collateral ligament (UCL) in- sis, joint space narrowing, and os- 1 year after surgery if the athlete has
juries are well described in throwing teophytes of the posteromedial full upper extremity strength and full
athletes, but are also found to be compartment.33 A standard throw- painless ROM.29,33
Intraoperative image of the medial

elbow after reconstruction of the
ulnar collateral ligament with a
palmaris longus autograft (white
arrow).
Intraoperative image after
transplantation of an osteochondral Coronal T2-weighted MRI image of
allograft plug for an osteochondritis the elbow demonstrating a complete
tear of the ulnar collateral ligament loading for 8 to 12 weeks, followed
dissecans lesion of the capitellum.
(red arrow). by PT. Alternatively, some higher
level athletes may proceed with
In addition, repetitive valgus stress Scaphoid Stress Fracture percutaneous screw fixation for
of the elbow in the immature gymnast Scaphoid stress fractures are rare in- earlier rehabilitation and return to
can lead to medial epicondyle juries that occur most commonly in sport as early as 2 months
apophyseal avulsion injuries. Peak young gymnasts.35,36 The injuries postoperatively.6,36
age of injury occurs at 11 to 12 years are different from the typical
before fusion of medial epicondyle scaphoid fracture seen in traumatic Distal Radius Physeal Injury
apophysis at 16 to 18 years.34 Pa- settings. The mechanism of injury is Distal radius physeal injury or gym-
tients often experience a sudden pop repetitive extreme wrist extension nast wrist occurs in young gymnasts
associated with swelling and pain with compressive axial loads and is with open epiphyseal plates. High-
over the medial epicondyle. Diag- particularly seen in gymnasts who impact forces during competition
nosis is made on plain radiographs, have rapidly increased their level result in repetitive injury to the epi-
which demonstrate the displaced of training. The vast majority of physeal plate.38
avulsed fragment. Surgical manage- these injuries occur at the scaph- Although the exact cause is
ment with open reduction and oid waist, the area most stressed unknown, it is thought that the injury
internal fixation is indicated for with forced extension as demon- may be the result of compromised
fractures with .5 mm of displace- strated by Majima et al’s biome- blood supply to the metaphyseal and
ment, fragment incarceration in the chanical study. 6,37 Handstands epiphyseal area leading to uncalcified
joint, gross elbow instability, or open and other static maneuvers com- chondrocytes. Unlike an adult wrist,
fracture. Return to full sport occurs monly performed in gymnastics were the immature wrist typically exhibits
around 3 months postoperatively.34 demonstrated to cause considerable negative ulnar variance, resulting in a
stress forces across the scaphoid higher distribution of load to the
Wrist waist.37 immature distal radius, especially
The wrist experiences high-impact This injury presents in the acute or during high compressive forces.
loading forces during gymnastic chronic setting as pain over the ana- Chronic compression can lead to
competition. These include axial tomical snuffbox that is aggravated growth arrest of the physis, causing
compression, torsional forces, and by wrist dorsiflexion. Scaphoid view premature closure of the ulnar aspect.
distractive forces, often while the radiographs may show an area of A deformity results when the distal
wrist is hyperextended. Compressive sclerosis at the scaphoid waist (Figure radius is shifted ulnar and volar.6,7
forces are reported to be as much as 13). The benchmark for diagnosis is Risk factors include the female sex,
16 times an athletes’ body weight.20 MRI (Figure 14). Most injuries are and repetitive activities with extreme
Events most associated with wrist treated nonsurgically with good re- wrist dorsiflexion.6
pain include the pommel horse, floor sults.35 The wrist is placed in a Patients present with dull radial
exercises, and parallel bars. thumb spica cast with no wrist wrist pain during the offending
July 1, 2019, Vol 27, No 13 465
AP radiograph of the wrist

Coronal T1-weighted MRI image of AP radiograph demonstrating
demonstrating sclerosis of the
the wrist demonstrating a stress stage 2 changes of the distal radial
scaphoid waist consistent with a
fracture of the scaphoid (red arrow). physis, including widening of the
scaphoid stress fracture (white
physis and cystic changes of the
arrow).
metaphysis.
Triangular Fibrocartilage
activity that is usually relieved with Complex Tears
rest. Pain at rest is a sign of more petition involve the lower extremity.5
severe injury. On examination, ten- This injury presents in skeletally
mature gymnasts secondary to the Traumatic lower extremity injuries
derness to palpation at the distal occur most commonly during land-
radial physis is noted. Degree of combination of positive ulnar variance
and chronic wrist weight-bearing. Pa- ings, tumbling, vaulting, and dis-
injury is divided into three stages mounts.39 Common injuries include,
based on the presence of radiographic tients present with ulnar-sided wrist
pain that is worse with ulnar deviation knee ligamentous injuries such as
changes. In stage 1, symptoms are anterior cruciate ligament tears,
present without radiographic and pronation. Other injuries such as
tendinitis and carpus instability should ankle sprains, and various foot
changes. Stage 2 demonstrates char- fractures. These injuries follow
acteristic radiographic changes such be ruled out. Patients present with
tenderness between the flexor carpi standard treatment and return to
as widening of the radial physis, cys- sport protocols and will not be dis-
tic changes of the metaphysis, ulnaris and ulnar styloid. Treatment
is initially conservative with rest, cussed in this review.
breaking of the distal aspect of the
epiphysis, and haziness within the NSAIDs, and PT. If symptoms
physis (Figure 15). Stage 3 involves persist, a period of wrist immobiliza- Summary
the addition of changes in ulnar tion typically 4 to 6 weeks is initiated.
variance. When radiographs are Improving wrist flexibility is important This review is intended to overview
negative, MRI is recommended to in the rehabilitative phase with atten- common injuries specific to gymnastics
further evaluate the physis and to tion to weight-bearing hand position and considerations in terms of recovery
rule out other causes. The mainstay because those who turn their hands and return to sport. Although not an
of treatment is rest from weight- outwards are at increased risk. Corti- exhaustive review, this should serve
bearing for at least 6 weeks depend- sone steroid injections can be consid- as a guide to framing your evaluation
ing on the stage. Rehabilitation ered for patients with persistent pain. and differential diagnosis when dealing
focuses on strengthening and pro- Arthroscopic débridement is reserved with gymnasts.
prioception training. Bracing treat- for individuals who do not improve
ment with the Gibson brace and a with conservative measures.6,7
Acknowledgments
palm pad is recommended. Surgery is
typically reserved for stage 3 injuries, Lower Extremity Injuries The authors would like to acknowl-
specifically patients with positive ulnar edge Georg Popa for his time and
variance and those at risk for physis More than 50% of the injuries that contribution toward creating several
arrest.4,6,7 occur during both practice and com- of the illustrations and figures.
13. Christoforidou Α, Patikas D, Bassa E, et al: shoulder in elite female gymnasts. Am J

References Landing from different heights: Biomechanical Orthop 2007;36:660-665.
and neuromuscular strategies in trained
gymnasts and untrained prepubescent girls. 26. Levine WN, Prickett WD, Prymka M,
References printed in bold type are J Electromyogr Kinesiol 2017;32:1-8. Yamaguchi K: Treatment of the athlete with
those published within the past 5 multidirectional shoulder instability.
14. Daneshvar DH, Nowinski CJ, McKee AC, Orthop Clin North Am 2001;32:475-484.
years. Cantu RC: The epidemiology of sport-related
27. Gaskill TR, Millett PJ: Management of
concussion. Clin Sports Med 2011;30:1-17.
1. Saluan P, Styron J, Ackley JF, Prinzbach A, multidirectional instability of the shoulder.
Billow D: Injury types and incidence rates in 15. McCrory P, Meeuwisse WH, Aubry M, J Am Acad Orthop Surg 2011;19:758-767.
precollegiate female gymnasts: A 21-year et al: Consensus statement on concussion in
experience at a single training facility. 28. Vidal LB, Bradley JP: Management of
sport: The 4th International Conference on
Orthop J Sports Med 2015;3: posterior shoulder instability in the athlete.
Concussion in Sport, Zurich, November
2325967115577596. Curr Opin Orthop 2006;17:164-171.
2012. J Athl Train 2013;48:554-575.
2. McAuley E, Hudash G, Shields K, et al: 29. Dugas J, Chronister J, Cain EL Jr, Andrews
16. Broglio SP, Cantu RC, Gioia GA, et al: JR: Ulnar collateral ligament in the
Injuries in women’s gymnastics: The state of
National Athletic Trainers’ Association overhead athlete: A current review. Sports
the art. Am J Sports Med 1988;16(suppl 1):
position statement: Management of sport Med Arthrosc Rev 2014;22:169-182.
S124-S131.
concussion. J Athl Train 2014;49:245-265.
3. Kerr ZY, Hayden R, Barr M, Klossner DA, 30. Churchill RW, Munoz J, Ahmad CS:
Dompier TP: Epidemiology of National 17. Guskiewicz K, Bruce S, Cantu R, et al: Research Osteochondritis dissecans of the elbow. Curr
Collegiate Athletic Association Women’s based recommendations on management of Rev Musculoskelet Med 2016;9:232-239.
Gymnastics Injuries, 2009–2010 through sport related concussion: Summary of the
National Athletic Trainers’ Association position 31. Baker CL III, Romeo AA, Baker CL Jr:
2013–2014. J Athl Train 2015;50:870-878.
statement. Br J Sports Med 2006;40:6-10. Osteochondritis dissecans of the capitellum.
4. Westermann RW, Giblin M, Vaske A, Am J Sports Med 2010;38:1917-1928.
Grosso K, Wolf BR: Evaluation of men’s 18. May KH, Marshall DL, Burns TG, Popoli
DM, Polikandriotis JA: Pediatric sports specific 32. Mihara K, Tsutsui H, Nishinaka N,
and women’s gymnastics injuries: A 10-year
return to play guidelines following concussion. Yamaguchi K: Nonoperative treatment for
observational study. Sports Health 2015;7:
Int J Sports Phys Ther 2014;9:242-255. osteochondritis dissecans of the capitellum.
161-165.
Am J Sports Med 2009;37:298-304.
5. Marshall SW, Covassin T, Dick R, Nassar 19. Standaert CJ, Herring SA: Spondylolysis: A
critical review. Br J Sports Med 2000;34: 33. Safran M, Ahmad CS, Elattrache NS: Ulnar
LG, Agel J: Descriptive epidemiology of
415-422. collateral ligament of the elbow.
collegiate women’s gymnastics injuries:
Arthroscopy 2005;21:1381-1395.
National Collegiate Athletic Association Injury
Surveillance System, 1988–1989 through 20. Overlin AJF, Chima B, Erickson S: Update
34. Redler LH, Dines JS: Elbow trauma in the
2003–2004. J Athl Train 2007;42:234. on artistic gymnastics. Curr Sports Med
athlete. Hand Clin 2015;31:663-681.
Rep 2011;10:304-309.
6. Wolf MR, Avery D, Wolf JM: Upper 35. Matzkin E, Singer DI: Scaphoid stress
extremity injuries in gymnasts. Hand Clin 21. McCleary MD, Congeni JA: Current
fracture in a 13-year-old gymnast: A case
2017;33:187-197. concepts in the diagnosis and treatment of
report. J Hand Surg Am 2000;25:710-713.
spondylolysis in young athletes. Curr
7. Webb BG, Rettig LA: Gymnastic wrist injuries. Sports Med Rep 2007;6:62-66. 36. Yamagiwa T, Fujioka H, Okuno H,
Curr Sports Med Rep 2008;7:289-295. Tomatsuri M, Tanaka J, Yoshiya S:
22. Panteliadis P, Nagra NS, Edwards KL, Surgical treatment of stress fracture of the
8. Prassas S, Kwon YH, Sands WA: Behrbalk E, Boszczyk B: Athletic population
Biomechanical research in artistic gymnastics: scaphoid of an adolescent gymnast. J Sports
with spondylolysis: Review of outcomes Sci Med 2009;8:702.
A review. Sports Biomech 2006;5:261-291. following surgical repair or conservative
9. Dufek JS, Bates BT: Biomechanical factors management. Glob Spine J 2016;6:615-625. 37. Majima M, Horii E, Matsuki H, Hirata H,
associated with injury during landing in Genda E: Load transmission through the
23. Morscher E, Gerber B, Fase J: Surgical wrist in the extended position. J Hand Surg
jump sports. Sports Med 1991;12:326-337. treatment of spondylolisthesis by bone Am 2008;33:182-188.
10. Koh M, Jennings L, Elliott B: Gymnastics: grafting and direct stabilization of
Role of joint torques generated in an spondylolysis by means of a hook screw. Arch 38. Roy S, Caine D, Singer KM: Stress changes
optimised Yurchenko layout vault. Sports Orthop Trauma Surg 1984;103:175-178. of the distal radial epiphysis in young
Biomech 2003;2:177-190. gymnasts: A report of twenty-one cases
24. Gillis CC, Eichholz K, Thoman WJ, Fessler and a review of the literature. Am J Sports
11. Koh TJ, Grabiner MD, Weiker GG: Technique RG: A minimally invasive approach to Med 1985;13:301-308.
and ground reaction forces in the back defects of the pars interarticularis:
handspring. Am J Sports Med 1992;20:61-66. Restoring function in competitive athletes. 39. Kirialanis P, Malliou P, Beneka A,
Clin Neurol Neurosurg 2015;139:29-34. Giannakopoulos K: Occurrence of acute
12. Kruse D, Lemmen B: Spine injuries in the lower limb injuries in artistic gymnasts in
sport of gymnastics. Curr Sports Med Rep 25. Caplan J, Julien TP, Michelson J, Neviaser relation to event and exercise phase. Br J
2009;8:20-28. RJ: Multidirectional instability of the Sports Med 2003;37:137-139.
July 1, 2019, Vol 27, No 13 467
Research Article
Business Modeling of Orthopaedic

Trauma in the Emergency
Department: An Untapped
Revenue Stream
Abstract
Robert C. Jacobs, MD Introduction: Emergency departments (EDs) and emergency
Tinh T. Huynh, MD medicine and orthopaedic residencies can be faced with financial
challenges while caring for patients. Procedures performed by
S. Elizabeth Ames, MD
residents are a potentially viable source of revenue that may make
Mitchel H. Tsai, MMM, MD orthopaedic coverage of the ED a financially viable service line.
James D. Michelson, MD Methods: A custom text-mining program was created and validated,
which allowed evaluation of all orthopaedic resident notes.
Procedures performed in the ED were quantified, allowing for the
calculation of professional fee billing data. The patients with distal
radius fractures were followed after fracture reduction through final
outpatient clinic follow-up to identify additional professional fee billing.
Results: Over a 1-year period, more than $445,000 in uncaptured
professional fees charged was identified in the 12 most common
Current Procedural Terminology codes for splint application and
fracture reduction in the ED. More than $395,000 of outpatient
professional revenue was received for patients who had reduction of
From the Department of Orthopaedics distal radius fractures in the ED.
and Rehabilitation, University of Conclusion: A notable, previously unrecognized and uncaptured
Vermont Larner College of Medicine,
source of revenue was identified and quantified. Professional fee
Burlington, VT (Dr. Jacobs, Dr. Ames,
and Dr. Michelson), the University of billing for distal radius fracture reduction in the ED did not have a
Vermont Larner College of Medicine negative effect on outpatient professional fee revenue received for
(Dr. Huynh), and the Departments of
Anesthesiology and Surgery
these patients.
(Dr. Tsai).
Correspondence to Dr. Jacobs:

Rcjacobs87@yahoo.com

immediate family member has
received anything of value from or has
T he regionalization of trauma
care has been shown to reduce
trauma-related morbidity and mor-
history and physical examination, rec-
ommendation of appropriate imaging,
a diagnosis of the injury, and a man-
stock or stock options held in a tality.1-6 To enhance operational agement plan. Residents perform sepa-
related directly or indirectly to the
function, many academic medical rate procedures as a part of this service,
subject of this article: Dr. Jacobs, centers have integrated an ortho- including laceration repairs, minor
Dr. Huynh, Dr. Ames, Dr. Tsai, and paedic trauma service (OTS), which surgical procedures, reduction of dis-
Dr. Michelson. often provides care in the emergency located joints, and reduction/splinting
J Am Acad Orthop Surg 2019;27: department (ED). Here, the residents of fractures. Although an OTS has
e612-e621 function as consulting providers under been proven to increase hospital
DOI: 10.5435/JAAOS-D-17-00742 the direct supervision of emergency revenue,7 many trauma centers have
medicine and orthopaedic attending experienced a decline in financial
Academy of Orthopaedic Surgeons. physicians who attest to the work viability against the current back-
performed. A consultation includes a drop of healthcare change.8-19
e612 Journal of the American Academy of Orthopaedic Surgeons
Robert C. Jacobs, MD, et al
Billing regulations continue to im- provide insight for the development was queried for all ED notes written
pede an institution’s ability to accu- of a viable business plan to support by 15 orthopaedic residents for the
rately capture the work performed at the costs associated with additional period July 1, 2013 through June 30,
some institutions. The documentation attending coverage. 2014. Notes identified for manual
of attending physician oversight for This study identifies the clinical review were one of three types: con-
orthopaedic residents consulting activities of orthopaedic residents in sult, procedure, or admission history
in the ED may be problematic for an ED setting to quantify the effect of and physical notes. Each note was
several reasons. Reimbursement de- incomplete direct supervision on the evaluated during the manual review
pends heavily on the use of Current capture of procedural coding and for any procedures and billable ser-
Procedural Terminology (CPT) codes billing. The hypothesis of the study vices performed by the orthopaedic
that are valued based on the severity was that a substantial amount of resident writing the note. Once the
of the patient’s condition and the revenue is generated through the manual review was completed, a
complexity of the care provided.20 clinical service provided by ortho- custom text-mining algorithm was
Ideally, a consultation performed by paedic residents in the ED, which is written using a text search program
an orthopaedic resident in the ED not captured by the billing process. (dtSearch, V 7.81; dtSearch) to
includes a patient evaluation and Using a custom text-mining algo- enable any EMR note to be searched
management (E&M), which may rithm, this study identifies clinical for certain words or phrases used in
require a procedure (ie, fracture procedures within routine clinical the description of procedures per-
reduction) representing clear, billable notes written by orthopaedic residents formed in the ED by orthopaedic
work. The attending orthopaedic in the ED. In conjunction with the residents. This program underwent
surgeon can bill for the E&M of the billing records from the hospital and multiple iterations, with subsequent
problem and a professional fee using medical system, this study estimates the manual note review cross-checks
CPT codes based on the procedure or potential revenue opportunity of the using 1-month blocks of orthopae-
procedures performed.21-23 Similarly, residents’ clinical activities. dic resident notes from the ED used
an ED physician may bill for a por- The specific subgoals of the study to improve the sensitivity and spec-
tion of the professional fee if they were to ificity of the program.
participated in the resident-performed (1) demonstrate the accuracy of a The final algorithm for the text-
procedure in the absence of an or- custom text-mining algorithm in mining program was validated
thopaedic attending. However, the ED identifying all notes written by against the manual review of all pa-
provider may not bill for the ortho- the orthopaedic residents in the tients treated with CPT code 25605
paedic resident’s E&M because of ED and the number of proce- (closed reduction and manipulation
concurrent ED E&M billing. Finally, dures they performed, of distal radius fracture, Colles’ type)
neither the CPT code professional fee (2) identify the percentage of billable and the hospital EMR records for
nor the orthopaedic E&M consult fee procedures that were captured each patient. CPT code 25605 was
can be billed when resident supervi- and billed by the hospital-based chosen because it is the most com-
sion is not documented by any at- coding process, and mon orthopaedic fracture reduction
tending physician. The latter two (3) identify the total uncaptured performed at our institution. Calcu-
scenarios engender notable financial professional fees from these lation of positive and negative pre-
shortfalls because of unrealized billing orthopaedic procedures, in the dictive values (PPVs and NPVs) and
in an academic setting where clinical context of the entire episode of sensitivity and specificity for the
revenue generated by residents can care for the fracture, including algorithm was performed using man-
cover the cost of training.24,25 their follow-up in the outpatient ual review of all the notes identified by
Despite the aspirational educa- clinics. text mining or of patients identified
tional goals of a residency program, it by the hospital-based coding.
may not be possible for an attending Separately, the hospital’s billing
physician to be present for the pro- Methods records from the ED for the 12-
cedural portion of patient care. month study period were queried
However, understanding the recov- The level I trauma hospital’s elec- to determine the payer mix (Com-
ered and lost revenue streams should tronic medical record (EMR), Epic, mercial, Medicaid, Medicare, Other,
An analysis of emergency department (ED) professional fee billing for orthopaedic trauma in a level I academic medical center. Included is
an evaluation of distal radius fractures seen in the ED by orthopaedic residents through final outpatient follow-up for patients receiving
surgical and nonsurgical care. Total revenue calculated for each pathway from the ED to final outpatient disposition has been calculated.
July 1, 2019, Vol 27, No 13 e613
Business Modeling of Orthopaedic Trauma
Figure 1
Flow diagram for CPT code 25605 from ED to outpatient clinic. CPT = Current Procedural Terminology, ED = emergency
department, UVM = University of Vermont
Self-Pay, Workers Comp) for the showing the patient’s progress from clinic, those getting surgery, those not
population being evaluated. The or- time of ED admission, diagnosis, following up in the medical center,
thopaedic resident notes were used reduction/procedural intervention, along with the associated professional
to generate a list of seven most and follow-up in clinic for ongoing fees billed for their CPT code 25605.
common fracture reduction CPT nonsurgical versus surgical manage- The total revenue for each pathway
codes and five most common splint ment. This allowed the evaluation of was used as the outcome utility of the
application CPT codes during the total procedural professional fee decision tree.
study period. The professional fee billing in the inpatient and outpa-
billed to the patient for each of the tient setting for all orthopaedic care.
12 selected CPT codes was identi- The most common orthopaedic CPT Results
fied. The reimbursement rates for code performed in the ED (25605)
each payer mix category were iden- was used as a base model. Each Distal Radius Fractures
tified for the 12 selected CPT codes. patient identified within the 12-month To validate the results generated by
A weighted average for each pro- period was followed from ED diag- the text-mining program, all notes
fessional fee was calculated for each nosis to final follow-up in the out- were identified containing the key
CPT code. The weighted average for patient clinic (even if the time of words for a distal radius fracture
each CPT code was multiplied by the follow-up carried beyond the period and reduction (CPT code 25605).
number of CPT codes in each of the of the initial diagnosis). The period for the study was the
12 categories to estimate the total Using the flow diagram, a decision 12 months (from July 1, 2013 to June
dollar amounts billed, collected, and analysis tree was constructed with 30, 2014). A total of 1,604 notes were
uncollected. decision-making nodes placed at identified, of which 230 were de-
To get a broader view of the bill- points where decisions between alter- tected as having fractures requiring
ing for fracture management that natives (ie, surgical versus nonsurgical reduction and 1,374 as not involving
extends after the ED visit, a decision care) were made. Each branch was distal radius fractures. Manual
tree (DPL, V7.0; Syncopation Soft- populated by billing data for the per- review of these notes for the perfor-
ware) (Figure 1) was constructed, centage of patients following up in mance of a fracture reduction showed
220 true positives and 1,384 true neg- Table 1

atives. Compared with the benchmark
Validation of Text-mining Algorithm
of manual chart review for distal radius
fractures, the text-mining algorithm Manual Review Technique
exhibited a sensitivity of 100%, speci- Variable True Positives True Negatives Total
ficity of 99.7%, PPV of 95.6%, and
NPV of 100% (Table 1). Data-mining program — — —
The 220 patients treated under the Positive 220 10 230
CPT code 25605 for distal radial frac- Negative 0 1,374 1,374
tures were examined using a decision Total 220 1,384 1,604
tree to determine the proportion of the Sensitivity 100%
billing that occurs after the ED visit, in Specificity 99.7%
the form of either office treatment or Positive predictive value 95.6%
surgery. For this analysis, the profes- Negative predictive value 100%
sional fee for reduction of the CPT code
CPT = Current Procedural Terminology, ED = emergency department
25605 in the ED was added to any fees The notes used for analysis were those written by orthopaedic residents for patients within the
incurred in the outpatient setting ED between July 1, 2013 and June 30, 2014, which involved management of a distal radius
(either clinic or outpatient surgical). fracture (CPT code 25605).
Analysis of the actual professional fees

collected (not the weighted-average
Figure 2
Flow diagram of ED to outpatient billing, with patient numbers and total reimbursed dollars. ED = emergency department,
UVM = University of Vermont
July 1, 2019, Vol 27, No 13 e615
Table 2
Realized Professional Fee Revenue From Outpatient Management of Distal Radius Fractures Seen in the ED (Also
See Figures 1 and 2)
Box No. of Fractures CPT Codes $ E&M Codes $ Cast Codes $ Total Revenue $ $ per Fracture
5 5 $16,125 $2,104 $218 $18,447 $3,689

7 6 $2,600 $1,311 $925 $4,836 $806
9 5 $0 $0 $0 $0 $0
13 77 $269,775 $25,101 $4,577 $299,453 $3,889
15 95 $20,369 $33,840 $18,963 $73,172 $770
17 32 $0 $0 $0 $0 $0
Total 220 $308,869 $62,356 $24,683 $395,908 $1,800
CPT = Current Procedural Terminology, E&M = evaluation and management, ED = emergency department
The notes used for analysis were those written by orthopaedic residents for patients in the ED between July 1, 2013 and June 30, 2014. Column 1
(box) directly correlates to Figures 1 and 2 and should be used together for reference.
Table 3
Percentage of the 5 Most Common Splint Applications That Were Identified and Billed by the Hospital-based Coding
System
CPT Code Splint Procedure Unbilled Billed Total % Unbilled
29105 Long arm 101 17 118 86

29125 Short arm 521 146 667 78
29130 Finger 132 50 182 72
29505 Long leg 28 2 30 93
29515 Short leg 203 43 246 82
Total 985 258 1,243 79

The notes used for analysis were those written by orthopaedic residents for patients in the ED between July 1, 2013 and June 30, 2014.
Table 4
Percentage of the 7 Most Common Fracture Reductions That Were Identified and Billed by the Hospital-based
Coding System
CPT Code Reduction Procedure Unbilled Billed Total % Unbilled
25605 Distal radius 200 20 220 91

26605 Metacarpal 50 0 50 100
27752 Tibial shaft 23 8 31 74
27788 Distal fibula 21 3 24 88
27810 Bimalleolar ankle 22 3 25 88
27818 Trimalleolar ankle 18 2 20 90
27825 Distal tibia 12 4 16 75
Total 346 40 386 90

estimate) associated with care to final $395,000. Most of this came from an outpatient basis, totaling approx-
management of the distal radius frac- the CPT codes for surgical manage- imately $299,000 (Figure 2). Of the
tures showed a total of approximately ment of the distal radius fracture on total professional fees collected,
Table 5
Expected and Realized Professional Fee Revenue From Splints Applied in the ED
Professional
Fee Billed? Total Possible Professional
Professional Charges
CPT Code Procedure No Yes Total % Unbilled Charges Unbilled
29105 Long arm 101 17 118 86 $95,739.51 $81,946.53

29125 Short arm 521 146 667 78 $88,352.13 $69,012.68
29130 Finger 132 50 182 72 $16,760.36 $12,155.87
29505 Long leg 28 2 30 93 $1,823.66 $1,702.08
29515 Short leg 203 43 246 83 $30,796.87 $25,413.68
Total 985 258 1,243 79 $233,472.53 $190,230.84

Table 6
Expected and Realized Professional Fee Revenue From Fracture Reductions Performed in the ED
Professional
Fee Billed? Total Possible
Professional Professional
CPT Code Procedure No Yes Total % Unbilled Charges Charges Unbilled
25605 Distal radius 200 20 220 91 $178,291.30 $162,083.00

26605 Metacarpal 50 0 50 100 $23,945.16 $23,945.16
27752 Tibial shaft 23 8 31 74 $24,762.39 $18,372.09
27788 Distal fibula 21 3 24 88 $15,400.09 $13,475.08
27810 Bimalleolar ankle 22 3 25 88 $17,403.10 $15,314.73
27818 Trimalleolar ankle 18 2 20 90 $14,390.89 $12,951.80
27825 Distal tibia 12 4 16 75 $12,981.63 $9,736.22
Total 346 40 386 90 $287,174.56 $255,878.08

approximately 6% resulted from percent of these procedures had no Medicare reimbursement rates for
casting, 16% from E&M codes, professional fee associated with each CPT code to 100% relative to
and 78% from CPT (or related) their billing (Table 3). For the seven the other payer sources, the weighted
codes (Table 2). most common fracture reduction average reimbursement rate could be
CPT codes, 391 procedures were calculated (148.7% Medicare rates)
performed; 90% had no professional and expected payment average for
All Fractures Managed in the fees billed in association with the each of the 12 CPT codes (Tables 5
Emergency Department fracture reduction procedure (Table 4). and 6).
One thousand two hundred forty- The billing data for all patients seen If every splinting procedure had
three procedures were performed in in the ED for the same 12-month appropriate attending attestation, a
the 12-month period (July 1, 2013 to period were also analyzed and had total of approximately $233,000 pro-
June 30, 2014), which were identi- the following payer mix: 36% Com- fessional fees would have been reim-
fied using the data-mining program mercial, 22% Medicaid, 29% Medi- bursed for the five splinting CPT codes.
for the five most common splint care, 3% Other, 7% Self-pay, and Roughly $43,000 in professional fees
application CPT codes. Seventy-nine 2% Workers compensation. Setting were actually billed and $190,000
July 1, 2019, Vol 27, No 13 e617
unbilled (Table 5). If every fracture the American Academy of Orthopae- notable number of clinical encounters
reduction procedure was appropri- dic Surgeons) included a new section did not have the appropriate attending
ately documented and billed, a total on practice management. The Accred- documentation of the procedure to
of approximately $287,000 would itation Council of Graduate Medical permit professional fee billing.
have been reimbursed for the seven Education stipulates that under This study highlights the impor-
reduction CPT codes. Approximately appropriate supervision, residency tance of a system-based perspective
$31,000 professional fees was actually programs should provide education of orthopaedic workflow starting
billed and $255,000 unbilled (Table that “ensures each resident’s devel- from the ED. Following the entire
6). The 1-year total of unbilled col- opment of the skills, knowledge, and episode of care provides a clearer
lections for these 12 frequent CPT attitudes required to enter the unsu- view of patient flow and a better
codes was approximately $445,000 pervised practice of medicine; and understanding of revenue streams.
(Tables 5 and 6). establishes a foundation for continued Restricting the study to billing from
professional growth.”32 Similarly, the the ED visit alone would have missed
Reno Orthopaedic Center Trauma more than 85% of all revenue for
Discussion Fellowship program also recognizes each distal radius fracture. Although
this to be a notable part of ortho- improved billing in the ED may
The OTS is one of several critical paedic education and has designed a decrease E&M coding at subsequent
service lines required for a hospital to specific curriculum to address the clinic visits, our study suggests that
maintain level I accreditation. How- multifaceted topic of the orthopaedic this would not have an overall dele-
ever, management of orthopaedic practice management.33-35 At our terious effect on subspecialty billing.
trauma in the ED is a financial bur- institution, a specific “Business of Patients with nonsurgically managed
den, resulting in 14% losses of oper- Orthopaedics” didactic curriculum distal radius fractures generated an
ating expenses for US level I trauma exists, which covers topics such as average per patient revenue of $806
centers.19 This is, in part, because of operating room efficiency, enterprise when their treatment was attested by
the necessity to maintain a capacity- finance, medical coding and billing, an ED attending, compared with an
based trauma service that emphases personal finance, and contract nego- average per patient revenue of $770
personnel, equipment and operating tiating strategies. Senior hospital when their treatment was not docu-
room availability over economic managers, administrators, community mented by an ED attending (Table 2).
considerations to permit adequate professionals, and private practice Surgically treated patients generated
response to major trauma. Revenue orthopaedic surgeons supplement the 4.1 times the revenue than did non-
optimization is also hindered by lectureship curriculum. The current surgically treated patients.
incomplete documentation and in- study is a direct result of the Business As shown in Table 2, 82 patients
adequate reimbursement.26-30 In of Orthopaedics curriculum. (37%) of the 220 went on to be
essence, these financial, operational, Operational and clinical processes treated with open reduction and
and organizational forces combine to that track patient encounters in the internal fixation and all 183 patients
make it difficult for trauma centers in ED can help track and recover lost with distal radius fractures (83%) of
today’s healthcare environment. reimbursements. The current system the 220 were treated with ED closed
In 2012, Appleton et al31 demon- in our ED is similar to that in many reduction (by an orthopaedic resi-
strated that 31% of total OTS rev- academic centers, without a dedicated dent) followed up at our outpatient
enue was attributable to nonsurgical orthopaedic surgeon to staff only clinic. Our patients are referred to
fracture management (eg, office vis- inpatient and ED orthopaedic con- our affiliated clinic through our
its, nonsurgical treatment fracture sults. Typically, an attending surgeon EMR and secure e-mail server where
billing, and orthopaedic consultations). will staff resident work while con- they are contacted by our schedulers
This clinical pathway represents nota- currently working in clinic or oper- the following day for an appropriate
ble educational opportunities and is ating. Any immediate or critical patient appointment. This may represent a
increasingly recognized as an important care needs are performed by attending unique practice environment because
component of an orthopaedic resi- ED physicians. Although this frag- no additional orthopaedic practices
dency curriculum. An increasing focus mented supervision system allows exist in the local market and follow-
is observed on practice management at orthopaedic residents a fair degree of up rates may not be as high in other
the American Academy of Orthopae- autonomy needed for educational pur- practice environments.
dic Surgeons national conference, and poses, it can lead to notable gaps in This practice model also may not be
in 2017, the orthopaedic resident in- clinical documentation at the attending applicable to the private practice set-
training examination (administered by level. The current study shows that a ting because of the use of residents
instead of midlevel providers for clini- and admissions are staffed with the existing information, which does not
cal care, including procedural reduc- appropriate service or on-call ortho- rely on the imperfect expertise of
tions. Resident physicians require paedic attending. A patient with a administrative coders, to recognize
attending level direct supervision and distal radius fracture that is reduced opportunities to improve attending
attestation for billing under Centers for by an orthopaedic resident and at- oversight of the residents in a way
Medicare and Medicaid Services re- tested by an ED provider is billed with that does not hamper their workflow
quirements; however, physician assis- an ED E&M code with a 54 modi- or education. The text-mining meth-
tants may bill for procedural and fier. When the patient subsequently odology can also provide orthopae-
consultative specialty services without follows up in the orthopaedic out- dic residency directors a method to
direct supervision. Midlevel providers patient clinic, the initial visit and any cross-check resident case records
also cannot attest to consults or pro- subsequent nonsurgical visits are with respect to nonsurgical proce-
cedures performed by resident physi- billed as an established patient with dures performed in the ED. This
cians. The reimbursement level for appropriate established patient level functionality should enable the
physician assistant billing in this sce- of service E&M coding with a 55 orthopaedic residency to verify the
nario typically is 80% of what an modifier up to the 90-day global Accreditation Council of Graduate
attending orthopaedic physician period under the CPT code 25605 Medical Education fracture care
would bill. This loss of potential that began at the time of the ED milestones for the residents.
revenue is balanced by the need for reduction. This process changes In closing, this methodology pro-
resident education in the academic when the decision for surgery is vides an opportunity for a notable
setting. made. In this scenario, the visit is revenue enhancement, which other-
The compensation models for the billed as the established E&M level wise represents care that is currently
orthopaedic and ED attendings also of service with a 57 modifier for the being provided for free. In a level I
merit discussion. Both departments surgical decision change to the plan trauma center that is trying to
are employed by the hospital and are of care. After surgery, clinic follow- maintain a fundamentally economi-
paid on a metric that reimburses up visits are billed under 90-day cally unsound, but clinically neces-
approximately 50% base salary and global period for the surgical CPT sary, trauma service, this revenue
50% relative value units (RVUs) gen- code performed. stream could help hospitals sustain a
erated. Specifically related to the Finally, if a patient is evaluated and vital resource for the community.
orthopaedic attendings, the RVU por- reduced by the orthopaedic resident Counterbalancing this observation is
tion of the compensation is negotiated and no ED or orthopaedic attending the recognition that costs are associ-
as a productivity target for the attestation is performed (for the ated with full-time staffing of an
upcoming year and is usually based on fracture reduction), they can then be orthopaedic attending in the ED to
the expected surgical/clinical produc- billed as a new patient visit level provide more oversight and docu-
tivity. Some incentive exists for the E&M with appropriate level of ser- mentation. An assessment of the
orthopaedic attending to directly staff vice when the patient follows up in balance between the revenue gener-
the procedural portion of the consult; our orthopaedic clinic. The ED pro- ated against the personnel costs
however, this is somewhat diminished vider can still bill for the ED E&M would be necessary to determine the
because the RVUs are not directly portion of the patient’s care. Because economic viability of changing the
reimbursed to the provider on a 1:1 no CPT code for distal radius fracture attending coverage practice in the ED
basis. The results of this study are lim- reduction with manipulation (25605) based on institutional factors, the
ited to the current practice model and was billed in the ED, the CPT code local community setting, and varied
may not be applicable to non-hospital 25600 (closed management of distal patient populations.
employment models; they will also radius fracture without manipulation)
vary based on the compensation can be billed instead. Subsequent out-
structure of each practice. patient clinic visits can then be billed Study Limitations
Because the compensation strategy under that CPT code under that 90-day
for attending orthopaedic surgeons global start date. The expected revenue was calculated
does not directly incentivize them to This study has demonstrated that using weighted averages for each of
staff reductions from home, out- it is possible to use a text-mining the 12 CPT codes. Given the vari-
patients seen in the ED are more often algorithm on primary clinical docu- ability of orthopaedic practice, this
staffed by ED attendings for proce- mentation to accurately identify or- may not reflect another institution’s
dures (consult attestations are physi- thopaedic procedures occurring in practice. However, this methodology
cian dependent). Inpatient consults the ED. It is possible to leverage this establishes a framework to compare
July 1, 2019, Vol 27, No 13 e619
different periods using CPT codes systems development. JAMA 1985;254: 16. Selzer D, Gomez G, Jacobson L,
1059-1063. Wischmeyer T, Sood R, Broadie T:
as a surrogate for payer reimburse- Public hospital-based level I trauma
ments. It should be noted that the 2. Guss DA, Meyer FT, Neuman TS, et al: centers: Financial survival in the
The impact of a regionalized trauma new millennium. J Trauma 2001;51:
actual collected revenue amounts system on trauma care in San Diego 301-307.
were used, not an estimated weighted County. Ann Emerg Med 1989;18:
average for patients with distal radius 1141-1145. 17. Thal ER, Rochon RB: Inner-city trauma
centers: Financial burdens or community
fractures. This framework allowed 3. Kane G, Wheeler NC, Cook S, et al: Impact saviors? Surg Clin North Am 1991;71:
of the Los Angeles county trauma system on
for more accurate and easier deter- 209-219.
the survival of seriously injured patients.
mination of dollar amounts at each J Trauma 1992;32:576-583. 18. Taheri PA, Butz DA, Lottenberg L,
decision-making node in the decision 4. Shackford SR, Hollingworth-Fridlund P,
Clawson A, Flint LM: The cost of trauma
center readiness. Am J Surg 2004;187:7-13.
tree (Figure 2) because multiple types Cooper GF, Eastman AB: The effect of
of surgical CPT codes were billed and regionalization upon the quality of trauma 19. U.S. Trauma Center Crisis: Lost in the
care as assessed by concurrent audit before Scramble for Terror Resources. National
collected in managing the initial and after institution of a trauma system: Foundation for Trauma Care, Irvine, CA,
injury CPT code 25605. Finally, the A preliminary report. J Trauma 1986;26: 2004.
CPT code 25605 is a common injury 812-820.
20. Fakhry SM: Billing, coding, and
managed in the ED with well-defined 5. West JG, Cales RH, Gazzaniga AB: documentation in the critical care
Impact of regionalization: The Orange environment. Surg Clin North Am 2000;80:
and limited management options. It
County experience. Arch Surg 1983;118: 1067-1083.
was chosen to verify the sensitivity, 740-744.
21. Barnes SL, Robinson BR, Richards JT: The
specificity, PPV, and NPV for all 6. Vali Y, Rashidian A, Jalili M, Omidvari devil is in the details: Maximizing revenue
patients who were evaluated for the AH, Jeddian A: Effectiveness of for daily trauma care. Surgery 2008;144:
inpatient/outpatient revenue review. regionalization of trauma care services: A 670-675.
systematic review. Public Health 2017;146:
Further research will be needed to 92-107. 22. Kirschner C: CPT 2017 Professional
extend these findings to other in- Edition. American Medical Association,
7. Harris MB, Cayen B: The organizational Chicago, IL, 2017, vol. 2017.
juries commonly managed in the ED. and financial viability of an orthopedic
trauma service. J Trauma 2009;67: 23. Ingenix: Optum Learning: Understanding
1359-1370. E/M Coding 2013. Optum, West Valley
City, UT, 2013.
Conclusion 8. Brackstone M, Doig GS, Girotti MJ:
24. Turner BC, Tsai MH, Black IH, Mathews
Surgical case costing: Trauma is
underfunded according to resource DM, Adams DC: Observations: Clinical
A text-mining program was validated intensity weights. Can J Surg 2002;45: revenue directly attributable to anesthesiology
and then used to identify and quantify a 57-62. residents. J Grad Med Educ 2014;6:384.
substantial potential additional source 9. Cohen JW, Krauss NA: Spending and 25. Govindarajan V: Delivering world-class
health care, affordably. Harvard Business
of revenue. Hospitals, orthopaedic de- service use among people with the fifteen
most costly medical conditions, 1997. Review 2013;91:117-122.
partments, and EDs can better eluci- Health Aff (Millwood) 2003;22:129-138.
26. Fakhry SM, Potter C, Crain W, Maier R:
date the staffing models and clinical Survey of national usage of trauma
10. Eastman AB, Rice CL, Bishop G,
workflows that make orthopaedic Richardson JD: An analysis of the critical response charge codes: An opportunity for
coverage of the ED a sustainable, problem of trauma center reimbursement. enhanced trauma center revenue. J Trauma
J Trauma 1991;31:920-925. 2009;67:1352-1358.
financially viable mission by incor-
porating the insights gained from this 11. Eastman AB, Bishop GS, Walsh JC, 27. Dezfuli B, Smith JL: Level of billing as a
Richardson JD, Rice CL: The economic function of resident documentation and
study. Future applications include orthopedic subspecialty at an academic
status of trauma centers on the eve of health
identifying the most cost-effective care reform. J Trauma 1994;36:835-844. multispecialty orthopedic surgery practice.
Orthopedics 2012;35:e1655-e1658.
clinical pathways for patients with 12. Enderson BL: Can trauma surgeons survive
specific fractures and the design of business medicine? J Trauma 2003;55: 28. Kapa S, Beckman TJ, Cha SS, et al: A
215-221. reliable billing method for internal medicine
operational processes to maintain resident clinics: Financial implications for
financial viability of OTS lines. 13. Lanzarotti S, Cook CS, Porter JM, Judkins an academic medical center. J Grad Med
DG, Williams MD: The cost of trauma. Am Educ 2010;2:181-187.
Surg 2003;69:766-770.
29. Evans DV, Cawse-Lucas J, Ruiz DR, Allcut
14. Miller TR, Levy DT: The effect of regional EA, Andrilla CH, Norris T: Family
References trauma care systems on costs. Arch Surg medicine resident billing and lost revenue: A
1995;130:188-193. regional cross-sectional study. Fam Med
References printed in bold type are 2015;47:175-181.
15. Rodriguez JL, Peterson DJ, Muehlstedt SG,
those published within the past 5 Zera RT, West MA, Bubrick MP: The 30. Andreae MC, Dunham K, Freed GL:
years. impact of managed care and current Inadequate training in billing and coding
governmental policies on an Urban as perceived by recent pediatric
1. Cales RH, Trunkey DD: Preventable Academic Health Care Center. Surgery graduates. Clin Pediatr (Phila) 2009;48:
trauma deaths: A review of trauma care 2001;130:539-544. 939-944.
31. Appleton P, Chacko A, Rodriguez EK: ProgramRequirements/CPRs_Section% 34. Althausen PL, Mead L: Practice
Financial implications of nonoperative 20VI_with-Background-and-Intent_2017- management. J Orthop Trauma 2014;
fracture care at an academic trauma center. 01.pdf. Accessed July 30, 2018. 28(7 suppl):S12-S17.
J Orthop Trauma 2012;26:617-619.
33. Althausen PL, Bray TJ, Hill AD: Reno 35. Bray TJ, Althausen PL: The Reno
32. ACGME: Common Program Requirements Orthopaedic Trauma Fellowship business Orthopaedic Clinic trauma business
Section VI with Bacground and Intent. 2017. curriculum. J Orthop Trauma 2014; curriculum: Introduction. J Orthop
http://www.acgme.org/Portals/0/PFAssets/ 28(7 suppl):S3-S11. Trauma 2014;28(7 suppl):S1-S2.
July 1, 2019, Vol 27, No 13 e621
Research Article
Clinical Approach in Youth Sports

Medicine: Patients’ and Guardians’
Desired Characteristics in Sports
Medicine Surgeons
Abstract
Jennifer J. Beck, MD Introduction: Adolescent athletes’ and their guardians’ preferences
Martha M. Murray, MD for sports medicine surgeon characteristics are unknown.
Methods: Unique, anonymous surveys regarding preferences in
Melissa A. Christino, MD
characteristics of sports medicine surgeons were given to both
patients (aged 10 to 18 years) and their guardians before being seen
by a sports medicine surgeon.
Results: Patients and their guardians reported shared decision
making as the most important surgeon characteristic, followed by
understanding patients’ sports and goals. A higher percentage of
male patients than female patients had a surgeon sex preference (P =
0.005); however, for both the groups, this ranked lower than other
surgeon qualities. Nearly all respondents reported that the adolescent
patient should be involved in medical decision making; yet, physician
selection was determined by the guardian 65% of the time.
From the Orthopedic Institute for Conclusions: Both adolescent patients and their guardians reported
Children/UCLA, Los Angeles, CA shared decision making and understanding patients’ sports and goals as
(Dr. Beck), the Boston Childrens
Hospital, Boston, MA (Dr. Murray), surgeon qualities that were more important to them than surgeon sex.
and the Childrens Hospital of Atlanta, Level of Evidence: Level IV, cross-sectional study
Atlanta, GA (Dr. Christino).
Correspondence to Dr. Beck:
jjbeck@mednet.ucla.edu
Dr. Beck or an immediate family
member serves as a board member,
P atient experiences and outcomes
have become important metrics
within the healthcare system and
practice, showing no notable pref-
erences toward surgeon sex, age,
religion, or race, but those who had a
owner, officer, or committee member
of the Pediatric Orthopaedic Society of have become targets of healthcare preference tended to choose sur-
North America. Dr. Murray or an improvement projects in recent years. geons of the same demographic as
immediate family member serves as a
board member, owner, officer, or
With a plentiful number of Internet themselves.11
committee member of American sites allowing patients to rate their In addition to patient preferences,
Academy of Orthopaedic Surgeons. physician and hospital experiences, the experience of medical decision
Dr. Christino or an immediate family little is known about what prefer- making has evolved from being pri-
member has stock or stock options
held in MIACH Orthopaedics and
ences patients have in regard to these marily physician driven to a shared
serves as a paid consultant to entities, particularly with respect to decision-making model between the
Musculoskeletal Transplant physician characteristics. Previous physician and the patient. This has
Foundation. studies have looked at the effects of been more extensively discussed
J Am Acad Orthop Surg 2019;27: physician sex1-9 and attire10 on within orthopaedic surgery litera-
479-485 patient satisfaction. These studies ture,12-14 with application to specific
DOI: 10.5435/JAAOS-D-18-00263 cross multiple surgical and medical situations such as upper extremity
specialties. The only study involving traumatic versus nontraumatic con-
Academy of Orthopaedic Surgeons. orthopaedic surgery is the survey of ditions,15 carpal tunnel syndrome
500 adult patients in a large urban management,16,17 and stable ankle
July 1, 2019, Vol 27, No 13 479
Clinical Approach in Youth Sports Medicine
fractures.18 Only one study evalu- veys were given to both patients and compare differences between patient
ated differences in the preferences of their guardians before being seen by and guardian responses using SPSS.
patients and parents regarding ado- one of seven sports medicine sur- Statistical significance was set at P ,
lescent idiopathic scoliosis surgery, geons in the clinical setting of a ter- 0.05.
which found that patients and pa- tiary care referral center. For
rents often did not agree on reasons 3 months around the time of the
for surgery or on concerns or ex- study, the volume of potentially eli- Results
pectations after surgery.19 gible patients was 361, 321, and 315
Sports participation and injury patients within this clinic. Surveys In a 1 month study period, 280 ado-
rates are increasing among adoles- were marked with a unique identifier lescent athletes filled out anonymous
cent athletes,20 resulting in increased to allow matching of anonymous paper surveys regarding preferences
appointments with sports medicine patient and guardian surveys. in sports medicine surgeons. Adoles-
surgeons. At a time when body self- Inclusion criteria were defined as cent patients’ average age was 14.7 6
image and independence are devel- patients aged 10 to 18 years, with a 2.2 years (age groups: 10 to 14
oping, adolescent athletes and their primary guardian in attendance at years; N = 119 and 15 to 18 years; N =
families are often faced with notable their appointment. Exclusion criteria 161), with responses given by 105
medical decisions that affect their included non-English readers and male patients (38%) and 175 female
quality of life and ability to participate patients without primary guardians patients (62%). No notable differ-
in beloved activities. Very little is at their visit. Eligible patients and ences were observed when patients
known about adolescent preferences guardians were offered enrollment in were divided into two age groups, so
for their care. A large study on ado- this study on check-in to their office results represent entire patient par-
lescent preferences for healthcare visit. Participation was voluntary. ticipation population. In the same
providers, without reference to a Answers were anonymous. 1 month study period, 256 adult
specific medical specialty, showed that A unique survey was developed guardians filled out matched corre-
physician sex was a notable variable with assistance of an academic survey sponding surveys regarding prefer-
in adolescent preferences21 and one in methodologist to ensure quality. ences in sports medicine surgeons.
seven adolescents show fear and Survey questions included basic Based on the study protocol, only
anxiety toward the physicians.22 demographics, referral information, one guardian was allowed to com-
Once an injury occurs, compliance reason for visit, preferred surgeon plete the survey per adolescent pa-
and adherence to medical treatment characteristics, and the most impor- tient. If more than one guardian was
can be affected by adolescent athletes’ tant surgeon characteristic, as well as present, those guardians determined
relationship with their sports medi- questions related to family medical who completed the survey. Guard-
cine surgeons. To date, no study decision making. Patient and guardians’ average age was 47.4 6 5.5
exists which has investigated ado- ian surveys were nearly identical and years, with responses given by 46
lescent athletes’ and their guardians’ are shown in Figures 1 and 2, male patients (18%) and 206 female
sports medicine surgeon preferences, respectively. Given the lack of med- patients (82%).
how they may differ from each other, ical terminology in the study, the Characteristics other than surgeon
and the role of the adolescent patient survey methodologist determined sex were reported to be very impor-
in shared decision making in these that the age range of 10 to 18 years tant to adolescents, particularly for
situations. In this study, we hypoth- was appropriate for the education female patients (see Supplement 1,
esized that a higher percentage of and reading level of this survey. Pa- Supplemental Digital Content 1,
patients would rank surgeon char- tients and guardians completed sur- http://links.lww.com/JAAOS/A222).
acteristics, including shared decision veys simultaneously, and study A greater percentage of female pa-
making, as very important than participants were allowed to ask tients than male patients reported
surgeon sex. questions of their guardians or the that surgeon compassion (72% ver-
research staff during survey com- sus 47%; P , 0.001) and being a
pletion. Patient and guardian surveys good listener (82% versus 61%; P ,
Methods were matched using the same ran- 0.001) were very important. Seventy
dom identification number. percent of female patients reported
Before initiation of this cross- Statistical analysis was performed that being treated as an individual
sectional study, approval was ob- with descriptive statistics for patient was very important, whereas only
tained from institutional review and guardian responses, as well as 58% of male patients did so (P =
board. Over a 1 month period, sur- univariate and agreement analysis to 0.02). Forty-seven percent of female
Jennifer J. Beck, MD, et al
Figure 1
Patient survey.
July 1, 2019, Vol 27, No 13 481
Figure 2
Guardian survey.
patients reported that surgeon as- 28%). Patients reported hearing good http://links.lww.com/JAAOS/A223).
sertiveness was very important, things about the surgeon (10%) as the Seventy-seven percent of patients aged
whereas only 34% of male patients third most common and most 10 to 14 years reported that their
did so (P = 0.03). Similarly, both important attribute, whereas guard- guardians chose the doctor compared
female and male patients reported ians reported treating their child as a with 56% of patients aged 15 to 18
that hearing good things about the unique individual (15%) as the third years who reported that their guard-
surgeon was very important (77% most common and most important ians chose the doctor (P , 0.001).
versus 65%); however, more female characteristic. Agreement was fair for When the guardians and patients
patients compared with male pa- the most important qualities of being decided the physician together, no
tients reported this attribute to be involved in decision making and statistical differences was found in
slightly more important (P = 0.045). understanding their sports and goals their surgeon preferences. However,
Both female and male adolescent (kappa: 0.236) and surgeon sex pref- when guardians made the decision
patients (85% versus 70%) reported erence (kappa: 0.258). alone, the patients were more likely to
surgeons involving them in decision In regard to surgeon sex, far fewer have different surgeon preferences
making and the surgeon under- adolescents reported that surgeon sex from their guardians, specifically on
standing their sports and goals as the was important than other character- the issue of sex (P , 0.05).
top important characteristics. istics, with male patients reporting a
Surgeon characteristics showed preference more frequently than
fewer notable differences between female patients (26% versus 12%; Discussion
male and female guardians. Female P = 0.005). Among the male patients,
guardians reported that hearing good 25% reported that they would Orthopaedic surgeons are seeing an
things about the surgeon was very prefer a male surgeon, less than 1% increasing number of adolescent pa-
important, more so than male guard- reported preference for a female tients for sports-related injuries.20
ians (83% versus 57%; P , 0.001). surgeon, and 74% had no prefer- With patient satisfaction at the
More female guardians compared ence. Among the female patients, 8% forefront of healthcare discussions,
with male guardians reported that reported preferring a female sur- desired characteristics in sports
understanding their child’s sports geon, 4% reported preference for a medicine surgeons in this growing
and goals was very important (86% male surgeon, and 88% had no population is unknown. In this
versus 67%; P , 0.003). Similar to preference (see Supplement 1, Sup- study, both adolescent patients and
the adolescent patients, both male plemental Digital Content 1, http:// their guardians reported shared
and female guardians reported that links.lww.com/JAAOS/A222). An decision making and understanding
surgeons involving them in decision even smaller percentage of guardians patients’ sports and goals as surgeon
making and the surgeon under- reported a surgeon sex preference, qualities that were more important
standing their sports and goals were with only 3% preferring male sur- to them than surgeon sex or repu-
the top important characteristics. geons, 3% preferring female sur- tation. Although shared decision
Comparing adolescent patient and geons, and 94% had no preference. making was highlighted as impor-
guardian responses (see Supplement A fair agreement existed between tant by both patients and their
2, Supplemental Digital Content 2, patients and their guardians on sur- guardians, only 33% of the time did
http://links.lww.com/JAAOS/A223), geon sex preference (see Supplement patients and their guardians decide
only fair to moderate agreement was 2, Supplemental Digital Content 2, on the physician together.
seen in all eight categories involving http://links.lww.com/JAAOS/A223). In contrast to the physician-driven
the importance of surgeon charac- Even though nearly all patients and and paternalistic medial decision
teristics between patients and guard- guardians reported that the adoles- making of past decades, healthcare
ians. Patient and guardian agreement cent patient should be involved in decision making has evolved into a
ranged from poor to fair with re- their medical decision making (pa- shared model. Bryant et al12 described
sponses to surgeon characteristics tients: 95%; guardians: 96%), phy- three treatment decision-making ap-
(intraclass correlation, 0.34 to 0.53). sician selection was determined by proaches: paternalistic, informed de-
Patients and guardians both reported the guardian 65% of the time and by cision making, and shared decision
being involved in the decision-making the patient 2% of the time, whereas making. It is important to understand
process as the most important sur- only 33% of the time did the adoles- the application of each decision-
geon characteristic (34% versus cent patient and guardian chose the making approach and its efficacy
31%), followed by understanding physician together (see Supplement 2, within each individual practice. Our
patients’ sports and goals (32% versus Supplemental Digital Content 2, study shows the importance of shared
July 1, 2019, Vol 27, No 13 483
decision making to both adolescent men. In 2014, the Association of teristics and may affect survey an-
athletes and their guardians in regard American Medical Colleges reported swers. Second, because of variations
to their sports medicine care. that orthopaedic surgery had the in how patients check-in to the clinic,
More recent orthopaedic surgery lowest proportion (16%) of full-time we were unable to calculate an exact
literature has described the effect of female faculty of any clinical or response percentage. We know that
shared decision making on orthopae- preclinical department.28 Male ado- for the 3 months around the time of
dic practices, reinforcing its impor- lescents may feel more comfortable the study, the volume of potentially
tance within our field.12-14 Similar to expressing their desires for a male eligible patients was 361, 321, and
this study, shared decision-making physician knowing the high likeli- 315. With 280 completed surveys,
approaches and benefits have been hood of having a male orthopaedic this suggests that the response per-
highlighted in upper extremity trau- surgeon. On the contrary, female centage would have been between
ma,15 carpal tunnel release,16,17 and adolescents must be more open- 70% and 90%, but we were unable to
ankle fractures18 literature. The in- minded to either sex as the likeli- verify this. Despite these limitations,
volvement of adolescents in their hood of having a female orthopaedic this study was a first step at attempt-
medical care is a challenge for their surgeon is very low. ing to assess patient and guardian
guardians and medical providers. As the only similar study performed preferences and how they may differ
During a time of psychological within orthopaedic surgery, Bridwell from each other. Last, patient and
growth and developing indepen- et al19 surveyed guardians and pa- guardian comprehension of survey
dence, adolescent decisions may tients before an upcoming scoliosis words and questions was not able to
change with growth, development, surgery. They had similar results to be assessed; therefore, accuracy of
and alterations in family or social our study that adolescent patient and responses cannot be guaranteed.
context.23 Pressures from coaches, guardian concerns and opinions
teammates, and peers may be un- regarding scoliosis surgery are dif-
derestimated in this population. ferent. Specifically, Bridwell noted Conclusions
Adolescents are more likely to focus that guardians and patients did not
In this study, both adolescent patients
on immediate, short-term treatment agree on reasons for having surgery or
and their guardians reported shared
effects and quality of life, whereas on their specific concerns or expec-
decision making and understanding
parents take a longer term view.24 tations about the surgery. Similarly,
patients’ sports and goals as surgeon
Careful explanations of details, our study only had fair agreement
qualities that were more important
while avoiding confusion, may between patients and guardians
to them than surgeon sex or repu-
enable adolescents to make informed regarding surgeon characteristics.
tation. Further study is warranted to
decisions without adding stress and Life experiences and alteration in
better understand the needs of young
confusion.25 Adolescents do appre- perspectives on present versus future
athletes because it relates to their
ciate being included in decisions effects may influence guardian and
healthcare treatment and outcomes.
and having their opinion valued,26 patient responses to these surveys.
and in this study, most adolescents
seeking sports medicine care wish to References
play active roles in their healthcare Limitations
decisions. References printed in bold type are
The current study showed that male This study had several limitations. those published within the past 5
adolescents had higher preference for First, the study was performed at one years.
male surgeons, whereas female ado- specific clinic and demographic fac-
1. Amir H, Beri A, Yechniely R, Levy YA,
lescents had less of a preference. This tors were not addressed as potential Shimonov M, Groutz A: Do urology
somewhat contradicts previously biases. The generalizability of this male patients prefer same-gender
cited studies. Proposed reasoning for study therefore cannot be determined urologist? Am J Mens Health 2018;12:
1379-1383.
this discrepancy could be because of from these data but will be addressed
sex-based role biases. Nationwide, in future studies by this research 2. Childs A, Friedman WH, Schwartz MP,
Johnson M, Royek AB: Female patients’
orthopaedic surgery continues to be a team. The aim of this study did not sex preferences in selection of
male-dominated specialty, with only include surgeon demographics gynecologists and surgeons. South Med J
2005;98:405-408.
5.3% of orthopaedic surgeons being beyond sex, nor attire, practice
women.27 Orthopaedic surgeons, model/location, skill set/training, or 3. Hall JA, Blanch-Hartigan D, Roter DL:
Patients’ satisfaction with male versus
and surgeons in general, are stereo- case volume. These may be more female physicians: A meta-analysis. Med
typically and actually predominantly important which questioned charac- Care 2011;49:611-617.
4. Hall JA, Roter DL: Do patients talk 13. Slover J, Shue J, Koenig K: Shared decision- provider gender and confidentiality in their
differently to male and female physicians? A making in orthopaedic surgery. Clin health care. J Adolesc Health 1999;25:
meta-analytic review. Patient Educ Couns Orthop Relat Res 2012;470:1046-1053. 131-142.
2002;48:217-224.
14. Smith MA: The role of shared decision 22. Woynarowska-Soldan M, Tabak I,
5. Kerssens JJ, Bensing JM, Andela MG: making in patient-centered care and Doroszewska A: The feelings of
Patient preference for genders of health orthopaedics. Orthop Nurs 2016;35: adolescents connected with medical visits
professionals. Soc Sci Med 1997;44: 144-149. and their perception of the physicians’
1531-1540. behavior. Dev Period Med 2014;18:
15. Hageman MG, Reddy R, Makarawung DJ,
331-342.
6. Lurie N, Slater J, McGovern P, Ekstrum J, Briet JP, van Dijk CN, Ring D: Do upper
Quam L, Margolis K: Preventative care extremity trauma patients have difference 23. Lipstein EA, Brinkman WB, Fiks AG, et al:
for women: Doe the sex of the physician preferences for shared decision-making An emerging field of research: Challenges in
matter? N Engl J Med 1993;329: than patients with nontraumatic pediatric decision-making. Med Decis
478-482. conditions? Clin Orthop Relat Res 2015; Making 2015;35:403-408.
473:3542-3548.
7. Nolen HA, Moore JX, Rodgers JB, Wang 24. Lipstein EA, Dodds CM, Lovells DJ,
HE, Walter LA: Patient preference for 16. Gong HS, Huh JK, Lee JH, Kim MB, Chung Denson LE, Britto MT: Making decisions
physician gender in the emergency MS, Baek GH: Patients’ preferred and about chronic disease treatment: A
department. Yale J Biol Med 2016;89: retrospectively perceived levels of comparison of parents and their
131-142. involvement during decision-making adolescent children. Health Expect 2014;
regarding carpal tunnel release. J Bone Joint 19:716-726.
8. Reid I: Patients’ preference for male or Surg Am 2011;93:1527-1533.
female breast surgeons: Questionnaire 25. Day E, Jones L, Langer R, Bluebond-
study. BMJ 1998;317:1051. 17. Nam KP, Gong HS, Bae KJ, Rhee SH, Lee Langer M: Current understanding of
HJ, Baek GH: The effect of patient decision-making in adolescents with cancer:
9. Weisman CS, Teitelbaum MA: Physician involvement in surgical decision making for
gender and the physician-patient A narrative systematic review. Palliat Med
carpal tunnel release on patient-reported 2016;30:920-934.
relationship: Recent evidence and relevant outcome. J Hand Surg Am 2014;39:
questions. Soc Sci Med 1985;20: 493-498. 26. Morgan R, Katzman D, Kaufman M,
1119-1127.
Goldberg E, Toulany A: Thanks for asking:
18. Hutchinson RH, Barrie JL: The effects of
10. Rehman SU, Nietert PJ, Cope D, Kilpatrick Adolescent attitudes and preferences
shared decision making in the conservative
AO: What to wear today? Effect of regarding the use of chaperones during
management of stable ankle fractures.
doctor’s attire on the trust and confidence physical examination. Pediatr Child Health
Injury 2015;46:1116-1118.
of patients. Am J Med 2005;118: 2016;21:191-195.
1279-1286. 19. Bridwell KH, Shufflebarger HL, Lenke LG,
27. AAOS Department of Research and
Lowe TG, Betz RR, Bassett GS: Parents’
11. Abghari MS, Takemoto R, Sadiq, Karia R, Scientific Affairs: American Academy of
and patients’ preference and concerns in
Phillips D, Egol KA: Patient perceptions Orthopedic Surgeons Orthopedic Practice
idiopathic adolescent scoliosis. Spine 2000;
and preferences when choosing an in the U.S. in 2016. Rosemont, IL,
25:2392-2399.
orthopaedic surgeon. Iowa Orthop J 2014; American Academy of Orthopedic
34:204-208. 20. Gottschalk AW, Andrish JT: Surgeons, 2017.
Epidemiology of sports injury in pediatric
12. Bryant D, Bednarski E, Gafni A: athletes. Sports Med Arthrosc Rev 2011; 28. Lautenberger DM, Dandar VM, Raezer
Incorporating patient preferences into 19; 2-6. CL: The State of Women in Academic
orthopaedic practice: Should the Medicine: The Pipeline and Pathways to
orthopaedic encounter change? Injury 21. Kapphahn CJ, Wilson KM, Klein JD: Leadership, Association of American
2006;37:328-334. Adolescent girls’ and boy’s preference for Medical Colleges, 2013-2014.
July 1, 2019, Vol 27, No 13 485
Research Article
Insurance Status Affects

Complication Rates After Total
Hip Arthroplasty
Abstract
David R. Veltre, MD Introduction: Previous studies have examined the relationship
David C. Sing, MD between total hip arthroplasty (THA) and insurance status in small
Paul H. Yi, MD cohorts. This study evaluates the effect of patient insurance status on
complications after primary elective THA using the Nationwide
Atsushi Endo, MD, MPH
Inpatient Sample.
Emily J. Curry, BA Methods: All patients undergoing primary elective THA from 1998 to
Eric L. Smith, MD 2011 were included. Patient demographics, comorbidities, and
Xinning Li, MD complications were collected and compared based on insurance type.
Multivariable logistic regression and a matched cohort analysis were
From the Department of Orthopaedic performed.
Surgery, Boston University School of
Results: About 515,037 patients (53.7% Medicare, 40.1% private
Medicine, Boston, MA (Dr. Veltre,
Dr. Sing, Dr. Smith, and Dr. Li), The insurance, 3.9% Medicaid/uninsured, and 2.2% other) were included,
Russell H. Morgan Department of who underwent elective THA. Privately insured patients had fewer
Radiology and Radiological Science,
Johns Hopkins University School of
medical complications (odds ratio, 0.80; P , 0.001), whereas patients
Medicine, Baltimore, MD (Dr. Yi), the with Medicaid or no insurance demonstrated no notable difference
Department of Epidemiology, Harvard (odds ratio, 1.03; P = 0.367) compared with Medicare patients.
T.H. Chan School of Public Health
(Dr. Endo), and the Department of
Similar trends were found for both surgical complications and
Epidemiology and Biostatistics, mortality, favoring lower complication rates for privately insured
Boston University School of Public patients. Furthermore, patients with private insurance tend to go to
Health (Ms. Curry), Boston, MA.
higher volume hospitals for total hip replacement surgery compared
Correspondence to Dr. Li: to those with Medicare insurance.
xinning.li@gmail.com
Discussion: Patients with government-sponsored insurance
Dr. Smith or an immediate family
(Medicare or Medicaid) or no insurance have higher risk of medical
member serves as a paid consultant to
Arthocare, Conformis, and DePuy; has complications, surgical complications, and mortality after primary
received research or institutional elective THA compared with privately insured patients. Insurance
support from Conformis and DePuy;
and serves as a board member, owner,
status should be considered an independent risk factor for stratifying
officer, or committee member of the patients before THA procedures.
American Orthopaedic Association.
immediate family member has received
M
anything of value from or has stock or ore than 300,000 total hip complications occur not infrequently
stock options held in a commercial
company or institution related directly or
arthroplasty (THA) surgeries after THA, inflicting considerable
indirectly to the subject of this article: are performed annually, making it physical, mental, and financial burden
Dr. Veltre, Dr. Sing, Dr. Yi, Dr. Endo, one of the most common surgeries on patients and their families and
Ms. Curry, and Dr. Li. performed in the United States.1 THA caretakers, as well as on the healthcare
J Am Acad Orthop Surg 2019;27: predictably results in tremendous system.5-7 Infection, in particular, is a
e606-e611 improvements in quality of life for devastating complication for both the
DOI: 10.5435/JAAOS-D-17-00635 patients suffering from osteoarthritis patient and the healthcare system after
and other hip pathologies, such as total joint arthroplasty (TJA).8
Academy of Orthopaedic Surgeons. osteonecrosis, rheumatoid, and post- Factors associated with complica-
traumatic arthritis.2-4 Unfortunately, tions after TJA have been well described
David R. Veltre, MD, et al
in the literature.9,10 In recent years, and admission types that were classi- tients who do not have any matching
increased attention has been paid fied as elective were included in the demographics with a corresponding
toward psychosocial and socioeco- analysis. Admissions that were classi- Medicare/private insurance patient.
nomic factors11 and their effect on fied as urgent or emergent and re- All the analyses were performed
outcomes and complications after visions were excluded from this study. using R 3.0.3 (R Foundation,
TJA, including treatment at a safety Patient demographics (sex, age, race, www.r-project.org).
net hospital,12 mental health prob- and admission type) and comorbid-
lems,13-15 and race.16 Insurance status ities using the Charlson comorbidity
has also previously been associated index were collected and analyzed. Results
with worse outcomes after TJA.17-20 Age was divided into four categories:
Previous studies assessing the rela- less than 40 years, 40 to 64 years, 65 Demographics
tionship between insurance status and to 79 years, and more than 80 years. Overall, 515,037 patients fulfilled cri-
TJA complications were all conducted All patient factors were separated teria for inclusion into the study. The
in small cohorts18 or primarily focused based on insurance type (Medicare, payer mix included 53.7% (276,695)
on comparisons between Medicaid private insurance, Medicaid/Uninsured, patients with Medicare, 40.1%
and Medicare populations.13 Our and “other”). The primary outcomes (206,692) patients with private insur-
study uses a large all-payer inpatient were medical and surgical complica- ance, 3.9% (20,153) patients with
healthcare database to evaluate the tions including mortality during Medicaid or no insurance, and 2.2%
effect of patient insurance status and the same hospitalization. Medical (11,497) patients from other programs
other associated patient factors on the complications included myocardial (ie, worker’s compensation, Civilian
medical complication, surgical com- infarction, pulmonary edema, venous Health and Medical Program of the
plication, and mortality rates after thromboembolism, cerebrovascular Uniformed Services, Civilian Health
primary elective THA. Bothhemiar- accident, acute kidney injury, pneu- and Medical Program of the Depart-
throplasties, revisions, and total hip monia, sepsis, and urinary tract ment of Veterans Affairs, Maternal
arthroplasties performed urgently for infection (UTI). Surgical complica- and Child Health Services Block Grant
trauma were excluded from this study. tions included septic shock, hema- [Title V of the Social Security Act], and
toma, accidental puncture, wound other governmental programs).
dehiscence, retained foreign body, Demographic information is re-
Methods postop infection, nonhealing surgical ported in Table 1 (see Supplemental
wound, and other surgical compli- Digital Content 1, http://links.lww.com/
Data were obtained from the cations. For statistical analysis, JAAOS/A250). Most of the patients
Healthcare Cost and Utilization Pro- Pearson’s chi-squared test was used were white, followed by black and
ject Nationwide Inpatient Sample to compare baseline characteristics Hispanic. Most of the patients with
(NIS) between 1998 and 2011. The on each insurance cohort. Multi- Medicaid and private insurance were
NIS is the largest national database variable logistic regression was used younger than 65 years. Patients with
of all-payer inpatient discharge infor- to determine the influence of insur- government-sponsored insurance ten-
mation, sampling approximately 20% ance type on complications. ded to be less healthy, as indicated
of all nonfederal US hospitals and in- A secondary analysis was performed by their higher comorbidity index.
cluding approximately 9 million hos- using a matched cohort comparing Stark contrast exists between the
pital admissions each year. Each NIS patients with Medicare with those with Medicaid/uninsured patients and the
entry includes International Classifi- private insurance. Each Medicare pa- private insured patients, where 46.2%
cation of Diseases, 9th Revision, tient was matched one-to-one with a of Medicaid/Uninsured patients versus
Clinical Modification diagnosis and private insurance patient on the basis only 28.7% of the private insured pa-
procedure codes of activity during the of age, sex, and other demographics tients went to low-volume hospitals
patient’s hospitalization at the time of and comorbidities using the coarsened (,149 cases per year) for their primary
discharge, as well as patient demo- exact matching algorithm. Matching hip replacement surgery (see Table 1,
graphics, hospital characteristics, and improves the covariate estimation in Supplemental Digital Content 1, http://
duration of stay. predicting outcomes through decreas- links.lww.com/JAAOS/A250).
Patients undergoing THA were ing the demographic imbalance be-
selected based on International Clas- tween the Medicare and the private
sification of Diseases, 9th Revision insurance groups. The matching algo- Overall Complications
procedural coding (total hip replace- rithm sorts patients based on matching The most common complication over-
ment, 81.51). Only primary surgeries demographics and then discards pa- all was cardiac events that occurred in
July 1, 2019, Vol 27, No 13 e607
Total Hip Arthroplasty Complications Insurance Type
Figure 1
Types of medical and surgical complications based on insurance types.
8.4% of Medicare patients, at almost pared with Medicare (odds ratio of having a medical complication.
three times the rate of patients on [OR], 0.8; 95% confidence interval The trend is similar for those with
private insurance (3.2%) (Figure 1). [CI], 0.78 to 0.82; P , 0.001) (see increasing comorbidity index, in which
UTIs and pneumonia were the next Table 2, Supplemental Digital Con- those with comorbidity index $3 are
two most frequent medical compli- tent 2, http://links.lww.com/JAAOS/ 20 times (OR, 20.44; 95% CI, 19.69 to
cations, both seen more commonly in A251). Additionally, patients with 21.23; P , 0.001) higher odds of
those with government-sponsored Medicaid or no insurance did not having a medical complication com-
insurance. UTI was seen in 2.6% of have a markedly higher risk of medical pared with those with a comorbidity
Medicaid/Uninsured patients and complications compared with patients index of zero.
3.5% of Medicare patients compared with Medicare (adjusted OR, 1.03;
with 1.7% in the private patient 95% CI, 0.97 to 1.09; P = 0.367).
population. Pneumonia was seen in Additionally, white patients had Surgical Complications
1.1% of Medicaid/Uninsured patients more medical complications than Similar to medical complications, in
and 1.2% of Medicare patients com- black, Hispanic, or Asian patients comparing the incidence of surgical
pared with 0.6% in patients with (adjusted OR, 0.89, 0.78, and 0.57, complications (see Table 2, Supple-
private insurance. The most compli- respectively, compared with white mental Digital Content 2, http://
cations seen overall, irrespective of the patients with P , 0.001). Increasing links.lww.com/JAAOS/A251) between
type of hip replacement surgery, was age and comorbidity index were insurance types, patients with private
found in the Medicare population, substantially associated with medical insurance had a statistically signifi-
followed by the Medicaid/Uninsured, complications. Compared with pa- cant lower risk of surgical compli-
other insurances, and privately tients younger than 40 years, pa- cations (OR, 0.92; 95% CI, 0.87 to
insured patient populations. tients aged 40 to 64 years had two 0.97; P = 0.002) compared with
times (OR, 2.23; 95% CI, 2.04 to patients with Medicare, whereas
2.44; P , 0.001), patients aged 65 to patients who have Medicaid or no
Medical Complications 79 years had nearly four times (OR, insurance had no notable difference
The multivariable logistic regression 3.58; 95% CI, 3.27 to 3.93; P , (OR, 1.04; 95% CI, 0.94 to 1.14;
analysis found that having private 0.001), and those older than 80 years P = 0.458) compared with patients
insurance was associated with less had six times (OR, 5.98; 95% CI, with Medicare insurance. Black,
overall medical complications com- 5.44 to 6.56; P , 0.001) higher odds Hispanic, and Asian patients had a
statistically significant increase in postoperative infection, postoperative complications or mortality risk com-
surgical complications compared anemia, and mortality. We found no pared with each other. In compari-
with white patients. Patients with notable difference in the rates of son, those with private insurance had
comorbidities had an increased risk peripheral vascular, gastrointesti- markedly fewer medical and surgical
of surgical complications, although nal, wound dehiscence, deep vein complications and decreased risk
the effect is not as dramatic as it is thrombosis, and pulmonary embo- of mortality. These findings are in
for medical complications. Finally, a lism complications between the two keeping with previous literature on
statistically significant increase is matched cohorts of Medicare and the joint arthroplasty which indicates
observed in surgical complications in private insurance patients after pri- decreased clinical outcomes and
younger patients (aged less than 40 mary elective THA. higher complications in patients with
years) and in male patients. government-sponsored insurance
(Medicaid), which may be associated
Mortality Discussion with their poor socioeconomic status
and access to care.17-19 Similar large
The overall mortality rate in our
Demand for primary hip arthro- database studies in the trauma litera-
study group was 0.1% (n = 693).
plasties is projected to rise dramati- ture also showed an increase in mor-
Compared with Medicare patients,
cally over the next decade and tality22-24 and poorer outcomes25,26 in
patients with Medicaid or no insur-
beyond.21 This rise in use makes it uninsured patients after trauma.
ance had no difference in mortality
even more imperative to identify and To add credence to our multivari-
(OR, 1.31; 95% CI, 0.81 to 2.1; P =
understand complications that oc- able regression analysis of all pa-
0.273) and those with private in-
cur after hip arthroplasty. To our tients, when accounting for patient
surance had a markedly lower mor-
knowledge, this is the largest and characteristics and potential con-
tality risk (OR, 0.56; 95% CI, 0.42
most comprehensive analysis of founding variables in a matched
to 0.75; P , 0.001). Female patients
outcomes and complications after cohort model, the rates of complica-
(OR, 0.65; 95% CI, 0.56 to 0.76;
THA as they relate to insurance tions were still markedly higher in
P , 0.001) had a lower risk of mor-
status. Our multivariable regression many categories in the Medicare
tality compared with male patients
analysis found that after primary group than in the private insurance
(see Table 2, Supplemental Digital
elective hip arthroplasty, postopera- group. All complications that had a
Content 2, http://links.lww.com/
tive in-hospital complication were statistically significant (P , 0.05)
JAAOS/A251). Patients who were
associated with the patient’s insur- difference between Medicare and
older and had more comorbidities
ance status and other patient factors. private insurance patients showed a
also had higher mortality risk.
A secondary analysis matching demo- higher relative risk for patients with
graphics and comorbidities between Medicare insurance. Medicare pa-
Matched Cohort Analysis Medicare and private insurance patients had a higher risk of compli-
The demographic information for tients confirmed these results, demon- cations because of central nervous
matched cohort between the Medi- strating that complications were more system (relative risk [RR], 1.25), car-
care and the private insurance pa- common in government-sponsored diovascular (RR, 1.18), respiratory
tients is reported in Table 3 (see insurance (Medicare) patients than (RR, 1.30), genitourinary (RR, 1.25),
Supplemental Digital Content 3, in those with private insurance after hematoma (RR, 1.07), postoperative
http://links.lww.com/JAAOS/A252), primary elective THA. infection (RR, 1.42), postoperative
which demonstrates similar baseline Multivariable regression analysis anemia (RR, 1.06), and mortality
characteristics and comorbidities from more than half of a million pa- (RR, 1.45). The most common com-
among the two study groups. Table 4 tients included in this study indicates plication, postoperative anemia,
(see Supplemental Digital Content 4, that patients with Medicare and occurred in 33.3% of Medicare pa-
http://links.lww.com/JAAOS/A253) those with Medicaid or no insurance tients compared with 31.5% of pri-
compares the proportions of com- were more likely to have medical vate insurance patients. All other
plications between the two insurance complications, most commonly car- medical complications occurred in less
groups when controlling for both diac complications, UTIs, and pneu- than 5% of the patients in either
demographic and risk factors. Patients monia, compared with private insurance population. This under-
with Medicare had a markedly higher insurance patients. However, pa- scores the importance of monitoring
risk of complications from central ner- tients with Medicare and Medicaid or all patients closely for postoperative
vous system, cardiovascular, respira- no insurance had no notable differ- anemia because it occurs in many
tory, genitourinary, hematoma/seroma, ence in both the medical and surgical patients recovering from THA.
July 1, 2019, Vol 27, No 13 e609
Total Hip Arthroplasty Complications Insurance Type
A similar study using the NIS standard arthritis, resulting in a of outcomes between insurance types
database performed by Browne greater risk of surgical complication should be performed to optimize
et al13 looked at complications in a due to complexity of hip pathology. patient outcomes after THA.
matched group of Medicaid and However, given the nature of the
non-Medicaid patients who under- study and limitations of results drawn
went primary hip and knee arthro- from a national database, we are References
plasty. They found a higher rate of unable to imply causality with this
wound dehiscence, infection, and association. The data are from a NIS References printed in bold type are
anemia in the Medicaid patients and a and only have information on a single those published within the past 5
hospital discharge, so no data are years.
higher rate of cardiac complications
in the non-Medicaid patients.20 For available on follow-up rates, read- 1. Products—Data Briefs—Number 186—
our analysis, we performed both a missions, or delayed complications that February 2015. CDC/NCHS, National
Hospital Discharge Survey, 2000–2010.
multivariable logistic regression to occur after the patient is discharged. 2016. http://www.cdc.gov/nchs/data/
analyze the overall effect of insurance Additionally, another interesting find- databriefs/db186.htm. Accessed April 19,
2016.
on outcomes and a matched cohort of ing was inhomogeneity between the
patients with Medicare versus private different groups in terms of where most 2. Ethgen O, Bruyere O, Richy F, Dardennes
C, Reginster JY: Health-related quality of
insurance. With the matching scheme, patients had their surgery performed life in total hip and total knee arthroplasty:
our analysis better used the whole because it relates to their insurance A qualitative and systematic review of the
national cohort of patients without status. Patients with private insurance literature. J Bone Joint Surg Am 2004;86-A:
963-974.
missing any “unmatched” patient were more likely to get their surgeries
data. Additionally, we excluded all done at high-volume centers (.293 3. Laupacis A, Bourne R, Rorabeck C, et al:
The effect of elective total hip replacement
hemiarthroplasties, revision total hip cases per year) and in an urban setting, on health-related quality of life. J Bone
arthroplasties, and any total hip whereas patients with Medicare or Joint Surg Am 1993;75:1619-1626.
replacement done in the urgent setting Medicaid/Uninsured were more likely 4. Daigle ME, Weinstein AM, Katz JN, Losina
to strengthen our findings while mini- to have their surgery performed at E: The cost-effectiveness of total joint
arthroplasty: A systematic review of
mizing bias. We believe that our results hospitals with annual volume of less published literature. Best Pract Res Clin
reflect the real-world population well, than 149 cases and in a rural setting. Rheumatol 2012;26:649-658.
especially in the overall complication Previous studies have demonstrated 5. Lavernia C, Lee DJ, Hernandez VH: The
and mortality rates after primary elec- that hospital volume alone may be increasing financial burden of knee revision
surgery in the United States. Clin Orthop
tive THA procedure. associated with increased mortality Relat Res 2006;446:221-226.
In our study, we were able to and complications rates.27,28 And
6. Crowe JF, Sculco TP, Kahn B: Revision
analyze a large number of patient re- finally, although statistical signifi- total hip arthroplasty: Hospital cost and
cords to find statistical trends that cance can be demonstrated in a large reimbursement analysis. Clin Orthop Relat
may not otherwise be discernible in database study such as this, the clin- Res 2003:175-182.
smaller, retrospective, single or mul- ical relevance of small but statistically 7. Kurtz SM, Ong KL, Schmier J, et al: Future
tisite studies. The database is a significant differences is an area of clinical and economic impact of revision
total hip and knee arthroplasty. J Bone
nationwide sampling of all elective debate. Joint Surg Am 2007;89(suppl 3):144-151.
primary total hip arthroplasties, so it Our data show that, in general,
8. Berend KR, Lombardi AV Jr, Morris MJ,
can be seen as broadly applicable. patients with government-sponsored Bergeson AG, Adams JB, Sneller MA: Two-
However, because of the nature of a (Medicare) or no insurance were stage treatment of hip periprosthetic joint
infection is associated with a high rate of
retrospective study, we are limited in sicker and older and had more medical infection control but high mortality. Clin
our ability to analyze causality. For and surgical complications including Orthop Relat Res 2013;471:510-518.
example, our study found that the mortality. Older patients or those with 9. Bozic KJ, Lau E, Ong K, et al: Risk factors
surgical complication rates are mark- more comorbidities were more likely to for early revision after primary TKA in
edly higher among patients younger have medical and surgical complica- Medicare patients. Clin Orthop Relat Res
2014;472:232-237.
than 40 years than in patients aged tions including death. Using a matched
between 40 and 80 years. One possible cohort to minimize confounding vari- 10. Jain NB, Guller U, Pietrobon R, Bond TK,
Higgins LD: Comorbidities increase
explanation for this finding is that ables between the groups of Medicare complication rates in patients having
younger patients (aged less than 40 and private insurance patients, the risk arthroplasty. Clin Orthop Relat Res 2005:
232-238.
years) may have more complex hip of postoperative medical and surgical
pathology if indicated for a THA (eg, complications including mortality was 11. Barrack RL, Ruh EL, Chen J, et al: Impact
of socioeconomic factors on outcome of
hip dysplasia, posttraumatic arthritic, also higher in Medicare patients. Fur- total knee arthroplasty. Clin Orthop Relat
osteonecrosis) than older patients with ther research investigating the disparity Res 2014;472:86-97.
12. Jergesen HE, Yi PH: Early complications demographics, and postoperative outcomes Insurance coverage is associated with
in hip and knee arthroplasties in a based on insurance payer type. J mortality after gunshot trauma. J Am Coll
Safety Net Hospital vs a University Arthroplasty 2012;27:1761-1765.e1. Surg 2010;210:280-285.
Center. J Arthroplasty 2016;31:
754-758. 18. Rosenthal BD, Hulst JB, Moric M, Levine 24. Tepas JJ III, Pracht EE, Orban BL, Flint
BR, Sporer SM: The effect of payer type on LM: Insurance status, not race, is a
13. Browne JA, Sandberg BF, D’Apuzzo MR, clinical outcomes in total knee determinant of outcomes from vehicular
Novicoff WM: Depression is associated arthroplasty. J Arthroplasty 2014;29: injury. J Am Coll Surg 2011;212:
with early postoperative outcomes 295-298. 722-727.
following total joint arthroplasty: A
nationwide database study. J Arthroplasty 19. Hinman A, Bozic KJ: Impact of payer type 25. Havens JM, Olufajo OA, Cooper ZR,
2014;29:481-483. on resource utilization, outcomes and Haider AH, Shah AA, Salim A:
access to care in total hip arthroplasty. J Defining rates and risk factors for
14. Ellis HB, Howard KJ, Khaleel MA, Arthroplasty 2008;23:9-14. readmissions following emergency
Bucholz R: Effect of psychopathology on general surgery. JAMA Surg 2016;151:
patient-perceived outcomes of total knee 20. Browne JA, Novicoff WM, D’Apuzzo MR: 330-336.
arthroplasty within an indigent Medicaid payer status is associated with
population. J Bone Joint Surg Am 2012; in-hospital morbidity and resource 26. Haider AH, Chang DC, Efron DT, Haut
94:e84. utilization following primary total joint ER, Crandall M, Cornwell EE III: Race
arthroplasty. J Bone Joint Surg Am 2014; and insurance status as risk factors for
15. Lavernia CJ, Alcerro JC, Brooks LG, Rossi 96:e180. trauma mortality. Arch Surg 2008;143:
MD: Mental health and outcomes in 945-949.
primary total joint arthroplasty. J 21. Kurtz S, Ong K, Lau E, Mowat F, Halpern
Arthroplasty 2012;27:1276-1282. M: Projections of primary and revision hip 27. Katz JN, Losina E, Barrett J, et al:
and knee arthroplasty in the United States Association between hospital and surgeon
16. Lavernia CJ, Alcerro JC, Contreras JS, from 2005 to 2030. J Bone Joint Surg Am procedure volume and outcomes of total
Rossi MD: Ethnic and racial factors 2007;89:780-785. hip replacement in the United States
influencing well-being, perceived pain, and medicare population. J Bone Joint Surg Am
physical function after primary total joint 22. Greene WR, Oyetunji TA, Bowers U, et al: 2001;83-A:1622-1629.
arthroplasty. Clin Orthop Relat Res 2011; Insurance status is a potent predictor of
469:1838-1845. outcomes in both blunt and penetrating 28. Doro C, Dimick J, Wainess R, Upchurch G,
trauma. Am J Surg 2010;199:554-557. Urquhart A: Hospital volume and inpatient
17. Martin CT, Callaghan JJ, Liu SS, Gao Y, mortality outcomes of total hip
Warth LC, Johnston RC: Disparity in total 23. Dozier KC, Miranda MA Jr, Kwan RO, arthroplasty in the United States. J
joint arthroplasty patient comorbidities, Cureton EL, Sadjadi J, Victorino GP: Arthroplasty 2006;21:10-16.
July 1, 2019, Vol 27, No 13 e611
Research Article
Interventions for Distal Radius

Fractures: A Network Meta-analysis
of Randomized Trials
Abstract
Christopher Vannabouathong, Introduction: The distal radius is the second most commonly
MSc fractured bone in the elderly population. Several approaches for
Nasir Hussain, MD management exist including both surgical and nonsurgical
Ernesto Guerra-Farfan, MD treatments. Evidence for alternative approaches remains
inconclusive and often conflicting.
Mohit Bhandari, MD, PhD,
FRCSC
Methods: Electronic database searches were conducted to identify
randomized trials that (1) enrolled adults with a distal radius fracture;
(2) compared external fixation, intramedullary nailing, K-wires, plaster
casting (PC), or plate fixation (PF) to one of the other treatments listed;
and (3) reported on a functional outcome that categorized patients as
“excellent” and “good” or reported fracture healing complication (loss
From the OrthoEvidence Inc, of reduction, malunion, delayed union, nonunion, and refracture).
Burlington, Ontario, Canada Data were synthesized using a Bayesian network meta-analysis.
(Mr. Vannabouathong), the
Department of Anesthesiology, Ohio
Odds ratios (ORs) with 95% credible intervals (CrIs) using a random-
State University College of Medicine, effects model were calculated.
Columbus, OH (Dr. Hussain), the Results: Thirty-eight trials were included. Total sample sizes ranged
Department of Traumatology,
Orthopaedic Surgery and Emergency,
from 30 to 461 patients. Although no differences were noted between
Hospital Vall d’Hebron, Barcelona, treatments at 3 months, PF was ranked the highest for improving
Spain (Dr. Guerra), and the functional outcomes at 6 and 12 months, and it demonstrated a
Department of Surgery, McMaster
University, Hamilton, Ontario, Canada
statistically significant difference compared with PC at 12 months
(Dr. Bhandari). (OR = 4.27; 95% CrI, 1.07 to 15.12). For reduction in fracture healing
Correspondence to Vannabouathong: complications, PF was ranked the highest again, showing
chris.vannabouathong@ significantly more favorable results relative to the four other
myorthoevidence.com interventions: OR = 0.25 (95% CrI, 0.07 to 0.86) versus external
Dr. Bhandari or an immediate family fixation, OR = 0.09 (95% CrI, 0.02 to 0.36) versus K-wire, OR = 0.01
member serves as a paid consultant
to Acumed, LCC, DJ Orthopaedics,
(95% CrI, 0.00 to 0.03) versus PC, and OR = 0.00 (95% CrI, 0.00 to
Ferring Pharmaceuticals, Sanofi- 0.35) versus intramedullary nailing.
Aventis, and Stryker. None of the Conclusions: A network meta-analysis of randomized trials revealed
following authors or any immediate
family member has received anything
that open reduction and internal fixation with a plate offers the best
of value from or has stock or stock results for adult patients with a distal radius fracture, in terms of early
options held in a commercial company and sustained functional recovery and a reduction in fracture healing
or institution related directly or
indirectly to the subject of this article:
complications. Determining whether one approach to PF or plate
Mr. Vannabouathong, Dr. Hussain, design is superior requires further study.
and Dr. Guerra.
J Am Acad Orthop Surg 2019;27:

e596-e605
DOI: 10.5435/JAAOS-D-18-00424 T he distal radius is one of the most
commonly fractured bones, with
approximately 18% of all adult or-
estimates in the United States alone
suggest an incidence of approxi-
mately 640,000 cases, and rising, per
Academy of Orthopaedic Surgeons. thopaedic fractures occurring in this year.1 With the high costs associated
region.1 Furthermore, annualized with the surgical management of
Christopher Vannabouathong, MSc, et al
distal radius fractures (DRFs), the Inclusion Criteria and Study the number of patients categorized as
principal questions for patients un- Selection “excellent” or “good” on a func-
dergoing treatment surround their tional scale at 3, 6, and $12 months.
We included RCTs that (1) followed
long-term benefit on function, mor- For fracture healing complications,
an adult or elderly population with
tality, and the risk of postoperative we compared the number of patients
DRFs; (2) compared EF, IMN, KW,
complications. Although the most who had an event over the entire
PC, or PF to one of the other treat-
recent guidelines developed by the length of the study.
ments listed; and (3) reported on a
American Academy of Orthopaedic We calculated odds ratios (ORs)
functional outcome that categorized
Surgeons (AAOS) are largely incon- with 95% credible intervals (CrIs)
patients as “excellent” and “good,”
clusive for the various surgical and using a random-effects model with
or reported fracture healing compli-
nonsurgical treatments for these vague priors.5 We generated the
cations. For this review, fracture
fractures,2 several options exist. corresponding network diagrams
healing complications included loss
The primary means of management and forest plots displaying all pair-
of reduction, malunion, delayed
of stable DRFs includes closed reduc- wise comparisons. Surface under the
union, nonunion, and refracture. We
tion under anesthesia with subsequent cumulative ranking curve (SUCRA)
excluded full-text publications that
stabilization with a plaster cast (plas- values were also presented, which
did not provide sufficient informa-
ter casting [PC]).3 Four different provides a numeric representation of
tion on outcomes or, for non-English
surgical interventions exist for pa- the overall ranking of a given treat-
publications, the data of interest
tients undergoing DRF repair: (1) ment relative to all other treatments
were not available in the English
closed reduction and percutaneous included in the network (ie, the
language abstract. Screening was
pinning with K-wires (KWs), (2) higher the value, the greater the
performed in duplicate, and any
external fixation (EF), (3) open re- likelihood that the treatment is top
discrepancies were resolved via a
duction and internal fixation (ORIF) ranked).5
third party.
with volar or dorsal locking plate
fixation (PF), and (4) intramedullary
nailing (IMN). However, clinical Data Extraction and Results
consensus on the topic is still unclear.3 Outcomes
A single, large-scale, randomized We extracted data related to the study Search Results
controlled trial (RCT) that com- design, participant demographics, The electronic literature search
pares all surgical and nonsurgical fracture characteristics, treatment yielded 9,666 citations (Figure 1). We
modalities for DRFs would be ideal, details, and measures on function and reviewed 364 full texts for final eli-
but difficult because of staggering fracture healing complications. gibility and 38 were included.6-43
costs. Thus, the objective of this
study was to perform a network Quality Assessment
meta-analysis (NMA) of random- Description of Included
The methodological quality of Studies
ized trials comparing the different
included trials was graded according
treatment modalities for DRF, in The included trials were published
to the Cochrane Risk of Bias (ROB)
terms of function and fracture heal- between 1988 and 2017 (see Appen-
tool.4
ing complications in the adult and dix B, Supplemental Digital Content
elderly populations. 2, http://links.lww.com/JAAOS/A219).
Data Analysis Studies were conducted across a wide
Data were synthesized using a range of geographical locations. One
Methods Bayesian NMA using the methods study clearly reported receiving
provided by Brown et al5 and the industry funding. Study follow-up
Search Strategy Canadian Agency for Drugs and ranged from 3 months to 7 years;
We conducted literature searches (see Technologies in Heath. The analysis 30 (78.9%) had $12 months of
Appendix A-Search strategy, Sup- was completed using the NetMetaXL follow-up. The Gartland-Werley
plemental Digital Content 1, http:// Microsoft Excel–based tool and scale and loss of fracture reduction
links.lww.com/JAAOS/A218) in the WinBUGS software.5 The WinBUGS were the most commonly reported
MEDLINE, Embase, Cochrane Li- code is based on the National functional measure and fracture
brary, CINAHL, AMED, PEDro, Web Institute for Health and Care Ex- healing complication, respectively.
of Science, WHO, clinicaltrials.gov, cellence Decision Support Unit The ROB assessment for all
and SIGLE databases. Series.5 For function, we compared included trials is presented in
July 1, 2019, Vol 27, No 13 e597
Interventions for Distal Radius Fractures
Figure 1 Finally, 32 studies were classified as

being low risk for selective reporting
bias because of sufficient reporting on
all outcomes that were stated in the
methods, or being preregistered and
having protocols available for review.
Patient demographics are summa-
rized in Appendix D (see Supple-
mental Digital Content 4, http://
links.lww.com/JAAOS/A221). Total
sample sizes ranged from 30 to 461
patients. The highest reported mean
age was 77.4 years, and the lowest
was 34.5 years. Most trials (25) had
less than 50% men in their sample.
Flow diagram of included studies. The most commonly used fracture
classification system was the AO
system, with most studies enrolling
Figure 2 patients with a type A or C DRF; the
Frykman classification system was
also used.
The corresponding network dia-
gram of overall treatment compar-
isons is available in Figure 2.
Network Meta-analysis of
Randomized Controlled
Trials
Function (Excellent or Good)
The NMA on functional outcomes at
3 months included six studies, total-
ing 277 patients. No studies evaluat-
ing IMN fixation were available for
this time point; therefore, this inter-
vention was not included in this
analysis. The results of the random-
effects model showed no statistically
significant differences in the ORs
between any of the pairwise com-
parisons (Figure 3). The correspond-
Overall network diagram. EF = external fixation, IMN = intramedullary nailing, ing SUCRA values were 80.1% for
KW = K-wires, PC = plaster casting, PF = plate fixation
KW, 63.3% for PF, 45.3% for EF,
and 11.3% for PC.
Appendix C (see Supplemental Digi- to treatment allocation. In contrast, Function at 6 months included six
tal Content 3, http://links.lww.com/ eight studies reported that outcome studies, totaling 325 patients. No
JAAOS/A220). Seventeen trials ade- assessors were either not blinded or studies evaluating IMN fixation were
quately described randomization that they were physicians directly available for this time point; there-
and used methods that were con- involved in the procedure. Fifteen fore, this intervention was not
sidered to have low ROB. Twenty studies had either a .80% follow-up included in this analysis. The results
trials had a low ROB for allocation rate, adequately described reasons of the random-effects model showed
concealment. Ten studies reported for missing data, or had balanced no statistically significant differences
that outcome assessors were blinded missing data between the groups. in the ORs between any of the
Figure 3
Forest plot of network meta-analysis showing an “excellent” or “good” functional outcome at 3 months. Results from the
random-effects (vague prior) model. CrI = credible interval, OR = odds ratio
pairwise comparisons (Figure 4). The random-effects model showed that

corresponding SUCRA values were patients managed with PF experi-
Discussion
87.3% for PF, 59.6% for KW, enced significantly lower odds of a
42.3% for EF, and 10.8% for PC. fracture healing complication com- Main Findings
The analysis on functional out- pared with EF (OR = 0.25; 95% CrI, A NMA of randomized trials re-
comes at $12 months included 17 0.07 to 0.86), KW (OR = 0.09; 95% vealed that ORIF with plates may
studies, totaling 1,123 patients. The CrI, 0.02 to 0.36), PC (OR = 0.01; offer the best results for patients
results of the random-effects model 95% CrI, 0.00 to 0.03), and IMN with a DRF, in terms of early and
demonstrated one significant out- (OR = 0.00; 95% CrI, 0.00 to 0.35) sustained posttreatment function (an
come between PC and PF (OR = (Figure 6). Both KW and EF were excellent or good result on a func-
4.27; 95% CrI, 1.07 to 15.12) in significantly more favorable than tional scale) and a reduction in frac-
favor of PF (Figure 5). All other PC for reducing the odds of a fracture healing complications. Also, PF
pairwise comparisons were not sig- ture healing complication, with was compared with EF, intra-
nificant. The corresponding SUCRA ORs of 0.07 (95% CrI, 0.01 to medullary nail fixation, K-wires, and
values were 83.0% for PF, 60.6% 0.34) and 0.03 (95% CrI, 0.01 to PC; the only instance where PF was
for IMN, 54.8% for KW, 45.1% for 0.11) versus PC, respectively. All not ranked the highest (according to
EF, and 6.5% for PC. other pairwise comparisons were the SUCRA analysis) was function at
not statistically significant. The 3 months. However, it was ranked
Fracture Healing Complications corresponding SUCRA values were the second highest after KW fixation.
The NMA on fracture healing com- 99.3% for PF, 72.4% for EF, The most recent AAOS clinical prac-
plications included 25 studies and 50.9% for KW, 21.9% for PC, and tice guideline on the treatment of
2,253 patients. The results of the 5.6% for IMN. DRFs was inconclusive on the use of
July 1, 2019, Vol 27, No 13 e599
Figure 4
casting and surgical interventions, comparing the different devices (Table spective cohort comparing the two
including locking plates in those 1). Biomechanical studies that have found no significant differences in
older than 55 years; however, ac- conducted cyclic load testing across a function scores and complica-
cording to the AAOS appropriate use number of plate designs demonstrated tions.45,48 A volar (or palmar) plate
criteria for DRFs, the appropriate- no difference between the plates.47,49 has a wider applicability and is more
ness ratings for both volar and dorsal A retrospective cohort found better commonly used; but a dorsal plate
plating range from “may be appro- DASH (Disabilities of the Arm, may be indicated for fractures classi-
priate” to “appropriate” across all Shoulder and Hand) scores and lower fied as dorsal shear fractures, dorsal
clinical scenarios.2,44 Patient demo- revision surgery rates in patients die-punch fractures, fractures where
graphics (age, sex distribution, and treated with low-profile dorsal plates reduction from the volar approach
fracture grade) and outcome re- versus the Synthes Pi plate.50 Obser- cannot be achieved, fractures with an
porting were quite variable across vational cohorts and one RCT showed associated scapholunate ligament
the included studies. Also, the RCT no significant difference in either injury, or comminuted fractures with a
evidence on IMN fixation is limited, radiological or clinical outcomes in bone defect that requires grafting.45,48
which is demonstrated by the larger comparisons of different volar plate Most of the studies included in this
uncertainty (CrIs) surrounding pair- systems.46,51,52 One RCT demon- review used a volar plate device (one
wise comparisons that included this strated that a volar plate may result in study used a dorsal approach and
intervention. more favorable functional outcomes three used a combination of both).
Various plates exist, and there is and lower complication rates relative The results of this NMA are con-
limited high-quality evidence directly to a dorsal approach, whereas a pro- sistent with those of previous meta-
Figure 5
analyses on the topic. Chaudhry nience to patients.53 The RCT evi- trials comparing PF with IMN is small,
et al53 also determined that volar dence on PF versus PC is more recent, and additional studies are needed to
locking plates provided better func- the earliest trial being published in make more definitive conclusions on
tional outcomes and reduced post- 2011.9 The results at longer-term their comparative efficacy and safety.
operative complications compared visits (12 months) are inconsistent, Although there appears to be some
with KW fixation. In the most recent but there is a clear trend toward benefit with PF, surgical intervention is
meta-analysis evaluating PF versus earlier (6 to 12 weeks) functional associated with its own set of risks,
EF, Li-hai et al54 found that PF results recovery with PF compared with PC; most notably related to wound heal-
in earlier return to function. There are also, PC is associated with increased ing;55 therefore, it is important to
also concerns with pin site infections risks of loss of reduction and mal- be mindful of this when treating a
with the use of EF devices, and its union that could require subsequent patient who may be at greater risk of
bulky nature may be an inconve- intervention.9,10,36 The number of developing a wound healing problem.
July 1, 2019, Vol 27, No 13 e601
Figure 6
Forest plot of network meta-analysis showing fracture healing complications. Results from the random-effects (vague prior)
model. CrI = credible interval, OR = odds ratio
Strengths and Limitations A limitation of our analysis in- they are to each other. The most
The strengths of our study were a cluded variation in patient demo- commonly used functional scale was
comprehensive and systematic search graphics; there was a wide range ages the Gartland-Werley (across 12 tri-
strategy, independent and duplicate included across trials, and different als), but the Green-O’Brien (4 trials),
screening for eligible studies, and use age groups may respond differently Sarmiento (3 trials), Lidstrom (2
of a NMA to compare numerous in- to these treatments. Additional re- trials), Mayo score, Horne, Cooney,
terventions for DRFs all at once, with search is also required to determine if and Jakim et al scales were also used
both direct and indirect high-quality therapy should be dependent on to assess function. Although all of
RCT evidence. The outcomes (func- fracture type or severity. There was these instruments are wrist-specific
tion and complications) are also con- also various functional scales used, measurements, there may be some
sidered patient important. and it is unclear how comparable variability with regard to the specific
Table 1
Prior Studies Comparing Different Plate Fixation Devices
First Author, Study
yr Design Device(s) Summary of Results/Conclusions
Abe et al45 PCS Aptus Radius volar, Medartis (volar plate) Except for wrist flexion, there were no other differences
SmartLock, Stryker; 2.4-mm LCPDRP, Synthes; in the clinical results between groups for both
Aptus Radius dorsa, Medartis (dorsal plate) subjective and objective parameters. There were no
significant differences in the complication rates
between the volar and dorsal plated groups. The
treatment of displaced intra-articular DRFs with a
dorsal versus a volar plate system demonstrated
similar clinical results. Postoperative complications
were not readily observed in the patients treated with a
dorsal plate. Certain fracture patterns are more
appropriately stabilized using a dorsal plate fixation.
Cho et al46 RC AO LCPDRP, Synthes Radiologic parameters were significantly improved after
Acu-Loc, Acumed surgery in both the groups. The mean Mayo scores
were similar. There were no significant differences
between the two groups with respect to both
radiological and clinical outcomes at the final follow-up.
The difference of implant design did not influence the
overall final outcomes.
Dahl et al47 BS Acu-Loc, Acumed All plates survived cyclic load testing. All plates showed
DVR, Hand Innovations increasing stiffness at higher loads. No failures
Extra-articular plate, Synthes occurred. All plate constructs met the anticipated
LoconVLS, Wright Medical demands of the loads transmitted across the wrist. It
Periarticular DR plate, Zimmer seems clear that fracture configuration, screw
SCS volar plate, Small Bone Innovations placement, cost, and surgeon familiarity with the
Volar distal plate, Synthes instrumentation should take priority in selecting a
SmartLock, Stryker plating system for DRF treatment.
Jakubietz RCT Aptusradius volar, Medartis (volar plate) The volar plate group demonstrated significantly better
et al48 AO Pi plate, Synthes (dorsal plate) range of motion and grip strength over the course of
12 months, and earlier functional recovery. Radiologic
results showed a significantly increased palmar tilt and
carpal sag in dorsal plates. Complications were also
more frequent in the dorsal plate group. This study has
shown the key advantage of the volar plate to be faster
recovery time, which will not only reduce the cost after
this injury but also may restore individual
independence especially in the elderly patient
population and possibly prevent nursing home
placement.
Kameiet al49 BS Acu-Loc, Acumed None of the plates failed during the cyclic loading. There
AO LCPDRP, Synthes were no significant differences among the five groups
DRV, Mizuho Ikakogyo for the failure load. All of the five plates provided
SmartLock, Stryker sufficient stability to permit 3,000 repeated motions of
Stellar, Japan Universal Technologies the digits after surgery for AO type C3 DRFs.
Rozental RC AO Pi plate, Synthes There was a significant difference in both the DASH
et al50 Acu-Loc, Acumed; DePuy plate, DePuy; T-plate, score and revision surgeries because of complications
Synthes (low-profile plates) in favour of low-profile plates.
Tanaka et al51 RCT Acu-Loc, Acumed Both groups demonstrated overall satisfactory function
VariAx, Stryker at 6 months, with no significant difference found
between the groups. Minimal loss of reduction was
demonstrated in both the groups.
von PCS 2.4 mm LCPDRP, Synthes The risk of experiencing a complication after ORIF was
Recumet al52 3.5 mm LCPDRP, Synthes not significant between the plates. Wrist function was
also similar between cohorts based on the range of
movement and grip strength measurements relative to
the contralateral healthy side. DASH, SF-36, and pain
scores were not significantly different between the
groups throughout the 2-yr period. No patient from
either treatment group had a step-off . 2 mm.
Differences in plate design did not influence the
overall final outcome of fracture fixation.
BS = biomechanical study, DASH = Disabilities of the Arm, Shoulder and Hand, DRF = distal radius fracture, LCPDRP = locking compression plate
distal radius plate, PCS = prospective cohort study, RC = retrospective cohort, RCT = randomized controlled trial
July 1, 2019, Vol 27, No 13 e603
domains evaluated within each scale distal radius: A systematic review. Bone fixation employing multiplanar K-wires
Joint J 2015;97-B:1370-1376. versus volar locked plating for dorsally
and their individual scoring systems. displaced fractures of the distal radius. Arch
Another consideration is that we 4. Cochrane Handbook for Systematic Orthop Trauma Surg 2013;133:595-602.
Reviews of Interventions Version 5.1.0.
grouped loss of reduction, malunion, Chichester, UK, The Cochrane 16. Grewal R, MacDermid JC, King GJ, Faber
delayed union, nonunion, and re- Collaboration; 2011. KJ: Open reduction internal fixation versus
percutaneous pinning with external fixation
fracture all as one composite outcome 5. Brown S, Hutton B, Clifford T, et al: A of distal radius fractures: A prospective,
(fracture healing complications). If an Microsoft-Excel-based tool for running and randomized clinical trial. J Hand Surg
critically appraising network meta- 2011;36:1899-1906.
investigator is only interested in analyses: An overview and application of
knowing the rate of just one of these NetMetaXL. Syst Rev 2014;3:110. 17. Gupta R, Raheja A, Modi U: Colles’
specific events, grouping all of them fracture: Management by percutaneous
6. Abbaszadegan H, Jonsson U: External crossed-pin fixation versus plaster of Paris
as one outcome makes this difficult to fixation or plaster cast for severely cast immobilization. Orthopedics 1999;22:
ascertain. Another potential limita- displaced Colles’ fractures? Prospective 1- 680-682.
year study of 46 patients. Acta Orthop
tion is that our analysis could not Scand 1990;61:528-530. 18. Hegeman JH, Oskam J, van der Palen J,
account for potential deviation of ten Duis HJ, Vierhout PAM: Primary
7. Abramo A, Kopylov P, Geijer M, Tagil M: external fixation versus plaster
patients from their assigned group. Open reduction and internal fixation immobilization of the intra-articular
For example, a patient may be as- compared to closed reduction and external unstable distal radial fracture in the elderly.
fixation in distal radial fractures: A Aktuelle Traumatol 2004;34:64-70.
signed to KW fixation before surgery randomized study of 50 patients. Acta
but then the surgeon decides to per- Orthop 2009;80:478-485. 19. Hollevoet N, Vanhoutie T, Vanhove W,
form another surgery instead because Verdonk R: Percutaneous K-wire fixation
8. Agrawal VRK: Distal radius fractures: A versus palmar plating with locking screws
of the presence of certain fracture comparative study between conservative for Colles’ fractures. Acta Orthop Belg
management and external fixation. JEMDS
characteristics. Our analysis was 2017;6:1093-1098.
2011;77:180-187.
dependent on how the investigators 20. Howard PW, Stewart HD, Hind RE, Burke
9. Arora R, Lutz M, Deml C, Krappinger D, FD: External fixation or plaster for severely
of the included studies chose to ana- Haug L, Gabl M: A prospective randomized displaced comminuted Colles’ fractures? A
lyze their study samples. Also, when trial comparing nonoperative treatment prospective study of anatomical and
data are sparse, Bayesian methodol- with volar locking plate fixation for functional results. J Bone Joint Surg Br
displaced and unstable distal radial 1989;71:68-73.
ogy with vague priors may have a fractures in patients sixty-five years of age
greater influence on the analysis.5 and older. J Bone Joint Surg 2011;93: 21. Hutchinson DT, Strenz GO, Cautilli RA:
2146-2153. Pins and plaster vs external fixation in the
treatment of unstable distal radial fractures:
10. Bartl C, Stengel D, Bruckner T, Gebhard F: A randomized prospective study. J Hand
Conclusions The treatment of displaced intra-articular Surg Br 1995;20:365-372.
distal radius fractures in elderly patients.
A NMA of randomized trials re- Dtsch Arztebl Int 2014;111:779-787. 22. Ismatullah: Efficacy of plaster casting
versus external fixation in comminuted
vealed that ORIF with plates may 11. Costa ML, Achten J, Parsons NR, et al: distal radius fractures. J Postgrad Med Inst
Percutaneous fixation with Kirschner wires
offer the best results for adult and versus volar locking plate fixation in adults
2012;26:311-316.
elderly patients with a DRF, in terms with dorsally displaced fracture of distal 23. Jenkins NH, Jones DG, Mintowt-Czyz WJ:
of early and sustained functional radius: Randomised controlled trial. BMJ External fixation and recovery of function
2014;349:g4807. following fractures of the distal radius in
recovery and a reduction in fracture young adults. Injury 1988;19:235-238.
healing complications. It is currently 12. Egol K, Walsh M, Tejwani N, McLaurin T,
Wynn C, Paksima N: Bridging external 24. Jeudy J, Steiger V, Boyer P, Cronier P, Bizot
unclear if one specific PF device is fixation and supplementary Kirschner-wire P, Massin P: Treatment of complex
better (or worse) than another. fixation versus volar locked plating for fractures of the distal radius: A prospective
unstable fractures of the distal radius: A randomised comparison of external
randomised, prospective trial. J Bone Joint fixation “versus” locked volar plating.
Surg Br 2008;90:1214-1221. Injury 2012;43:174-179.
References
13. Földhazy Z, Ahrengart L: External fixation 25. Kapoor H, Agarwal A, Dhaon BK:
References printed in bold type are versus closed treatment of displaced distal Displaced intra-articular fractures of distal
radial fractures in elderly patients: A radius: A comparative evaluation of results
those published within the past 5 randomized controlled trial. Curr Orthop following closed reduction, external
years. Pract 2010;21:288-295. fixation and open reduction with internal
fixation. Injury 2000;31:75-79.
1. Nellans KW, Kowalski E, Chung KC: The 14. Goehre F, Otto W, Schwan S, Mendel T,
epidemiology of distal radius fractures. Vergroesen PP, Lindemann-Sperfeld L: 26. Khan JI, Hussain FN, Mehmood T, Adil O:
Hand Clin 2012;28:113-125. Comparison of palmar fixed-angle plate A comparative study of functional outcome
fixation with K-wire fixation of distal of treatment of intra articular fractures of
2. Lichtman DM, Bindra RR, Boyer MI, et al: radius fractures (AO A2, A3, C1) in elderly distal radius fixed with percutaneous
Treatment of distal radius fractures. J Am patients. J Hand Surg Eur Vol 2014;39: Kirschner’s wires vs T-plate. Pak J Med Sci
Acad Orthop Surg 2010;18:180-189. 249-257. 2017;33:709-713.
3. Jordan RW, Saithna A: Defining the role of 15. Gradl G, Gradl G, Wendt M, Mittlmeier T, 27. Leung F, Tu YK, Chew WY, Chow SP:
intramedullary nailing for fractures of the Kundt G, Jupiter JB: Non-bridging external Comparison of external and percutaneous
pin fixation with plate fixation for intra- of fractures of distal radius (AO type B 46. Cho CH, Lee SW, Jung GH: Comparison of
articular distal radial fractures: A and C): Volar plating versus nonoperative radiological and clinical outcomes of
randomized study. J Bone Joint Surg Am treatment. J Orthop Sci 2014;19: internal fixation using two different volar
2008;90:16-22. 537-544. plates for distal radius fractures. Hand Surg
2014;19:357-362.
28. Ludvigsen TC, Johansen S, Svenningsen S, 37. Stoffelen DV, Broos PL: Kapandji pinning
Saetermo R: External fixation versus or closed reduction for extra-articular distal 47. Dahl WJ, Nassab PF, Burgess KM, et al:
percutaneous pinning for unstable Colles’ radius fractures. J Trauma 1998;45: Biomechanical properties of fixed-angle
fracture: Equal outcome in a randomized 753-757. volar distal radius plates under dynamic
study of 60 patients. Acta Orthop Scand loading. J Hand Surg 2012;37:1381-1387.
1997;68:255-258. 38. ur Rahman O, Khan MQ, Rasheed H,
Ahmad S: Treatment of unstable 48. Jakubietz MG, Gruenert JG, Jakubietz
29. Radwan M: Displaced distal radius intraarticular fracture of distal radius: POP RG: Palmar and dorsal fixed-angle plates
fractures presented late: A randomized, casting with external fixation. J Pak Med in AO C-type fractures of the distal
prospective comparative study of two Assoc 2012;62:358-362. radius: Is there an advantage of palmar
methods of treatment. Internet J Orthop plates in the long term? J Orthop Surg
Surg 2009;13(1). 39. Venkatesh RB, Maranna GK, Res 2012;7:8.
Narayanappa RK: A comparative study
30. Marcheix PS, Dotzis A, Benko PE, Siegler J, between closed reduction and cast 49. Kamei S, Osada D, Tamai K, et al: Stability
Arnaud JP, Charissoux JL: Extension application versus percutaneous K-wire of volar locking plate systems for AO type
fractures of the distal radius in patients fixation for extra-articular fracture distal C3 fractures of the distal radius:
older than 50: A prospective randomized end of radius. J Clin Diagn Res 2016;10: Biomechanical study in a cadaveric model. J
study comparing fixation using mixed pins Rc05-Rc09. Orthop Sci 2010;15:357-364.
or a palmar fixed-angle plate. J Hand Surg
Eur Vol 2010;35:646-651. 50. Rozental TD, Beredjiklian PK, Bozentka
40. Vural OO, Guvenir H, Ozalp RT, Akkaya
DJ: Functional outcome and complications
MG, Yercan HS: Closed reduction and
31. Mellstrand Navarro C, Ahrengart L, following two types of dorsal plating for
casting versus Kapandji technique for the
Tornqvist H, Ponzer S: Volar locking plate unstable fractures of the distal part of the
treatment of Colles fractures: A prospective,
or external fixation with optional addition radius. J Bone Joint Surg Am 2003;85-A:
randomized clinical trial. Joint Dis Relat
of K-wires for dorsally displaced distal 1956-1960.
Surg 2008;19:55-60.
radius fractures: A randomized controlled
study. J Orthop Trauma 2016;30:217-224. 51. Tanaka H, Hatta T, Sasajima K, Itoi E,
41. Williksen JH, Frihagen F, Hellund JC,
Aizawa T: Comparative study of treatment
Kvernmo HD, Husby T: Volar locking
32. Plate JF, Gaffney DL, Emory CL, et al: for distal radius fractures with two different
plates versus external fixation and adjuvant
Randomized comparison of volar locking palmar locking plates. J Hand Surg Eur Vol
pin fixation in unstable distal radius
plates and intramedullary nails for unstable 2016;41:536-542.
fractures: A randomized, controlled study. J
distal radius fractures. J Hand Surg 2015;
Hand Surg 2013;38:1469-1476. 52. von Recum J, Matschke S, Jupiter JB, et al:
40:1095-1101.
Characteristics of two different locking
33. Pritchett JW: External fixation or closed 42. Young CF, Nanu AM, Checketts RG: compression plates in the volar fixation of
medullary pinning for unstable Colles Seven-year outcome following Colles’ type complex articular distal radius fractures.
fractures? J Bone Joint Surg Br 1995;77: distal radial fracture: A comparison of two Bone Joint Res 2012;1:111-117.
267-269. treatment methods. J Hand Surg Br 2003;
28:422-426. 53. Chaudhry H, Kleinlugtenbelt YV, Mundi
34. Rodriguez-Merchan EC: Plaster cast versus R, Ristevski B, Goslings JC, Bhandari M:
percutaneous pin fixation for comminuted 43. Zehir S, Calbiyik M, Zehir R, Ipek D: Are volar locking plates superior to
fractures of the distal radius in patients Intramedullary repair device against volar percutaneous K-wires for distal radius
between 46 and 65 years of age. J Orthop plating in the reconstruction of extra- fractures? A meta-analysis. Clin Orthop
Trauma 1997;11:212-217. articular and simple articular distal radius Relat Res 2015;473:3017-3027.
fractures; a randomized pilot study. Int
35. Rozental TD, Blazar PE, Franko OI, Orthop 2014;38:1655-1660. 54. Li-hai Z, Ya-nan W, Zhi M, et al: Volar
Chacko AT, Earp BE, Day CS: Functional locking plate versus external fixation for
outcomes for unstable distal radial fractures 44. Hammert WC, Kramer RC, Graham B, the treatment of unstable distal radial
treated with open reduction and internal Keith MW: AAOS appropriate use criteria: fractures: A meta-analysis of randomized
fixation or closed reduction and Treatment of distal radius fractures. J Am controlled trials. J Surg Res 2015;193:
percutaneous fixation: A prospective Acad Orthop Surg 2013;21:506-509. 324-333.
randomized trial. J Bone Joint Surg Am
2009;91:1837-1846. 45. Abe Y, Tokunaga S, Moriya T: Management 55. Harris CL, Kuhnke J, Haley J, et al: Best
of intra-articular distal radius fractures: Practice Recommendations for the
36. Sharma H, Khare GN, Singh S, Volar or dorsal locking plate-which has Prevention and Management of Surfical
Ramaswamy AG, Kumaraswamy V, fewer complications? Hand (N Y) 2017;12: Wound Complications, North York, ON,
Singh AK: Outcomes and complications 561-567. Canada, Wound Care Canada, 2018.
July 1, 2019, Vol 27, No 13 e605
Review Article
Musculoskeletal-based Patient-
reported Outcome Performance
Measures, Where Have We
Been—Where Are We Going
Abstract
Kent Jason Lowry, MD While health care evolves from volume to value, there is increasing
William Timothy Brox, MD interest by payors to use patient-reported outcomes (PROs) to
determine value and more specifically, quality from the patient’s
Peggy L. Naas, MD, MBA
perspective. This article reviews the current state of PROs and
Creighton Collins Tubb, MD discusses future directions. Specifically, this article will review the
George F. Muschler, MD current history and background of PROs; it will highlight the
Warren Dunn, MD, MPH perspective of the National Quality Forum and review the efforts of
the musculoskeletal community related to PROs. Goals, positive
aspects, limitations, and barriers related to PROs will be discussed.
Additionally, development considerations and strategies will be
highlighted. Finally, development recommendations from the
American Academy of Orthopaedic Surgery position statement
“principles for musculoskeletal based PRO performance
measurement development” will be introduced.
P atient-reported outcomes (PROs)

were first used in a research set-
ting in the 1980s. Following the
that is reported directly by the patient.
The definition of the National Quality
Forum (NQF) is “any report of the
report of Institute of Medicine from status of a patient’s health condition,
2001, an increasing interest was health behavior, or experience with
found in using PROs to link quality health care that comes directly from
and reimbursement programs with the patient, without interpretation of
From Northland Orthopaedics,
Rhinelander, WI (Dr. Lowry), Kaiser selected segments of clinical care. The the patient’s response by a clinician or
Permanente Fresno Medical Center, use of PROs is now migrating into anyone else.”2
Fresno, CA (Dr. Brox), the American routine patient care.1 As a result, an A “patient-reported outcome mea-
Academy of Orthopaedic Surgeons, increasing interest and focus on the
Rosemont, IL (Dr. Naas), New
sure” (PROM) takes the concept of a
Braunfels Orthopaedic Surgery & use of PROs were observed to assess PRO one step further by defining the
Sports Medicine, New Braunfels, TX the quality and effect of reimburse- actual instrument, tool, or metric
(Dr. Tubb), the Cleveland Clinic, ment. This article reviews the current that can be used to measure a PRO.
Cleveland, OH (Dr. Muschler), and the state of PROs related to musculo-
School of Medicine and Public Health, Again, the NQF defines PROMs as
University of Wisconsin-Madison, skeletal based care and discusses “various tools (eg, instruments, scales,
Madison, WI (Dr. Dunn). future directions. or single-item measures) that enable
J Am Acad Orthop Surg 2019;27: researchers, administrators, and oth-
e589-e595
Background ers to assess patient-reported health
DOI: 10.5435/JAAOS-D-18-00429 status for physical, mental, and social
The term “patient-reported outcome” well-being.”3
Academy of Orthopaedic Surgeons. (PRO) is a generic term that can In 2004, the National Institutes of
encompass any outcome assessment Health established the Roadmap for
July 1, 2019, Vol 27, No 13 e589
Musculoskeletal-based Patient-reported Outcome
Medical Research and from this effort second measurement using the same have not been widely adopted for
developed the PROM Information PROM at a defined window of time. clinical use outside research settings in
System (PROMIS) with the goal of The resulting comparison is reported the United States.”3 Also the NQF
providing clinicians and researchers as a change score and is used as a recognizes “several method-related
access to standardized, valid, and measure of the effectiveness of a given challenges such as aggregating patient
flexible PRO data collection tools to intervention.3 data on PROMs to measure perfor-
assess health status that assess physi- mance at multiple levels of analysis (eg,
cal, mental, and social well-being from individual, group practice, organiza-
the patient’s perspective. The National Quality Forum tion) and use of proxy respondents.3
Examples of PROMs include and Patient-reported The NQF recognizes that future de-
PROMIS-10, Hip disability and Outcomes velopment will require further exami-
Osteoarthritis Outcome Score (HOOS), nation of the best practices for using
Knee disability and Osteoarthritis The NQF is a not-for-profit, nonpar- proxy respondents, assessing the
Outcome Score (KOOS), and the like. tisan, membership-based organization equivalency of data obtained from
Therefore, a PRO may include the that works to catalyze improvements different PROMs, calibrating multiple
capture of more than one PROM in health care. The NQF board mem- individual-level PROMs to a standard
because some PROMs provide met- bers represent a range of health care scale, aggregation of multiple PRO-
rics of general health (eg, PROMIS- leaders from the public and private PMs to complete a PRO picture, using
10, VR-12), and other PROMs are sectors. They include providers, pur- technology to reduce patient burden
designed to provide more disease chasers, consumers, and also rep- and increase response rates, and how
specific measures (eg, HOOS, KOOS). resentatives from the Agency for PROMs will be incorporated and data
However, “PRO performance mea- Healthcare Research and Quality, the standards developed into electronic
surement” (PRO-PM) captures yet Centers for Disease Control and Pre- medical record (EMR).3
another concept, that is, the evolution vention, the Centers for Medicare and
of the use of PROMs as a means for Medicaid Services, and the Health Re-
the measurement of clinical perfor- sources and Services Administration. The Musculoskeletal
mance, quality, and even value in The NQF released its “PROs in Community and Patient-
medicine. These PRO-PMs are being Performance Measurement” report reported Outcomes
considered as a method by which to in early 2013.3 This report con-
determine reimbursement. This might cludes that “patients remain an The American Academy of Ortho-
be done responsibly, according to the untapped resource in assessing the paedic Surgery (AAOS) and the
NQF, if reimbursement were “based quality of health care.” To achieve orthopaedic community are embrac-
on PRO data (from one or more PRO- the goal of accessing this untapped ing the importance of PROs. This is
PMs) aggregated for an entity deemed resource, the NQF acknowledges evidenced by the work of the Ortho-
as accountable for the quality of care two key goals: (1) “to engage patients paedic Quality Institute, the AAOS
or services delivered. Such entities can by building capacity and infrastruc- Summit on PROs, the Quality Out-
include (but would not be limited to) ture to capture PROs routinely and comes Data work group review of
long-term support services providers, (2) to use these data to develop PROMs and data collections systems,
hospitals, physician practices, or ac- performance measures to allow for and the work of the AAOS Perfor-
countable care organizations.”3 Gen- accurate appraisals of quality and mance Measures Committee.
erally accepted PRO-PM assessment efficiency.”3 In 2015, the AAOS released infor-
includes a baseline or preintervention The NQF also recognizes several mation statement 1044, Principles of
PROM assessment followed by a challenges. In particular, “PROMs Patient Reported Outcome Measures
Dr. Lowry or an immediate family member serves as a paid consultant to Security Health Plan and serves as a board member, owner, officer,
or committee member of the American Academy of Orthopaedic Surgeons, ASTM, and AMA—PQPI. Dr. Naas an immediate family member
serves as a board member, owner, officer, or committee member of the American Academy of Orthopaedic Surgeons, American Society of
Anesthesiologists, Steering Committee, Perioperative Surgical Home Collaborative, University of Minnesota Department of Orthopaedic
Surgery, Liaison, Orthopaedic Surgeon Well-being Project. Dr. Tubb or an immediate family member serves as a board member, owner,
officer, or committee member of the American Academy of Orthopaedic Surgeons and American Association of Hip and Knee Surgeons.
Dr. Muschler or an immediate family member has received royalties from Fortus; serves as a paid consultant to the National Institutes of
Health; serves as an unpaid consultant to Parker Hannifin; and has received research or institutional support from Fortus. Neither of the
following authors nor any immediate family member has received anything of value from or has stock or stock options held in a commercial
company or institution related directly or indirectly to the subject of this article: Dr. Brox and Dr. Dunn.
Kent Jason Lowry, MD, et al
(PROM) Reporting,4 with included ments, and its dashboard system al- the members’ understanding of the
PROM selection criteria. lows for aggregated site-specific topic, and its associated evidence,
The AAOS Summit on PRO de- reports on these additional PROMs. and therefore improve the final prod-
velopment related to total joint ar- Participants in the registry can submit uct. In 2018, the AAOS released
throplasty was convened later in PROM data via collection within the Position Statement #1188: Principles
2015. Participants included repre- AJRR PRO platform, data from inter- for Musculoskeletal Based PRO-PM
sentatives from orthopaedic organ- nal PROM collection systems submit- Development, which was developed
izations (American Association of ted via .csv or .xls, or via an external by the PMC before merging with the
Hip and Knee Surgeons, AAOS, The data collection vendor (pending ven- EBQV and establishes processes for
Hip Society, The Knee Society, and dor agreement with AJRR). PRO-PM development.7
American Joint Replacement Regis-
try [AJRR]), Center for Medicare
& Medicaid Services (CMS), Yale Performance Measures Goals and Positive Aspects
New Haven Health Service Corpo- Committee and Patient- of Patient-reported
ration (YNHHSC)/Center for Out- reported Outcome Measure Outcome Measure and
comes Research & Evaluation and Patient-reported Patient-reported Outcome-
(CORE), National Committee for Outcome-Performance Performance Measurement
Quality Assurance , Mathematica, Measurement Development Use
CECity, and Blue Cross Blue Shield.
This summit recommended a hybrid In 2014, the AAOS created the Per- Using PROMs “provides insights into
approach to PROM use and encour- formance Measurement Committee patients’ experiences of symptoms,
aged CMS to use a staged approach to (PMC) and charged it with facilitat- quality of life, and functioning; values
selecting candidate risk variables.5 ing the development and use of per- and preferences; and goals for health
In February 2016, the AAOS formance measures related to the care.”8 Incorporating PROMs into
Quality Outcomes Data work group musculoskeletal system. The PMC PRO-PMs allows an objective as-
reviewed the existing PROMs and has defined processes for perfor- sessment of a health care intervention
recommended a list of PROMs to mance measurement development to influence the patient’s life. This
consider for PRO-PM development and systematically established work concept is simple but represents a
and detailed the criteria used for groups to develop performance paradigm shift from how medicine
PROM selection.6 measures based on previously estab- has previously measured success.
lished AAOS Clinical Practice Guide- The most basic level of medical
lines (CPGs) and Appropriate Use training focuses on treatment outcome
The American Joint Criteria (AUC). The PMC established (eg, radiographic union of a fracture;
Replacement Registry work groups in the area of Arthritis stability of a ligament reconstruction;
Pain and Function Assessment, Hip range of motion following joint re-
The AJRR, established in 2010, Fractures in the Elderly, Anterior placement surgery). In contrast, a PRO-
launched its PRO data submission plat- Cruciate Ligament Injury, and Sur- PM determines the treatment outcome
form in 2015. The AJRR recommends gical Management of Osteoarthritis from the perspective of the patient’s
using both a health-related quality of life of the Knee. In 2018, the PMC specific experience and goals. This re-
measure and a joint-specific measure merged into the Committee on Evi- quires understanding a preintervention
when collecting PRO data. The Registry dence-Based Quality and Value (baseline) state and the resultant
has recommended PROMs for national (EBQV), which also oversees the change to a postintervention (follow-
benchmarking (health-related quality development of AAOS CPGs and up) state. The magnitude of change is a
of life measure measures VR-12 or AUC. The AAOS Performance Mea- measure of performance of the health
PROMIS-10 Global Health and sures are based on AAOS CPGs and care system related to that individual
joint-specific measures HOOS, JR. reside alongside AAOS AUC in the patient. In aggregate, PRO-PM data
and KOOS, JR) based on their partic- Quality Cycle. The incorporation of highlight the value orthopaedic sur-
ipation in the 2015 Patient-Reported Performance Measures into the Com- geons bring to the lives of their patients.
Outcome Summit for Total Joint Ar- mittee on EBQV also enables the Also PRO-PMs enable objective
throplasty Summit. The AJRR also committee members serving as Over- comparison of treatment options using
provides a compiled list of PROMs for sight Chairs on work groups to follow a standardized scale. Integrating PRO-
hospitals that would like to collect data a topic from CPG to AUC, and finally PM baseline, follow-up, and change
outside of these recommended instru- to PM. This continuity would deepen scores with patient demographics,
July 1, 2019, Vol 27, No 13 e591
disease burden, disease severity and pate is sharing symptoms, informa- or environmental patient-related fac-
comorbidity data enables further lay- tion about treatments, and outcomes. tors) will influence patient access to
ers of analysis that may define key More than 600,000 people share on care or penalize providers providing
variables, factors, and interactions. the site www.PatientsLikeMe, and care to higher-risk patients. As a result,
Such predictive modeling could greatly researchers have harvested those data the implications of modifiable and
enhance shared decision making, to advance treatment and comparative more importantly, nonmodifiable
treatment comparisons, and even outcomes. patient-specific risk factors on
comparisons between clinicians and PRO-PM remains an elusive topic in
health systems. Using such data and need of more clarity. Additional re-
models, a provider could educate a Limitations/Barriers to search related to these topics is
patient on what treatment options are Patient-reported Outcome- urgently needed. Unfortunately, the
available, the range of probable out- Performance Measurement role and effect of patient-related risk
comes and likelihood of improvement, Use factors on PRO-PM is difficult to eval-
and the effect of modifiable factors uate until there is an agreement and
(such as obesity, diabetes, chronic Implementation of PRO-PM will be an availability for a given PRO-PM, which
obstructive pulmonary disease, con- important advance, but what are the further highlights the need for a road-
gestive heart disease, or disease in other barriers and limitations? Missing data map related to PRO-PM development.
joints) that can and should be ad- elements both at a point in time and The American College of Surgeons
dressed. This level of understanding has missed collection at a data set window (ACS) have begun early investigations
the potential to greatly enable the con- are significant problems when using into risk stratification through the
cept of a truly individualized patient- these data sets to analyze and report National Surgical Quality Improve-
informed consent that has already been care. Thankfully, there are algorithms ment Program data collection efforts
applied to many patients facing major to interpolate missing data and analyze and the development of the ACS risk
decisions in such settings such as pros- the significance of missing data. Au- calculator (developed “using data
tate cancer and heart surgery. thors should be reporting measures collected from more than 3.8 million
Looking beyond the individual pa- of data quality, including measures of surgeries from 740 hospitals partici-
tient encounter, PRO-PMs provide the adherence to collection windows and pating in ACS National Surgical
opportunity for providers and systems missing data. Investigators need stan- Quality Improvement Program from
to share and compare benchmark data dardized methods to consider the 2012 to 2016”).11 Although the risk
and assess their quality in comparison interplay of generalized musculoskeletal calculator is primarily used as a
to quality in other centers. This could problems, systemic pathology, and patient assessment tool, the effect of
have a strong positive effect, enabling mental health with localized musculo- patient risk factors related to PRO-PM
the rapid discovery and dissemination skeletal pathology. Currently, these on patient access and reimbursement
of best practice models and patient- measures are limited if available at all. models remains unclear.
specific care paths. Risk stratification and its effect on Recognizing this fundamental shift
Further beyond individual patient PRO-PM’s continue to be a signifi- in performance measurement assess-
care, national health care policymakers cant and under evaluated aspect of ment, the California Orthopedic Soci-
and health care payers would like to PRO-PM development. In August of ety commissioned a work group with
use these data to make more informed 2015, a summit for PRO develop- the goal to provide “recommendations
decisions on the relative value of vari- ment related to total joint arthro- on practical, cost-effective processes,
ous procedures or resources and to plasty was convened, and this work and standards to encourage more
drive health care value initiatives. They group recommended that “CMS widespread, consistent use of PRO
also hope to use PROM and PRO-PM should use a staged approach in se- instruments for orthopaedic surgeons
data to guide the design of incentives lecting the candidate variables for risk treating shoulder, hand, spine, foot,
and disincentives that may reduce over adjustment . . .. It is essential that risk and ankle conditions.”12 This review
utilization, guide patients to “higher adjusted data be collected or access highlights the need for risk stratifica-
value” providers, and to more fairly to care for certain patients will be tion of PRO data and summarizes
reimburse caregivers serving higher- limited in the future.”10 Despite these the potential orthopedic comorbidities/
risk patients. recommendations, payers continue to risk factors by specialty.
Prioritization of health care goals has seek PRO-PM without a clear un- Socioeconomic factors are not cur-
improved, so outcomes to be measured derstanding of how failure to prop- rently included in the US federal gov-
reflect those most important to patients erly apply risk stratification (whether ernment approach to risk ratification
themselves.9 Patients want to partici- socioeconomic, behavioral, medical, including patient outcomes. Authors
need to analyze whether differences in population of interest and be able to generate a calibrated bank of items.
measured patient functional outcomes discriminate subjects at the extremes The test starts with the same starting
are explained by population socio- of the spectrum. When an instrument item for all, then based on the
economic risk factors.13-17 is unable to discriminate subjects on response to the first item, subsequent
Patient compliance factors can be the extreme ends of a spectrum, it is items are selected to obtain the most
problematic in use of PRO-PM. typically because of floor and/or ceil- information possible using item se-
Problems include patient administra- ing effects. The ceiling effect occurs lection algorithms. Scoring rules are
tion using different administration when the data gathered are unable to built into the algorithm, as well as
techniques, surrogate responses, and reliably distinguish between subjects stopping rules. Another advantage of
language barriers. There may also be with the highest scores, whereas the CAT is immediate score reporting as
challenges regarding secondary gain. floor effect occurs when there is a the final element of the algorithm.
If the patient has a secondary agenda lower limit of reliable discrimination The disadvantage of CAT is that it
regarding disability or medical medi- among respondents. A simple exam- requires expertise and specific soft-
cation prescription, their response ple of this is measuring income level ware that can be expensive.
validity may be problematic. with predetermined categories: de-
Furthermore, there are technology, pending upon the ranges in each cat- Patient-reported Outcome
legal, and clinical process limitations. egory it is possible to create these Performance Measurement as a
Problems include adherence to data effects where you cannot distinguish Practice Comparison Tool
dictionaries. There are data sharing differences in income in the lowest The presumption that there will be a
limitations in the United States that category (floor) or the highest cate- change in a subject’s health status fol-
make it difficult and sometimes gory (ceiling). Depending on what is lowing an intervention is the basis of
impossible to pool and report data being measured, this may not be det- measuring change. Using some esti-
across large populations. rimental; however, if it is important to mate to quantify this prepost change, a
When PRO-PM data are used to distinguish between those at the ex- change score, is intuitive and often
compare populations of treatment tremes of a measurement scale, this used to make comparisons, such as
protocols, it is critical to be clear that can be very problematic and should be between different interventions or
the underlying populations are com- taken into consideration when devel- between different providers. However,
parable. For example, if age differences oping a PRO-PM. some serious limitations to some
are present between two populations common change scores exist that
being compared, comparing mortality Potential Use of Computerized should be noted. If the postmeasure-
may be associated with the age differ- Adaptive Testing ments are independent of the pretest
ence rather than treatment. Computerized adaptive testing (CAT) measurements, then a simple, or
Furthermore, complexity and varia- is a test administered by computer that absolute, difference (post value 2 pre
tion of the PRO-PM development may dynamically adjusts itself to the trait value) is easy to interpret and calcu-
result in competing and conflicting level of each examinee while the test is late. However, most of the time, the
implementation needs. The problem being administered. The CAT has the biggest predictor of the post value is
of minimally clinically important dif- advantage of lessening responder the pre value; in other words, they are
ference is noted.18,19 Because a statis- burden by eliminating items that not independent. Subjects with the
tically measurable difference is noted, would not apply to the examinee, highest prescores will have the largest
there may not be a clinically relevant thereby substantially shortening the pre-post differences. It is impossible
difference in outcomes between treat- time needed to administer the test. For to determine whether one patient
ment protocols. instance, if a functional scale is con- achieved a better outcome than an-
structed to measure a wide range of other by comparing simple change
function from sedentary to high levels scores unless you also know if their
Key Patient-reported
of physical function, items at the lower prescores were comparable. Likewise,
Outcome-Performance
end of physical activity, such as comparing average change scores
Measurement Development walking a few blocks, would be of no among different providers is similarly
Considerations and value in identifying a subject’s func- difficult to interpret because the pro-
Strategies tional status that would subsequently viders may differ significantly in the
answer yes to an item that asks if they mix of cases being treated. Although
Floor and Ceiling Effects are capable of running three miles. the absolute difference has these
Ideally, a PRO-PM should measure Typically, CAT tests are designed limitations, it is at least a symmetric
the entire range of health status in the using item response theory used to measurement. Conversely, percent
July 1, 2019, Vol 27, No 13 e593
change (post 2 pre/pre) is an asym- (3) Utilization of consistent PROM Publications/2012/12/Patient Reported_
Outcomes_in_Performance_Measurement.
metric measurement bound on the left tools across varying diagnosis. aspx. Accessed October 31, 2018.
at 2100 and not bound on the right (4) Guide future PROM develop-
4. American Academy of Orthopaedic
to infinity. Also, these can be very ment with the goal of being Surgeons: Information Statement:
skewed on the right when the post- nonproprietary (free use by clini- Principles of Patient Reported Outcome
values are much larger than the cians), have adaptability/usability Measures (PROMs) Reporting. 2015.
https://www.aaos.org/uploadedFiles/
prevalues. For these reasons, the best across differing diagnoses (is PreProduction/About/Opinion_Statements/
approach is to use a regression model, crosscutting), ensures that the advistmt/1044%20Principles%20of%
20Patient%20Reported%20Outcome%
such as analysis of covariance, with measured outcome is changeable 20Measures%20(PROMs)%20Reporting.
the postvalue as the dependent vari- by the measured intervention, and pdf. Accessed October 31, 2018.
able and the prevalues are adjusted patient-centered characteristics,
5. American Association of Hip and Knee
for by including them as an inde- which facilitate identification of Surgeons: Patient Reported Outcomes
pendent covariate along with other outcome goals important to in- Summit for Total Joint Arthroplasty
Report. 2015. http://www.aahks.org/wp-
covariates of interest such as the dividual patients and leads to content/uploads/2016/08/AAHKS-2015-
treatment and/or provider. enhanced patient engagement Outcomes-Summit-Report.pdf. Accessed
with his or her care. October 31, 2018.

American Academy of Surgeons: Patient Reported Outcome
Summary Measures. 2016. https://www.aaos.org/
Orthopaedic Surgeons quality/performance_measures/patient_
Patient-reported Outcome- In conclusion, PRO-PMs carry the
reported_outcome_measures/. Accessed
October 31, 2018.
Performance Measurement potential to profoundly change the
Development delivery of health care and positively
Surgeons: Position Statement #1188:
Recommendations influence an individual patient’s abil- Principles for Musculoskeletal Based
ity to make decisions related to the Patient Reported Outcome-performance
Recognizing the many challenges Measurement Development. 2018. https://
care being recommended. However, www.aaos.org/uploadedFiles/
noted above related to PRO-PM de- given the young state of PRO-PM PreProduction/About/Opinion_Statements/
velopment and the evolving devel- development and the often-competing position/1188%20Principles%20for%
opment needs, the AAOS believes 20Musculoskeletal%20Based%20Patient
PRO-PMs under development, the %20Reported%20Outcome-Performance
that this development should be AAOS believes in creating uniform %20Measurement%20Development.pdf.
guided by uniform, strategic goals. development strategies for musculo-
Accessed October 31, 2018.
Based on the work of the AAOS’s skeletal based PRO-PMs, which 8. Lavallee D, Chenok K, Love R, et al:
PMC, the AAOS is recommending will further advance PRO-PM im-
Incorporating patient-reported outcomes into
health care to engage patients and enhance
uniform PRO-PM development plementation and acceptance. From care. Health Affairs 2016;25:575-582.
goals, with the intent to create a this belief, the AAOS supports the 9. Hsiao B, Fraenkel L: Incorporating the
stable foundation for future devel- “principles for musculoskeletal based patient’s perspective in outcomes research.
opment. These recommendations PRO-PM development.” Curr Opin Rheumatol 2017;29:144-149.
include the following:
10. Patient Reported Outcomes Summit for
(1) A three-staged PRO-PM de- Total Joint Arthroplasty Report. American
velopment and implementation References Association of Hip and Knee Surgeons,
2015. http://www.aahks.org/wp-content/
process consisting of Stage 1: uploads/2016/08/AAHKS-2015-
Routine PRO use in clinical based References printed in bold type are Outcomes-Summit-Report.pdf. Accessed
practice—a process measure, those published within the past 5 June 24, 2017.
Stage 2: Pre- and posttreatment years. 11. ACS NSQIP: Surgical Risk Calculator.
comparison—a process mea- https://riskcalculator.facs.org/
1. Institute of Medicine: Crossing the Quality
RiskCalculator/about.html. Accessed
sure, and Stage 3: Outcome- Chasm: A New Health System for the 21st
September 18, 2018.
Century. Washington, DC, National
based PRO-PM comparison of Academies Press, 2001. 12. Glassman JR, Unti L, ABidi N: Patient-
pre- and posttreatment. Reported Functional Outcomes: How to
2. U.S. Food and Drug Administration:
(2) Time frames for administering Guidance for industry, patient-reported collect and Report Risk-Adjusted
Musculoskeletal Patient-Reported
PROMs in performance mea- outcome measures: Use in medical product
development to support labeling claims. Functional Outcome Data in an
surement are collected and de- Fed Resist 2009;74:65132-65133. Orthopedic Practice. California
fined in a predictable manner Orthopedic Associations, 2016. https://coa.
3. National Quality Forum: Patient Reported org/docs/WhitePapers/
consisting of a Preintervention, Outcomes (PROs) in Performance. 2013. PROWhitePaperFinal.pdf. Accessed
Early, Mid, and Late time frames. https://www.qualityforum.org/ September 26, 2018.
13. Daugbjerg SB, Cesaroni G, Ottesen B, 15. Qasim M, Andrews RM: Despite overall ethnicity and socio-economic differences in
Diderichsen F, Osler M: Effect of improvement in surgical outcomes since breast cancer surgery outcomes. Cancer
socioeconomic position on patient 2000, income-related disparities persist. Epidemiol 2015;39:745-751.
outcome after hysterectomy. Acta Health Aff (Millwood) 2013;32:
Obstet Gynecol Scand 2014;93: 1773-1780. 18. Jaeschke R, Singer J, Guyatt GH:
926-934. Measurement of health status: Ascertaining
16. Barrack RL, Ruh EL, Chen J, et al: Impact the minimal clinically important difference.
14. Neuburger J, Hutchings A, Black N, van of socioeconomic factors on outcome of Control Clin Trials 1989;10:407-415.
der Meulen JH: Socioeconomic differences total knee arthroplasty. Clin Orthop Relat
in patient-reported outcomes after a hip or Res 2014;472:86-97. 19. McGlothlin A, Lewis R: Minimal clinically
knee replacement in the English National important difference: Defining what really
Health Service. J Public Health (Oxf) 2013; 17. Akinyemiju TF, Vin-Raviv N, Chavez- matters to patients. JAMA 2014;312:
35:115-124. Yenter D, Zhao X, Budhwani H: Race/ 1342-1343.
July 1, 2019, Vol 27, No 13 e595
Research Article
Obstructive Sleep Apnea and

Arthroscopic Rotator Cuff
Repair—Are Complication Rates
Really Increased?
Abstract
Jourdan M. Cancienne, MD Background: A few investigations exist which evaluate the influence
Stephen F. Brockmeier, MD of obstructive sleep apnea (OSA) on complications after arthroscopic
rotator cuff repair.
Matthew J. Deasey, MD
Methods: A database was queried for patients undergoing rotator
Brian C. Werner, MD cuff repair with and without OSA and further subdivided into those with
From the Department of Orthopaedic and without a billing code for a continuous positive airway pressure
Surgery, University of Virginia, (CPAP) device. Thirty-day and 6-month adverse events were
Charlottesville, VA.
assessed.
Correspondence to Dr. Werner: Results: After regression analysis, patients with OSA had markedly
bcw4x@virginia.edu
increased emergency department visits and hospital admission (P ,
Dr. Brockmeier or an immediate family
0.05). This risk was mitigated by CPAP orders compared with control
member is a member of a speakers’
bureau or has made paid subjects. Patients without CPAP use had markedly increased risks of
presentations on behalf of and serves emergency department visits, hospital admission, and respiratory
as a paid consultant to Arthrex,
Zimmer Biomet, and Exactech; has
complications compared with control subjects (P , 0.05).
received research or institutional Conclusions: Patients with OSA have higher risks of emergency
support from Arthrex, Zimmer Biomet; department visits and hospital admissions postoperatively;
serves as a board member, owner,
officer, or committee member of the
however, a CPAP order appears to mitigate this risk. The independent
American Orthopaedic Society for risk imparted by OSA for the studied complications was markedly
Sports Medicine, the American lower than other comorbidities.
Shoulder and Elbow Surgeons, the
International Society of Arthroscopy,
Knee Surgery, and Orthopaedic
Sports Medicine, MidAtlantic Shoulder
and Elbow Society. Dr. Werner or an
immediate family member serves as a
board member, owner, officer, or
O bstructive sleep apnea (OSA) is
the most common sleep disor-
der and is characterized by partial or
after inpatient procedures, including
early readmission and post-discharge
death.1,5-9 Consequently, the suitabil-
committee member of the American complete airway collapse during ity of ambulatory surgery in patients
Orthopaedic Society for Sports
Medicine and the American Shoulder
sleep.1 Recent studies reported that with OSA has become increasingly
and Elbow Surgeons. Neither of the up to 20% of patients presenting for controversial because of similar con-
following authors nor any immediate elective surgery carry a diagnosis of cerns despite a paucity of data on
family member has received anything OSA, and approximately 80% of perioperative complications in patients
of value from or has stock or stock
options held in a commercial company
these patients are newly diagnosed undergoing ambulatory surgery with
or institution related directly or on preoperative screening.2,3 Given OSA.
indirectly to the subject of this article: the rising prevalence and identifica- With nearly 300,000 procedures
Dr. Cancienne and Dr. Deasey. tion of new cases of OSA, the pro- performed annually, arthroscopic
J Am Acad Orthop Surg 2019;27: portion of patients with OSA rotator cuff repair (RCR) represents
486-492 presenting for elective, ambulatory one of the most common ambulatory
DOI: 10.5435/JAAOS-D-18-00069 surgery is expected to drastically orthopaedic procedures in the United
increase.4 Previous literature has States.10 Of all ambulatory surgeries
Academy of Orthopaedic Surgeons. associated OSA with an increased performed at our institution, we
risk for perioperative complications have anecdotally seen the largest
Jourdan M. Cancienne, MD, et al
amount of scrutiny from anesthesia mation derived from insurance re- surgery (control group). The following
about performing RCR in a beach cords of both private-payer and ICD-9 and ICD-10 codes were used to
chair position in patients with OSA. Medicare providers. The database identify the presence or absence of
However, evidence is scarce in the contains all information for patients OSA: ICD-9 327.23 and ICD-10-D-
current literature to provide evidence coded by Current Procedural Termi- G47.33. The study group was then
for this and guide the perioperative nology (CPT) and International Clas- further divided into a cohort of patients
treatment of patients with OSA sification of Diseases, 9th revision with a reimbursed prescription for a
undergoing RCR. This is particularly (ICD-9) and 10th revision (ICD-10) CPAP before the surgical date and a
important given that patients un- codes. Additionally, the private in- second cohort of those without an
dergoing RCR may be at an even surance database contains prescrip- identified prescription for a CPAP.
greater risk of postoperative com- tion records and laboratory values. CPAP prescriptions were identified
plications in the presence of OSA All provided data are completely de- using CPT codes E0601 and 94660.
because of certain procedure-specific identified and have thus been deemed Detailed patient information for all
variables. For instance, preopera- exempt by our Institutional Review study groups and the control group,
tively patients are often given a sca- Board. For this study, the private- including demographics and co-
lene nerve block that could result payer database with patients from morbidities, is provided in Table 1.
in phrenic nerve paralysis, fluid Humana was used because it provides
extravasation into the chest wall in- prescription information that was Complications
traoperatively may result in de- necessary to determine CPAP usage,
The main outcome measure was
creased chest wall expansion, and which is not provided in the Medicare
adverse events or complications rele-
postoperatively these patients often files. When the analysis was per-
vant to OSA postoperatively. Six
sleep in an upright position for informed, the database covered patients
major outcome measures were eval-
creased comfort. Any number of these from 2007 to 2016-Q1.
uated, all within 30 days postopera-
factors could possibly exacerbate a
tively, except mortality. These
baseline diagnosis of OSA and predis- Study Cohorts included an emergency department
pose these patients to complications
Patients who underwent arthroscopic visit, hospital admission, PE, acute
not seen after other ambulatory
rotator cuff repair in an ambulatory MI, respiratory arrest, and in-
surgery. Given an increasing number
setting were queried in the database hospital mortality. The outcomes
of ambulatory RCR procedures
using CPT code 29827. Additional were all determined using CPT and
performed in patients with OSA, it is
inclusion criteria were a minimum ICD-9/10 codes. Unfortunately,
of high clinical significance to de-
postoperative database exposure of because of database limitations, the
termine the relationship between
6 months, which was required to limit diagnosis associated with presenta-
OSA and perioperative complications,
the inclusion of patients who changed tion to an emergency department or
including readmission, emergency
health insurance during the study admission to a hospital is unable to be
department visitation, pulmonary
period and may have been transient obtained.
embolism (PE), myocardial infarc-
within the database. Exclusion crite-
tion (MI), respiratory arrest, and in-
ria included (1) previous rotator cuff Statistical Analysis
hospital mortality after ambulatory
repair on the ipsilateral shoulder
RCR. Therefore, this study evaluates Complication rates were first pre-
within the database period and (2)
OSA as an independent risk factor sented for all study groups and the
patients who underwent concomitant
for perioperative complications after control group. Next, a binomial
open rotator cuff repair to eliminate
ambulatory RCR and further this multivariable logistic regression
any potential bias.
analysis by examining the effect of analysis was used to examine the
continuous positive airway pressure independent effect of OSA and CPAP
(CPAP) on risk mitigation. Study Groups prescriptions while controlling for
Two groups of patients were then additional covariates: age at the time
formed from those meeting the of surgery, sex, body mass index,
Methods inclusion and exclusion criteria. The alcohol abuse, tobacco use, knee
first group contained patients with osteoarthritis diagnosis, inflammatory
Database diagnosed OSA at or before the sur- arthritis diagnosis, depression, diabetes
This study used the PearlDiver patient re- gical date (study group) and the sec- mellitus, hypertension, hyperlipidemia,
cords database (www.pearldiverinc.com). ond group contained patients without peripheral vascular disease, coronary
This database contains patient infor- diagnosed OSA on or before the date of artery disease, congestive heart failure,
July 1, 2019, Vol 27, No 13 487
Sleep Apnea and Rotator Cuff Repair
Table 1
Patient Demographics and Comorbidities
Control Group (n = 35,859) OSA Group (n = 4,529)
Characteristic n % n %
Demographics
Age, yr
Less than 40 801 2.2 35 0.8
40-49 2,709 7.6 240 5.3
50-59 6,858 19.1 966 21.3
60-69 13,660 38.1 1,913 42.2
70-79 10,480 29.2 1,278 28.2
801 1,351 3.8 97 2.1
Sex (male) 18,661 52.0 2,628 58.0
Obesity (BMI, 30-39.9 kg/m2) 7,338 20.5 1,432 31.6
Morbid obesity (BMI, 401 kg/m2) 2,944 8.2 1,393 30.8
Tobacco use 5,891 16.4 913 20.2
Alcohol abuse 1,246 3.5 190 4.2
Comorbidities
Diabetes mellitus 11,943 33.3 2,483 54.8
Hyperlipidemia 27,694 77.2 4,052 89.5
Hypertension 26,420 73.7 4,100 90.5
Peripheral vascular disease 4,044 11.3 802 17.7
Congestive heart failure 3,492 9.7 997 22.0
Coronary artery disease 9,301 25.9 1,903 42.0
Chronic kidney disease 5,035 14.0 939 20.7
Chronic lung disease 6,882 19.2 1,583 35.0
Chronic liver disease 2,709 7.6 679 15.0
Thyroid disease 9,067 25.3 1,514 33.4
Depression 9,349 26.1 1,985 43.8
Hemodialysis 267 0.7 59 1.3
chronic kidney disease, chronic lung these, 4,529 patients (11.2%) had a (0.66% versus 0.35%), MI (0.22%
disease, chronic liver disease, thyroid diagnosis of OSA, whereas 35,859 versus 0.16%), respiratory arrest
disease, hypercoagulability, and current patients (88.8%) did not have a (0.91% versus 0.35%), and in-
hemodialysis use. P , 0.05 was used as diagnosis for OSA and were used as hospital mortality within 6 months
the threshold for statistical significance. the control group. Of the patients (0.11% versus 0.09%) (Table 2).
with OSA, 1,976 (43.6%) had a When comparing unadjusted com-
reimbursed CPAP prescription. Pa- plication rates between patients with
Results tients with OSA had a higher inci- OSA with and without a CPAP
dence of most medical comorbidities order, patients with OSA with a
In total, 3,923 patients were excluded studied (Table 1), which was subse- CPAP prescription trended toward
from the analysis. This included 985 quently controlled for in the logistic lower complication rates than those
patients who had undergone a previ- regression analysis. without a CPAP order (Table 2).
ous rotator cuff surgery on the ipsi- Before controlling for demograph- After controlling for confounding
lateral extremity and 2,938 patients ics and medical comorbidities, pa- covariates with a regression analysis,
who underwent a concomitant open tients with OSA had higher rates of patients with a diagnosis of OSA
rotator cuff repair procedure at the 30-day emergency department visits had a markedly higher incidence of
time of surgery. A total of 40,388 (8.77% versus 5.63%), hospital emergency department visits (odds
unique patients were included. Of admission (2.89% versus 1.52%), PE ratio [OR], 1.25 [1.11 to 1.41];
Table 2
Comparison of Complications After Arthroscopic Rotator Cuff Repair
Control Group OSA Overall OSA With CPAP
(N = 35,859) (N = 4,529) (N = 1,976) OSA With No CPAP (N = 2,553)
Complication N % N % N % N %
Emergency department visit (30 d) 2,018 5.63 397 8.77 159 8.05 238 9.32
Hospital admission (30 d) 545 1.52 131 2.89 59 2.99 72 2.82
Pulmonary embolism (30 d) 125 0.35 30 0.66 10 0.51 20 0.78
Myocardial infarction (30 d) 59 0.16 10 0.22 4 0.20 6 0.24
Respiratory arrest (30 d) 127 0.35 41 0.91 15 0.76 26 1.02
In-hospital mortality (6 mo) 32 0.09 5 0.11 2 0.10 3 0.12
CPAP = continuous positive airway pressure, OSA = obstructive sleep apnea
P = 0.0003) and hospital admission for a CPAP with those without a without OSA undergoing RCR had
(OR, 1.23 [1.06 to 1.43]; P = 0.034) diagnosis of OSA undergoing RCR, markedly higher rates of emergency
within 30 days compared with con- patients with OSA and a CPAP order department visits (OR, 1.38 [1.19 to
trol subjects without OSA (Table 3). had similar rates of all perioperative 1.60]; P , 0.0001), hospital read-
No notable difference was observed complications studied (Table 3). mission (OR, 1.24 [1.05 to 1.47]; P =
in the incidence of PE, MI, and After controlling for confounding 0.013), and respiratory arrest (OR,
respiratory arrest within 30 days covariates with a regression analysis, 1.57 [1.00 to 2.48]; P = 0.05) within
between these two groups. patients with a diagnosis of OSA 30 days (Table 3). No notable dif-
When comparing patients with a without a prescription for a CPAP ference was observed in the incidence
diagnosis of OSA and a prescription compared with control subjects of PE, MI, and respiratory arrest
Table 3
Statistical Comparison
OSA Overall Versus OSA With CPAP OSA With No CPAP OSA With No CPAP
Control Group Versus Control Group Versus Control Group Versus With CPAP
P P
Complication OR [95% CI] P Value OR [95% CI] Value OR [95% CI] P Value OR [95% CI] Value
Emergency 1.25 [1.11-1.41] 0.0003 1.10 [0.92-1.31] 0.307 1.38 [1.19-1.60] ,0.0001 1.26 [1.01-1.56] 0.039
department
visit (30 d)
Hospital 1.23 [1.06-1.43] 0.034 1.18 [0.88-1.58] 0.263 1.24 [1.05-1.47] 0.013 1.01 [0.70-1.44] 0.998
admission
(30 d)
Pulmonary 1.41 [0.91-2.18] 0.124 1.55 [0.87-2.76] 0.139 1.27 [0.72-2.24] 0.405 1.12 [0.53-2.37] 0.767
embolism
(30 d)
Myocardial 1.27 [0.63-2.59] 0.504 1.14 [0.44-2.97] 0.785 1.39 [0.54-3.56] 0.495 1.13 [0.32-4.04] 0.853
infarction
(30 d)
Respiratory 1.32 [0.89-1.94] 0.164 1.03 [0.58-1.81] 0.932 1.57 [1.00-2.48] 0.050 1.53 [0.80-2.94] 0.202
arrest
(30 d)
In-hospital 1.25 [0.46-3.41] 0.660 1.01 [0.29-3.52] 0.987 1.54 [0.35-6.72] 0.564 1.77 [0.21-14.72] 0.596
mortality
(6 mo)
CPAP = continuous positive airway pressure, OR = odds ratio, OSA = obstructive sleep apnea
July 1, 2019, Vol 27, No 13 489
within 30 days between the two based on expert opinion, often the effect of OSA on ambulatory
groups (Table 3). leading to increased scrutiny and orthopaedic surgery is limited and
Finally, patients with OSA and an prolonged postoperative monitoring largely related to hospital admission
order for a CPAP were compared times in the authors’ institution. postoperatively. Liu et al11 reviewed
with those with OSA without an Shoulder arthroscopy is of particular 206 patients who underwent ambu-
order for a CPAP with a regression concern for patients with OSA latory orthopaedic procedures and
analysis. Patients without CPAP pre- because of several procedure-specific were monitored for 24 hours post-
scriptions had a markedly higher factors. These include scalene pe- operatively on continuous pulse
incidence of emergency department ripheral nerve blockade with the oximetry. They reported that 34% of
visits (OR, 1.26 [1.01 to 1.56]; P = possibility of phrenic nerve paralysis, patients had hypoxemia in the
0.039) within 30 days. No other catheter placement, fluid extravasa- postoperative care unit and these
notable differences were observed in tion into the chest wall during surgery, episodes were not associated with
the studied perioperative complica- patient positioning postoperatively, any increase in adverse outcomes or
tions between these two groups and a depressed respiratory drive re- unplanned hospital admission, sug-
(Table 3). sulting from surgical narcotic use. gesting that overnight observation is
Across all regression analyses was However, the present study suggests unnecessary after such procedures.
the consistent finding that although that although patients with OSA have This study is likely subject to type II
OSA was markedly associated with slightly higher risks of emergency error given the low numbers of pa-
the complications of interest, the department visits and hospital ad- tients analyzed. By contrast, the
association of other medical co- missions within 30 days after RCR, present study reviews more than
morbidities with the studied adverse no association is noted between OSA 4,500 patients with OSA undergoing
events was more substantial. Most and increased incidences of other RCR and reports a markedly in-
notable was for postoperative emer- perioperative complications including creased risk of hospital admission
gency department visits, where the PE, MI, and respiratory arrest within and emergency department visitation
effect of alcohol abuse, depression, 30 days, and in-hospital mortality within 30 days in patients with OSA
hypercoagulable state, heart disease, within 6 months postoperatively. undergoing ambulatory RCR com-
and liver disease were all more Additionally, an order for CPAP ap- pared with those without OSA.
markedly associated with the com- pears to mitigate any increased risk However, in the subgroup of pa-
plication than OSA. Similarly, other for complications in patients with tients with a prescription for CPAP
comorbidities were more markedly OSA. Finally, the independent risk machine, each of these risks was
associated with hospital readmission, conferred by OSA for any of the mitigated. These data suggest that
including depression, hypercoagula- complications was markedly less than compared with other patients
ble state, lung disease, heart disease, numerous other demographic or undergoing ambulatory surgery,
and liver disease. medical comorbidity variables, indi- patients diagnosed with OSA
cating that aggressive management of without a CPAP order may be at
other medical comorbidities associ- higher risk of hospital readmission
Discussion ated with OSA may be equally as and emergency department visitation
important in reducing the risk for in the early postoperative period. In
The latest data set from the National complications after arthroscopic addition, the institution of CPAP in
Survey of Ambulatory Surgery RCR. the perioperative period might miti-
reported that more than 39 million Most of the literature evaluating gate these risks completely.
ambulatory surgical procedures are the effect of OSA on perioperative Despite well-established cardio-
performed annually in the United complications after orthopaedic sur- pulmonary risks that OSA poses in
States and arthroscopic rotator cuff gery has focused on inpatient total the perioperative period after inpa-
repair is the second most common joint arthroplasty.8 These studies tient procedures, no study to date
upper extremity procedure.10 Un- have reported higher rates of post- evaluates the relationship between
fortunately, as ambulatory surgery operative PE and hypoxia and OSA and cardiopulmonary compli-
continues to increase in popularity overall higher resource use.12 Al- cations after ambulatory orthopaedic
globally, so does the prevalence of though it can be assumed that such surgery.1 Given this lack of evidence,
sleep apnea.4 Currently, the evidence higher complication rates would also patients undergoing ambulatory
regarding the safety of ambulatory be found in ambulatory orthopaedic orthopaedic surgery with OSA,
shoulder arthroscopy for patients procedures, a few studies substanti- especially those such as RCR with
with OSA is limited and largely ate this. Furthermore, literature on regional anesthesia increasing the
risk of phrenic nerve blockade, might administrative databases.16,17 First, admission, and respiratory arrest
face more scrutiny and monitoring although we used a regression anal- within 30 days. However, a pre-
postoperatively. The present study is ysis in an attempt to control for scription for CPAP in patients with
the first to isolate a cohort of patients several known confounders for the OSA appeared to mitigate all afore-
with OSA undergoing ambulatory studied outcomes after ambulatory mentioned risks in these patients.
RCR and to report no differences in RCR, several other variables are not Additionally, the independent effect
the incidence of cardiopulmonary identifiable within the database imparted by OSA for any of the
perioperative events or in-hospital which are likely to contribute to the studied adverse events or complica-
mortality. With a rise in both num- outcomes studied. Second, the results tions was much lower than numerous
ber of ambulatory RCR and preva- of the present study are dependent other medical comorbidity variables,
lence of OSA, it is important to study on the quality of data and the potentially indicating that aggressive
the effect of OSA on perioperative accuracy with which these proce- optimization of other medical co-
complications to prevent the overuse dures and prescriptions are coded. morbidities associated with OSA may
of resources such as unnecessary Thus, miscoding and noncoding by be equally or more important in
inpatient procedures and prolonged physicians can be sources of error. reducing the risk for complications
postoperative monitoring. Third, CPAP compliance is unable to after arthroscopic RCR.
Among the various management of be quantified using this database.
OSA, CPAP has been reported as the Similarly, we are unable to stratify
References
most effective and widely used.13 It the degree of OSA within this pop-
has been shown to effectively reduce ulation and rely on our large sample References printed in bold type are
nocturnal events and improve daytime size to accurately capture a repre- those published within the past 5
sleepiness, cognitive function, and sentative population of patients with years.
overall well-being in patients with OSA undergoing ambulatory RCR.
1. Raveendran R, Chung F: Perioperative
OSA.14,15 The effect of preoperative Fourth, reasons for readmission after
consideration of obstructive sleep apnea in
CPAP usage on perioperative adverse any ambulatory surgery are complex ambulatory surgery. Anesthesiol Clin 2014;
events after inpatient procedures has and multifactorial. Although we 32:321-328.
been well studied. A recent meta- agree that reporting the reason or 2. Finkel KJ, Searleman AC, Tymkew H, et al:
analysis reviewing six studies and diagnosis of why a patient was re- Prevalence of undiagnosed obstructive sleep
apnea among adult surgical patients in an
more than 900 patients concluded no admitted, such as failure to extubate, academic medical center. Sleep Med 2009;
notable difference in such events respiratory distress, or certain hos- 10:753-758.
between patients with and without pital policies, would add important 3. Stierer TL, Wright C, George A, Thompson
perioperative CPAP therapy.13 By data to our analysis, we are un- RE, Wu CL, Collop N: Risk assessment of
obstructive sleep apnea in a population of
contrast, no current literature eval- fortunately unable to identify these patients undergoing ambulatory surgery. J
uates the effect of perioperative CPAP reasons within the database. Finally, Clin Sleep Med 2010;6:467-472.
therapy and postoperative adverse we are also limited in determining 4. Peppard PE, Young T, Barnet JH, Palta M,
events after ambulatory surgery. In the the type of anesthesia received by Hagen EW: Hla KM: Increased prevalence
of sleep-disordered breathing in adults.
present study, we are able to show a each patient and acknowledge that Am J Epidemiol 2013;177:1006-1014.
mitigation of adverse events, including patients receiving regional blockade
5. Memtsoudis SG, Della Valle AG, Besculides
emergency department visitation, re- for these surgeries may be at higher MC, Gaber L, Laskin R: Trends in
admission, and respiratory arrest, with risk for phrenic nerve palsy, causing demographics, comorbidity profiles, in-
prescription of a CPAP in the peri- an increased risk of the studied hospital complications and mortality
associated with primary knee arthroplasty.
operative period. Although the effect outcomes rather than OSA. J Arthroplasty 2009;24:518-527.
of this therapy is likely multifactorial, 6. Gupta RM, Parvizi J, Hanssen AD, Gay PC:
by controlling for its effect using a Postoperative complications in patients
multivariable logistic regression anal- with obstructive sleep apnea syndrome
Conclusions undergoing hip or knee replacement: A
ysis, we attempted to isolate the case-control study. Mayo Clin Proc 2001;
independent effect of this variable and Similar to the broader literature, 76:897-905.
show at the very least an association approximately 11.2% of patients 7. Kaw R, Pasupuleti V, Walker E,
with a CPAP prescription and risk undergoing ambulatory RCR had a Ramaswamy A, Foldvary-Schafer N:
Postoperative complications in patients
reduction. preoperative diagnosis of OSA. After with obstructive sleep apnea. Chest 2012;
This study has several important controlling for covariates, OSA was 141:436-441.
limitations, many of which are associated with a higher risk of 8. Naqvi SY, Rabiei AH, Maltenfort MG,
inherent to all studies using large, emergency department visitation, re- et al: Perioperative complications in
July 1, 2019, Vol 27, No 13 491
patients with sleep apnea undergoing total orthopedic surgery patients with diagnosis rotator cuff repair: A case report. Case Rep
joint arthroplasty. J Arthroplasty 2017;32: of obstructive sleep apnea: A retrospective Orthop 2013;2013:801752.
2680-2683. observational study. Patient Saf Surg 2010;
4:9. 15. Montserrat JM, Ferrer M, Hernandez L,
9. Opperer M, Cozowicz C, Bugada D, et al: et al: Effectiveness of CPAP treatment
Does obstructive sleep apnea influence 12. D’Apuzzo MR, Browne JA: Obstructive in daytime function in sleep apnea
perioperative outcome? A qualitative sleep apnea as a risk factor for syndrome: A randomized controlled study
systematic review for the society of postoperative complications after with an optimized placebo. Am J Respir
anesthesia and sleep medicine task force on revision joint arthroplasty. J Arthroplasty Crit Care Med 2001;164:608-613.
preoperative preparation of patients with 2012;27:95-98.
sleep-disordered breathing. Anesth Analg 16. Cancienne JM, Gwathmey FW, Miller MD,
2016;122:1321-1334. 13. Nagappa M, Mokhlesi B, Wong J, Wong Werner BC: Tobacco use is associated with
DT, Kaw R, Chung F: The effects of increased complications after anterior
10. Jain NB, Higgins LD, Losina E, Collins J, continuous positive airway pressure on cruciate ligament reconstruction. Am J
Blazar PE, Katz JN: Epidemiology of postoperative outcomes in obstructive Sports Med 2016;44:99-104.
musculoskeletal upper extremity sleep apnea patients undergoing
ambulatory surgery in the United States. surgery. Anesth Analg 2015;120: 17. Cancienne JM, Brockmeier SF, Werner BC:
BMC Musculoskelet Disord 2014;15:4. 1013-1023. Tobacco use is associated with increased
rates of infection and revision surgery after
11. Liu SS, Chisholm MF, John RS, Ngeow J, 14. Yamamoto T, Tamai K, Akutsu M, primary superior labrum anterior and
Ma Y, Memtsoudis SG: Risk of Tomizawa K, Sukegawa T, Nohara Y: posterior repair. J Shoulder Elbow Surg
postoperative hypoxemia in ambulatory Pulmonary embolism after arthroscopic 2016;25:1764-1768.
Review Article
Rationale for Biologic

Augmentation of Rotator
Cuff Repairs
Abstract
Raffy Mirzayan, MD The structural integrity of rotator cuff repair (RCR) has been a primary
Alexander E. Weber, MD focus for shoulder surgeons seeking long-term clinical and functional
success. Improvements in surgical techniques have allowed for
Frank A. Petrigliano, MD
superior initial biomechanical fixation. However, tendon healing
Jorge Chahla, MD, PhD remains a significant clinical problem even after rigid time-zero repair.
The lack of long-term healing has led to increased interest in biologic
augmentation to improve tendon-to-bone healing. This interest has
led to a rise in the investigation of small molecular therapies, cell-
From the Department of Orthopaedic based strategies, and tissue-derived treatments offering surgeons a
Surgery, Southern California
Permanent Medical Group, Baldwin
new therapeutic toolbox for potentially improving RCR long-term
Park, CA (Dr. Mirzayan), the Division outcomes. However, the delivery, efficacy, and safety of these
of Sports Medicine, Department of treatments remain under investigation. Additional well-designed,
Orthopaedic Surgery, Keck School of
Medicine, University of Southern
high-level studies are of paramount importance in creating evidence-
California, Los Angeles, CA based guidelines for the implementation of new biologic solutions.
(Dr. Weber, Dr. Petrigliano), and the This review article discusses the current preclinical, translational, and
Sports Medicine and Shoulder
Division, Department of Orthopaedic
clinical experience with and rationale for biologic augmentation in
Surgery, Rush University Medical RCR.
Center, Chicago, IL (Dr. Chahla).
Dr. Mirzayan or an immediate family
member is a member of a speakers’
bureau or has made paid
presentations on behalf of Arthrex;
has received research or institutional
R otator cuff tears (RCTs) are a
frequent cause of shoulder pain
and disability, with approximately
healing or re-tearing associated with
RCR surgery is noted.4 Rotator cuff
augmentation could be considered in
support from Arthrex and Joint 98 per 100,000 people in the United patients with risk factors for failure
Restoration Foundation; and has
stock or stock options held in
States undergoing rotator cuff re- to heal an RCR such as increasing
Alignmed. Dr. Petrigliano or an pair (RCR) annually, accounting for age (mostly .65 years), multiple
immediate family member is a 1,360 cases per 100,000 patients seen tendon involvement (.1), large tear
member of a speakers’ bureau or has for orthopaedic disease or injury size (.2 cm), retraction (.2 cm),
made paid presentations on behalf of
Zimmer Biomet and Stryker. Neither of
every year.1,2 The success of surgery high-grade fatty infiltration of the
the following authors nor any is often determined by the eradica- muscles (Goutallier .2).5-7
immediate family member has tion of pain and the return of
received anything of value from or has shoulder function. When evaluating
stock or stock options held in a
Rotator Cuff Repair Failure
the RCR literature, whether tradi-
related directly or indirectly to the tional open (deltoid take down), RCR failure should be divided into
subject of this article: Dr. Weber and mini-open (deltoid splitting), or clinical or mechanical failure. Clini-
Dr. Chahla. arthroscopic, a high level of patient cal failure is the subjective reporting
J Am Acad Orthop Surg 2019;27: satisfaction is noted.3 Surgery alle- of the patients that they are still
468-478 viates pain and leads to good sub- having pain and are limited in the use
DOI: 10.5435/JAAOS-D-18-00281 jective outcome scores. However, of their shoulder. Mechanical failure
despite the resolution of pain and occurs when there is a loss of
Academy of Orthopaedic Surgeons. the improvement in subjective out- fixation/repair, failure to heal, or a
comes, a high rate of incomplete re-tear of the tendon. Mechanical
Raffy Mirzayan, MD, et al
failure can only be detected with of daily living and a better range of the structural integrity rates to max-
imaging studies. Paradoxically, the active flexion (129° 6 20° compared imize patients’ strength and function.
structural integrity (or lack thereof) with 71° 6 41°) compared with the
does not necessarily correlate with shoulders that had a large recurrent
the clinical outcome.8 A patient may defect. The integrity of the rotator cuff Timing of Rotator Cuff
be satisfied with the surgery, but the at the time of follow-up, not the size of Repair Failures
rotator cuff tendon may remain the tear at the time of repair, was the
structurally compromised. Galatz major determinant of the outcome of To better understand why RCRs fail,
et al8 raised awareness to this issue RCR. They also found that the quality it is important to examine when they
by obtaining postoperative ultraso- of the rotator cuff tissue and the fail. Several studies have sequentially
nography in patients who had potential for a durable repair deteri- followed RCR with imaging studies
undergone repair of large and mas- orated with the patient’s age. Nho to aid in the understanding of when
sive RCTs and reported that 94% of et al10 retrospectively analyzed pa- RCRs fail. Miller et al13 performed
the repairs had “re-torn.” Since the tients in an arthroscopic rotator cuff arthroscopic double-row repair on
publication of that study, the focus registry who had ultrasonography 22 consecutive patients with large
of many shoulder surgeons has been documentation of cuff integrity and RCTs (.3 cm). The patients then
directed at improving the structural found that shoulders with cuff integ- underwent serial ultrasonography
integrity rates. Unfortunately, the rity at 1 year had significantly higher examinations at 2 days, 2 weeks,
term “re-tear” does not adequately external rotation strength (P , 6 weeks, and 3, 6, 12, and 24 months
capture what may be occurring bio- 0.05). Iannotti et al11 evaluated 113 after surgery. Of the 22 arthro-
logically and a better terminology RCRs with MRI confirmation of scopically repaired RCTs, 9 (41%)
should be used to define structural rotator cuff integrity performed at demonstrated recurrent tears. Seven
failure. Currently, the literature uses 52 weeks after surgery; the mean re-tears of the 9 (78%) occurred
the term “re-tear,” implying that the ratio of the scapular abduction within 3 months of surgery; 2 of
tendon healed to the tuberosity and strength of the affected shoulder to these 7 occurred while the patient
then tore again. It is unknown that of the normal, contralateral was in a sling during the postoper-
whether these are in fact repeat tears shoulder was only 75% in the sub- ative period. The other 2 (22%)
of a previously well-healed repair or jects lacking cuff integrity, whereas occurred between 3 and 6 months.
whether the tendon never fully it was 92% in those with an intact No re-tears occurred after 6 months.
healed after attempted repair. We repair (P = 0.0026). Similar findings were demonstrated
believe a more universal terminology Boileau et al30 performed arthro- in a multicenter, prospective, non-
of “structural integrity rate” may be scopic RCR in 65 patients and later randomized study of a single cohort
better suited to describe the state of performed postoperative CT arthro- of patients by Iannotti et al.11 An
the tendon-bone interface. gram of MRI between 6 months and arthroscopic transosseous equivalent
3 years. They found that patients who RCR was performed in 113 patients
had a healed rotator cuff had a sig- with a range of tear sizes from 1 to
Importance of Rotator Cuff nificantly higher Constant score (85.7 4 cm. Postoperative MRIs were ob-
Integrity versus 78.9; P = 0.02) and higher tained at 2, 6, 12, 16, 26, and
shoulder strength of shoulder eleva- 52 weeks. A re-tear occurred in 19
Studies have shown that when the tion (7.3 versus 4.7 kg; P = 0.001). In cases of the 113 (17%). One re-tear
rotator cuff heals, patients have the largest study correlating the (5%) was identified at 2 weeks, zero
increased rotator cuff strength.9-12 structural integrity of RCR with re-tears at 6 weeks, 7 re-tears (37%)
Harryman et al9 evaluated 105 RCRs functional outcomes, Collin et al12 at 12 weeks, 5 re-tears (26%) at
with an average age of 60 years reviewed the records of 210 shoulders 16 weeks, 5 re-tears (26%) at
(range, 32 to 80 years) at an average that had a postoperative MRI at 10 26 weeks, and 1 re-tear (5%)
of 5 years after surgery. The authors years post-op and found that the total at 52 weeks. The mean time to re-
correlated functional outcomes of Constant score (P , 0.005), espe- tear was 19.2 weeks. This number
patients with the integrity of the cially the strength component (P , may be skewed as there were less
rotator cuff assessed with ultraso- 0.001), was significantly correlated to frequent imaging intervals further
nography. They found that the repair integrity. Based on these stud- out from surgery. Approximately
shoulders in which the repaired cuff ies, the goal of shoulder surgeons 42% of re-tears occurred in the first
was intact at the time of follow-up ought to be not only improving pa- 3 months and 68% occurred in
had a better function during activities tient’s symptoms but also improving the first 4 months. The authors
July 1, 2019, Vol 27, No 13 469
Biologic Augmentation of Rotator Cuff Repair
concluded that “rotator cuff healing revision surgery and was found to be improve load to failure, but we know
is prolonged and there is an oppor- tendon pulling through sutures in 19 based on the previously cited studies
tunity to speed healing by protecting revision cases (86%). Two (9%) that the repair is still inadequate due
the repair from excessive loading.”11 were re-tears adjacent to the repair to weakness at the tendon-suture
Hernigou et al14 performed an site and 1 (5%) was anchor pullout. interface. Consequently, strategies
arthroscopic single-row repair on 45 In biomechanical testing of various to improve RCR integrity include
patients with tears less than 3 cm. RCR constructs, the most common attempts to reinforce the native
The patients were then followed failure is at the suture-tendon inter- tendon by mechanical or biologic
with monthly ultrasonography for face.17,18 The reason for suture means. These approaches may
the first 24 months. They found that pullout is the poor tissue quality of involve augmenting the native ten-
8 (17.8%) had a re-tear between 2 the rotator cuff tendon. Codman don or enhancing the biologic
and 3 months postoperative and an suggested that degenerative changes healing milieu. Biologic augmenta-
additional 8 (17.8%) had a re-tear occur in the rotator cuff tendons as tion can play a substantial role in
between 3 and 6 months, with an we age, and the diminished biome- strengthening this interface.
average time to re-tear at 3.4 months. chanical properties of the tendon
What can be surmised from these may impede its ability to retain the
studies is that the failures happen rel- suture. A host of histopathologic Dermal Allografts
atively early after repair, that rotator changes has been shown to occur in
cuff healing is delayed, and can take 6 ruptured tendons such as collagen Historical reports of freeze-dried
to 12 months to occur, if at all. So as degeneration, disordered arrange- allograft tendons have shown incon-
surgeons, what we need to accomplish ment of collagen fibers, and greater sistent outcomes with some cata-
is to either strengthen the repair so the quantities of type III collagen pro- strophic graft failures and foreign
tendon stays attached to the tuberosity duced by tenocytes from ruptured body reactions.20 In response, off-the-
longer or speed up healing of the tendons with a decrease in type I shelf porcine intestinal submucosal
tendon before failure occurs. collagen.19 In addition, a decrease in membrane “patches” rose in popu-
fibroblast population and the num- larity, which unfortunately resulted in
ber of blood vessels occurs as the size some intense local inflammatory re-
Cause of Rotator Cuff of the rotator cuff tear increases.19 actions and early graft failures.21,22
Repair Failure A commercially available bio-
inductive bovine collagen implant
Now that it is better understood Mechanical Augmentation has also been used in RCR augmen-
when repairs fail, it is possible to to Strengthen Rotator Cuff tation.23 This implant is intended as
discern why they fail. Early failures Repair an onlay over the RCR to add col-
occur due to inadequate mechanical lagen and thicken the cuff and does
repair, so it is imperative to analyze Surgeon-controlled variables are not provide any mechanical support
where the weak link is in the repair available which can strengthen the or advantage. To date, the only
construct. The commercially avail- repair (increase the load to failure) at clinical report is for the repair of
able high-strength sutures and an- time zero such as anchor type, suture partial-thickness tears.
chors currently have superior material, knot type, stitch configura- Dermal allografts, otherwise
biomechanical characteristics and tion, size and shape of a tissue- known as acellular dermal matrices
have been shown not to be the mode penetrating instrument, and size of (ADMs), have been the subject of
of failure. The “weak link” is the tissue bite. Anchors and high- extensive clinical and preclinical
believed to be the suture-tendon strength sutures, regardless of manu- evaluation and can significantly
interface.13,15 facturer, have been maximized in their increase the ultimate load to failure.
This finding has been demonstrated load to failure capacity. Ponce et al15 These grafts are processed to remove
in several studies. Cummins et al evaluated stitch configuration, size donor cells, leaving behind the
prospectively followed 342 consecu- and shape of a tissue-penetrating extracellular matrix, which is mostly
tive RCRs performed by a single instrument, and size of the tissue composed of type I collagen. There
surgeon.16 Twenty-one (6%) subse- bite in a laboratory setting and found are several commercially available
quently required a revision RCR, that stitch configuration had the most ADMs with different methods of pro-
with one patient undergoing two significant increase in load to failure. cessing and sterilization, as well as
revision repairs. The mode of failure Other than stitch configuration, handling characteristics.24 Given their
was documented at the time of there are no other ways a surgeon can biomechanically proven superior
suture pullout strength,25,26 ADMs with MRI. No difference was noted in

function to strengthening the repair
Enhancing the Biologic the structural integrity rates between
while allowing an optimized environ- Healing of Rotator Cuff the control group and vented group
ment for host cells and growth factors Repair for medium-sized tears (81.2% versus
to promote repair site healing. 86.4%, respectively; P = 0.62).
Extensive research and investigations
Multiple biomechanical studies have However, a significant improvement
are underway to improve the second
evaluated ADMs in rotator cuff repair occurred in the structural integrity
component of RCR: tendon-to-bone
model.17,26,27 Barber et al17 demon- rates for large to massive tears
healing.
strated in a cadaver RCT model an between the control group and vented
increase in mean failure strength in group (71.4% versus 95.5% respec-
augmented repairs with ADMs (325 Marrow Venting tively; P = 0.025). Osti et al36
Newtons) compared with cadaveric Procedures (Microfracture) conducted a prospective, randomized
controls (273 Newtons) (P = 0.047). trial with 29 controls and 28 vented
Beitzel et al27 evaluated ADM aug- Microfracture of the greater tuber- during arthroscopic RCR. Although a
mentation in a cadaver RCR model osity was popularized by Snyder and significant difference existed in func-
and found a statistically significant Burns32 who coined the term tional outcome and pain scores for the
increase in load to failure in ADM- “crimson duvet” for the bed of vented group at 3 months, this dif-
augmented repairs versus non- bloody fluid blanketing the greater ference did not carry through at the
augmented controls (575.8 N versus tuberosity after microfracturing al- final follow-up (minimum 2 years).
348.9 N; P = 0.025). lows the bone marrow elements Furthermore, postoperative MRI
These biomechanical findings have (mesenchymal stem cell [MSC], pla- performed at the final follow-up
been supported clinically. A level 2, telets, growth factors) to surface. showed no difference in the struc-
prospective, randomized controlled Dierckman et al33 retrospec- tural integrity rates between the
study by Barber et al18 evaluated 42 tively reviewed 52 patients (53 groups (89.7% versus 92.9%,
patients with .3 cm, two-tendon shoulders) who underwent a single- respectively; P = 0.67). One of the
RCTs repaired arthroscopically. row RCR of tears between 2 and 4 cm limitations of this study was that
Twenty-two patients were random- with marrow stimulation. At the authors do not report the size of
ized to single-row arthroscopic a minimum follow-up of 24 months, the RCT, and as elucidated by the
repair and 20 patients to single-row MRI revealed rotator cuff healing in previous study, there may be a dif-
arthroscopic repair augmented by 48 of 53 shoulders (91%). The limi- ference in healing rates in larger tears.
ADMs by an onlay technique as tation of this study was the absence Based on these studies, micro-
described by Labbe.28 At an average of a control group. Milano et al,34 in a fracture (venting) at the RCR site may
follow-up of 24 months, 85% of the prospective, randomized study, com- be beneficial in improving healing
augmented repairs were intact on pared 40 control subjects with 40 rates or large and massive tears but
MRI at follow-up, compared with patients who had a microfracture in do not play a role for small- and
40% in the control group (P , 0.05). conjunction with RCR. Overall, the medium-sized tears.
Agrawal retrospectively reviewed 14 structural integrity rate was 52.6% in
patients with either RCTs greater the control group and 65.7% in the
than 3 cm or recurrent RCTs (may venting group, without a significant Platelet-Rich Plasma
be less than 3 cm) that were arthro- difference between the groups (P =
scopically repaired with a double-row 0.256). However, a subgroup analysis Platelet-rich plasma (PRP) is an
technique with ADM augmentation. by tendon size showed that for large autologous concentrate of a patient’s
Postoperative MRI obtained at an tears, the microfracture group had a own blood that is injected into or
average of 16.8 months revealed higher structural integrity rate com- onto the injured soft tissue in an
85.7% of repairs to be intact, with pared with the control group (60% (6 effort to promote a healing response.
14.3% having recurrent tears of less of 10) versus 12% (1 of 8), respec- The release of growth factors from
than 1 cm.29 These clinical studies tively; P = 0.04). Although this sub- platelet alpha-granules enhances cell
demonstrate that RCRs augmented group analysis may be underpowered, proliferation of tenocytes and pro-
with ADMs appear to have a much the findings have been substantiated motes the synthesis of extracellular
higher rate of structural integrity on by Taniguchi et al35 who followed 44 matrix cell proliferation, chemotaxis,
postoperative imaging compared with control subjects and 67 vented cell differentiation, and angiogenesis
what has been previously reported in arthroscopic RCR by a single surgeon to enhance the healing process.37 The
the literature.11,30,31 with a minimum 12-month follow-up literature has been inconsistent on the
July 1, 2019, Vol 27, No 13 471
outcomes of PRP use in RCR. Sev- trials including 1,147 patients ana- and preparation method. BMC has
eral studies have demonstrated that lyzed separately the effect of PRP been used for decades in Europe
although PRP had no effect on the and platelet-rich fibrin (PRF) on with a long-term follow-up study
clinical outcomes after RCR, it did RCR. PRF is clotted after it is col- detailed later which shows significant
have an impact on the structural lected, is immediately activated, and benefit in RCR surgery, and until
integrity, suggesting that PRP pro- is sutured at the bone-tendon inter- only recently, there have been no
motes tendon healing to bone. face. The authors found no differ- studies on the use of adipose-derived
Other studies have shown that it did ence in healing rates between PRF stem cells in RCR surgery.14,42
not have an effect on the structural and controls, as well as patient satis-
integrity rate.38-40 Vavken et al38 faction, Constant score, and American
in a meta-analysis of 13 published Shoulder and Elbow Surgeons score. Bone Marrow Concentrate
reports between 2010 and 2014 However, they found that PRP
found that for small- and medium- improved the structural integrity rates Hernigou et al14 performed an
sized tears (,3 cm), the risk ratio compared with the controls overall arthroscopic single-row repair on 45
for re-tear was 0.60, consistent (82.8% versus 69.5%; P , 0.05), matched pair patients with tears less
with a significant difference in favor small to medium tears (77.6% versus than 3 cm. The treatment group
of PRP use (P = 0.038). Warth et al39 61.75; P , 0.05), and medium to received BMC at the time of the
in a meta-analysis and meta-regression large tears (93.3% versus 73.5%; P , repair, whereas the control group
analysis of eight level I and II studies 0.05). In addition, PRP use leads to did not (nonrandomized). The pa-
found that if the RCT size was greater significantly improved visual analog tients were matched for age, sex,
than 3 cm in AP length, the PRP- scale at 30 days and at the final dominance, and tear size. The pa-
treated group exhibited decreased re- follow-up, as well as improved Con- tients were then followed clinically
tear rates after double-row repairs stant and UCLA scores but no differ- and with imaging studies. Monthly
(25.9% versus 57.1%, respectively; ence in American Shoulder and Elbow ultrasonography was performed
P = 0.046). Cai et al40 in a meta- Surgeons scores. More importantly, for the first 24 months. MRIs were
analysis of five level I studies found the authors were able to differentiate performed at 3, 6, 12, and
that although no statistically signifi- between leukocyte-rich and leukocyte- 24 months, and 10 years postoper-
cant differences existed between poor PRP formulations and found that atively. The authors found that bone
groups in the overall outcome scores the leukocyte-poor formulations had marrow-derived MSC injection dur-
(P . 0.05), there were better healing significant improvement in healing ing RCR enhanced the healing rate
rates in patients treated with PRP rates compared with controls (83% and improved the quality of the re-
(P , 0.03) who had small/moderate versus 69.1%, respectively; P , 0.05), paired surface. All 45 pairs (100%)
full-thickness tears. whereas the leukocyte-rich for- with MSC augmentation had healed
One of the main reasons for the mulations did not (69.5% versus by 6 months, compared with just 30
inconsistency in the literature is that 59.3%; P = 0.36). A summary of all of 45 non-MSC repairs (67%). BMC
not all PRP is the same with regard to level I studies with PRP used in rota- injection was also protective against
the concentration, content, prepara- tor cuff repair is shown in Table 1. rupture through the most recent
tion method, and delivery technique. follow-up (10 years postoperatively).
There are various commercially At this time point, 87% of patients in
available systems that produce dif- Mesenchymal Stem Cells the MSC-treated group had intact
ferent concentrations of PRP com- cuffs, but just 44% were intact in the
pared with normal levels.41 In the Currently, two autologous sources of unaugmented group. A greater
same patient, there can be a high stem cells are available which can be number of transplanted MSCs also
degree of variability in platelet con- commercially used: adipose-derived appeared to propagate tissue integ-
centration throughout the day. stem cells and bone marrow concen- rity as those with cuff tears at any
Other factors such as activation of trate (BMC).56 Both can be har- time during the follow-up had
platelets to create a gel or clot, as vested, prepared, and then reinjected received fewer MSCs compared with
well as the number of white blood in an in-office setting with minimal those who maintained a successful
cells, can have an effect on the donor site morbidity. repair. Another significant finding in
healing potential and inconsistency As for PRP, a high variability exists this study was that no re-tears were
in the findings. in the concentration on progenitor present in patients who received
A recent meta-analysis by Hurley cells that is dependent on the age of BMC MSC in the first 6 months after
et al42 of 18 randomized controlled the donor, location of the harvest, surgery. It is important to highlight
Table 1
Level 1 Studies With a Minimum of 12-Month Follow-up
Differences in
Type of Study Treatment Control Re-tear/Healing Second Look or Functional
PRP (Year) Group (n) Group (n) Follow-Up Rates Imaging Scores Conclusion
PRP
Leukocyte Ebert et al43 27 28 42 No MRI—no CS, OSS, ASES, qDASH, Significant postoperative
poor (2017) differences. GRC. The only clinical improvements
difference was that the and high levels of patient
PRP group had better satisfaction were
strength CS subscale. observed in patients at
midterm after
supraspinatus repair.
Although pain free,
maximal abduction
strength was greater in
the midterm after PRP
treatment; repeated
applications of PRP
delivered at 7 and
14 days after surgery
provided no additional
benefit to tendon
integrity.
Flury et al44 60 60 24 No US and MRI—PRP CS, OSS, pASES, A single intraoperative
(2016) re-tear rate qDASH, EQ5D. Pain injection of pure PRP on
lower, 12.2% vs level—no differences. the reconstructed
20.8%. footprint of the
supraspinatus tendon
showed no significant
effect on the clinical and
patient-reported
outcomes up to
24 months after
arthroscopic rotator cuff
repair compared with an
intraoperative injection
of ropivacaine within the
subacromial space.
However, a similar time-
limited, pain-reducing
effect was noted
between the two
treatments. It remains
unclear whether an
improvement in patient
outcomes, notably in
nonsmoking patients,
can be achieved with
locally administered
growth factors in the
form of pure PRP.
Pandey 52 50 12 Yes Doppler US— CS, UCLA, ASES, VAS— Superior structural healing
et al45 vascularity in the VAS sign. Better until of arthroscopic repair of
(2016) PRP group 6 months; UCLA better the large rotator cuff tear
repair site at at 6 and 12 months and with a single-row
3 months (P , CS at 12 and technique when treated
0.05) and in 24 months for PRP. by moderately
peribursal tissue concentrated PRP
until 12 months. compared with controls.
PRP was also seen to
accelerate the
vascularity of the rotator
cuff and the surrounding
tissues in the early
phase. PRP is beneficial
in reducing the re-tear of
large tears.
(continued )
ASES, American Shoulder and Elbow Surgeons; CS, Constant score; GRC, Global Rating of Change; MRI, magnetic resonance imaging; OSS,
Oxford Shoulder Score; pASES, patient American Shoulder and Elbow Surgeons; PRP, platelet-rich plasma; quick DASH, EQ5D, EuroQol 5
Dimensions; SER, strength in external rotation; SPADI, Shoulder Pain and Disability Index; SST, Simple Shoulder Test; SSV, subjective shoulder
value; UCLA, University of California–Los Angeles, VAS = visual analog scale.
July 1, 2019, Vol 27, No 13 473
Table 1 (continued )
Differences in
Jo et al46 37 37 12 Yes MRI—PRP re-tear CS, UCLA, ASES, VAS, Compared with repairs
(2015) rate lower 3% vs SPADI, SST— without PRP
20%. Better comparable, except augmentation, the
cross-sectional VAS worst pain. current PRP preparation
area in the PRP and application methods
group. for medium to large
rotator cuff repairs
significantly improved
the quality, as evidenced
by a decreased re-tear
rate and increased CSA
of the supraspinatus but
not the speed of healing.
Malavolta 27 27 24 No MRI—no CS, UCLA, ASES, VAS— PRP prepared by
et al47 differences. the only significant apheresis and applied in
(2014) difference was UCLA the liquid state with
at 12 months. thrombin did not
promote better clinical
results at 24-month
follow-up. Given the
numbers available for
the analysis, the re-tear
rate also did not change.
Ruiz-Moneo 32 31 12 No MRI—no UCLA, patient No differences in rotator
et al48 differences. satisfaction—no cuff healing or
(2013) differences. improvements in
function were observed
in the 1-year
postsurgical clinical and
radiologic follow-up
assessments.
Jo et al49 24 24 12 Yes MRI or CT—PRP CS, UCLA, ASES, VAS, The application of PRP for
(2013) re-tear rate SPADI, SST— large to massive rotator
lower, 20% vs comparable, except for cuff repairs significantly
55.6%. Better overall function (P = improved structural
cross-sectional 0.043). outcomes, as evidenced
area in the PRP by a decreased re-tear
group. rate and increased CSA
of the supraspinatus
compared with repairs
without PRP
augmentation. Although
no significant difference
existed in clinical
outcomes, except the
overall shoulder function
after 1-year follow-up,
better structural
outcomes in the PRP
group might suggest
improved clinical
outcomes at longer-term
follow-up.
Leukocyte Zhang 30 30 12 Yes MRI—PRP re-tear CS, UCLA, DASH VAS, The local injection of PRP
rich et al50 rate lower, 13% ROM—no differences. into a primary
(2016) vs 30%. arthroscopic double-row
cuff repair resulted in
lower recurrence rates
than repairs without the
novel biologic
augmentation material.
(continued )
Table 1 (continued )
Differences in
Randelli 26 27 24 No MRI—no CS, UCLA, VAS, SST The results of our study
et al51 differences. SER—better at showed autologous
(2011) 3 months for PRP. PRP reduced pain in the
Comparable at 6, 12, first
and 24 months. postoperative months.
The long-term results of
subgroups of grade 1
and 2 tears suggest that
PRP positively affected
cuff rotator healing.
Platelet-rich
fibrin
Leukocyte Weber 30 30 12 No MRI—no UCLA, VAS, SST, Platelet-rich fibrin matrix
poor et al52 differences. ROM—no differences. was not shown to
(2013) significantly improve
perioperative morbidity,
clinical outcomes, or
structural integrity.
Although longer-term
follow-up or different
PRP formulations may
show differences, early
follow-up does not show
significant improvement
in perioperative
morbidity, structural
integrity, or clinical
outcome.
Castricini 43 45 20.2 No MRI—no CS—no differences. This study does not
et al53 differences. support the use of
(2011) autologous PRFM for
augmentation of a
double-row repair of a
small or medium rotator
cuff tear to improve the
healing of the rotator
cuff. Given the
heterogeneity of PRFM
preparation products
available in the market, it
is possible that other
preparations may be
more effective.
Zumstein 17 18 12 No MRI—no SSV, CS, VAS, SST, Arthroscopic rotator cuff
et al54 differences. ROM—no differences. repair with the
(2016) application of L-PRF
yielded no beneficial
effect on clinical
outcome, anatomic
healing rate, mean
postoperative defect
size, and tendon quality
at 12 months of follow-
up.
Leukocyte Gumina 39 37 13 Yes MRI—no re-tears CS—no differences. The use of the platelet-
rich et al55 in the PRP group leukocyte membrane in
(2012) vs 8%. the treatment of rotator
cuff tears improved
repair integrity
compared with repair
without a membrane.
However, the
improvement in repair
integrity was not
associated with greater
improvement in the
functional outcome.
July 1, 2019, Vol 27, No 13 475
that the authors examined the cells compared with 28.5% (10 of the 35) occur early at the suture-tendon
only in a quantitative fashion and in the conventional group (P , interface or later due to delayed
that the standard criteria from the 0.001). Of note, all available data on healing of the rotator cuff. The
International Society for Cellular ADSC have been reported on this suture-tendon interface can be
Therapy were not tested to designate a single study, and thus, the results augmented with dermal allografts
cell population as “MSCs”: (1) should be considered experimental. (ADMs), and healing of the tendon
culture-expanded cells that adhere to Additional randomized, prospective can be expedited with biologics.
tissue culture plastic; (2) cells that studies are needed to determine the Biologic augmentation has the
retain the capability for tri-lineage efficacy of this treatment. potential to improve tendon healing
differentiation (bone, cartilage, and The aforementioned studies of after RCR, but additional high-
adipose); (3) cells expressing CD105, biologic therapies for RCR have level studies are needed to trans-
CD73, and CD90 (with 95% preva- focused on improving healing at the late preclinical findings into clinical
lence); and (4) cells lacking expression bone-tendon interface. Conversely, applications.
of CD45, CD34, CD14 or CD11b, the treatment of muscle atrophy
CD79 alpha or CD19, and HLA-DR and fibroadipogenic degeneration
References
surface molecules.57 observed in the setting of large tears
Ellera Gomez et al58 reported sim- has largely been ignored. This is References printed in bold type are
ilar results to those of Hernigou despite the fact that both atrophy and those published within the past 5
et al in a small case series of 14 pa- degeneration have been correlated years.
tients treated with BMC injection with poor functional outcome and
1. Zhang AL, Montgomery SR, Ngo SS,
during mini-open RCR. Although decreased strength after RCR60 and
Hame SL, Wang JC, Gamradt SC: Analysis
this study lacked a control group for have been shown to be largely irre- of rotator cuff repair trends in a large
adequate comparison, 13 of the 14 versible even in the setting of com- private insurance population. Arthroscopy
2013;29:623-629.
patients had substantial clinical plete tendon-to-bone healing.61 This
improvement at 1-year follow-up has prompted an investigation into 2. Colvin AC, Egorova N, Harrison AK,
Moskowitz A, Flatow EL: National trends
and all tendons were confirmed the use of both bone marrow and in rotator cuff repair. J Bone Joint Surg Am
intact by MRI. In summary, there are adipose-derived stem cells to regen- 2012;94:227-233.
only two studies with limited scien- erate injured rotator cuff mus- 3. Ji X, Bi C, Wang F, Wang Q: Arthroscopic
tific rigor which does not allow for a cle.62,63 In a recent study by versus mini-open rotator cuff repair: An up-
to-date meta-analysis of randomized
strong conclusion in regard to BMC Eliasberg et al,62 a local injection of controlled trials. Arthroscopy 2015;31:
use in rotator cuff surgery. adipose-derived human perivascular 118-124.
stem cells was shown to diminish 4. McElvany MD, McGoldrick E, Gee AO,
muscle atrophy and fibroadipogenic Neradilek MB, Matsen FA III: Rotator cuff
Adipose-Derived Stem repair: Published evidence on factors
degeneration in a small animal associated with repair integrity and clinical
Cells model of massive RCT. The authors outcome. Am J Sports Med 2015;43:
suggest that these cells act as both 491-500.
Kim et al59 published the first clinical regenerative precursors and a source 5. Wylie JD, Baran S, Granger EK, Tashjian
results of adipose-derived stem cells of paracrine growth factors to stim- RZ: A comprehensive evaluation of
factors affecting healing, range of motion,
as an augment to rotator cuff surgery ulate local satellite cell activity.62 strength, and patient-reported outcomes
in a case-control series of 35 patients Further large animal preclinical and after arthroscopic rotator cuff repair.
undergoing arthroscopic repair. Pa- Orthop J Sports Med 2018;6:
human clinical studies are necessary 2325967117750104.
tients matched for age, sex, and tear to determine the effectiveness of
6. Park JS, Park HJ, Kim SH, Oh JH:
size were followed for a minimum of myo-regeneration on improving Prognostic factors affecting rotator cuff
12 months postoperatively. The au- both rotator cuff healing and ulti- healing after arthroscopic repair in small to
thors found no difference at the final medium-sized tears. Am J Sports Med 2015;
mate functional outcome. 43:2386-2392.
follow-up in regard to the range
of motion, pain, or functional out- 7. Mall NA, Tanaka MJ, Choi LS, Paletta GA
Jr: Factors affecting rotator cuff healing. J
comes measures. However, the Summary Bone Joint Surg Am 2014;96:778-788.
structural integrity of the tendon as
8. Galatz LM, Ball CM, Teefey SA, Middleton
indicated by MRI proved to be sig- Although patients do well clinically WD, Yamaguchi K: The outcome and
nificantly better in the adipose- after RCR surgery, mechanical repair integrity of completely
arthroscopically repaired large and massive
derived stem cell–treated group, failure of the repair can be as high as rotator cuff tears. J Bone Joint Surg Am
with only 14.3% tears (5 of the 35) 94% in massive tears. Failures can 2004;86-A:219-224.
9. Harryman DT II, Mack LA, Wang KY, 22. Iannotti JP, Codsi MJ, Kwon YW, Derwin prospective randomized study. Arthroscopy
Jackins SE, Richardson ML, Matsen FA III: K, Ciccone J, Brems JJ: Porcine small 2013;29:802-810.
Repairs of the rotator cuff. Correlation of intestine submucosa augmentation of
functional results with integrity of the cuff. J surgical repair of chronic two-tendon 35. Taniguchi N, Suenaga N, Oizumi N, et al:
Bone Joint Surg Am 1991;73:982-989. rotator cuff tears. A randomized, controlled Bone marrow stimulation at the footprint
trial. J Bone Joint Surg Am 2006;88: of arthroscopic surface-holding repair
10. Nho SJ, Adler RS, Tomlinson DP, et al: 1238-1244. advances cuff repair integrity. J Shoulder
Arthroscopic rotator cuff repair: Elbow Surg 2015;24:860-866.
Prospective evaluation with sequential 23. Schlegel TF, Abrams JS, Bushnell BD,
ultrasonography. Am J Sports Med 2009; 36. Osti L, Del Buono A, Maffulli N:
Brock JL, Ho CP: Radiologic and clinical
37:1938-1945. Microfractures at the rotator cuff footprint:
evaluation of a bioabsorbable collagen
A randomised controlled study. Int Orthop
implant to treat partial-thickness tears: A
11. Iannotti JP, Deutsch A, Green A, et al: Time 2013;37:2165-2171.
prospective multicenter study. J Shoulder
to failure after rotator cuff repair: A Elbow Surg 2018;27:242-251.
prospective imaging study. J Bone Joint 37. Sundman EA, Cole BJ, Karas V, et al: The
Surg Am 2013;95:965-971. anti-inflammatory and matrix restorative
24. Acevedo DC, Shore B, Mirzayan R:
mechanisms of platelet-rich plasma in
Orthopedic applications of acellular human
12. Collin P, Kempf JF, Mole D, et al: Ten-year osteoarthritis. Am J Sports Med 2014;42:
dermal allograft for shoulder and elbow
multicenter clinical and MRI evaluation of 35-41.
surgery. Orthop Clin North Am 2015;46:
isolated supraspinatus repairs. J Bone Joint
377-388, x. 38. Vavken P, Sadoghi P, Palmer M, et al:
Surg Am 2017;99:1355-1364.
Platelet-rich plasma reduces retear rates
25. Adams JE, Zobitz ME, Reach JS Jr, An KN, after arthroscopic repair of small- and
13. Miller BS, Downie BK, Kohen RB, et al:
Steinmann SP: Rotator cuff repair using an medium-sized rotator cuff tears but is not
When do rotator cuff repairs fail? Serial
acellular dermal matrix graft: An in vivo cost-effective. Am J Sports Med 2015;43:
ultrasound examination after arthroscopic
study in a canine model. Arthroscopy 2006; 3071-3076.
repair of large and massive rotator cuff
22:700-709.
tears. Am J Sports Med 2011;39:
39. Warth RJ, Dornan GJ, James EW, Horan
2064-2070. 26. Barber FA, Herbert MA, Coons DA: MP, Millett PJ: Clinical and structural
14. Hernigou P, Flouzat Lachaniette CH, Tendon augmentation grafts: outcomes after arthroscopic repair of full-
Delambre J, et al: Biologic augmentation of Biomechanical failure loads and failure thickness rotator cuff tears with and without
rotator cuff repair with mesenchymal stem patterns. Arthroscopy 2006;22:534-538. platelet-rich product supplementation: A
cells during arthroscopy improves healing meta-analysis and meta-regression.
27. Beitzel K, Chowaniec DM, McCarthy MB,
and prevents further tears: A case- Arthroscopy 2015;31:306-320.
et al: Stability of double-row rotator cuff
controlled study. Int Orthop 2014;38:
repair is not adversely affected by scaffold 40. Cai YZ, Zhang C, Lin XJ: Efficacy of
1811-1818.
interposition between tendon and bone. platelet-rich plasma in arthroscopic repair
15. Ponce BA, Hosemann CD, Raghava P, Tate Am J Sports Med 2012;40:1148-1154. of full-thickness rotator cuff tears: A meta-
JP, Sheppard ED, Eberhardt AW: A analysis. J Shoulder Elbow Surg 2015;24:
28. Labbe MR: Arthroscopic technique for 1852-1859.
biomechanical analysis of controllable
patch augmentation of rotator cuff repairs.
intraoperative variables affecting the
Arthroscopy 2006;22:1136.e1131-1136. 41. Chahla J, Cinque ME, Piuzzi NS, et al: A
strength of rotator cuff repairs at the suture-
call for standardization in platelet-rich
tendon interface. Am J Sports Med 2013; 29. Agrawal V: Healing rates for challenging plasma preparation protocols and
41:2256-2261. rotator cuff tears utilizing an acellular composition reporting: A systematic review
16. Cummins CA, Murrell GA: Mode of failure human dermal reinforcement graft. Int J of the clinical orthopaedic literature. J Bone
for rotator cuff repair with suture anchors Shoulder Surg 2012;6:36-44. Joint Surg Am 2017;99:1769-1779.
identified at revision surgery. J Shoulder 30. Boileau P, Brassart N, Watkinson DJ, 42. Hurley ET, Lim Fat D, Moran CJ, Mullett H:
Elbow Surg 2003;12:128-133. Carles M, Hatzidakis AM, Krishnan SG: The efficacy of platelet-rich plasma and
17. Barber FA, Herbert MA, Boothby MH: Arthroscopic repair of full-thickness tears platelet-rich fibrin in arthroscopic rotator cuff
Ultimate tensile failure loads of a human of the supraspinatus: Does the tendon really repair: A meta-analysis of randomized
dermal allograft rotator cuff augmentation. heal? J Bone Joint Surg Am 2005;87: controlled trials. Am J Sports Med 2019;47:
Arthroscopy 2008;24:20-24. 1229-1240. 753-761.
18. Barber FA, Burns JP, Deutsch A, Labbe 31. Derwin KA, Badylak SF, Steinmann SP, 43. Ebert JR, Wang A, Smith A, et al: A
MR, Litchfield RB: A prospective, Iannotti JP: Extracellular matrix scaffold midterm evaluation of postoperative
randomized evaluation of acellular human devices for rotator cuff repair. J Shoulder platelet-rich plasma injections on
dermal matrix augmentation for Elbow Surg 2010;19:467-476. arthroscopic supraspinatus repair: A
arthroscopic rotator cuff repair. randomized controlled trial. Am J Sports
32. Snyder SJ, Burns J: Rotator cuff healing and Med 2017;45:2965-2974.
Arthroscopy 2012;28:8-15.
the bone marrow “crimson duvet” from
19. Longo UG, Berton A, Khan WS, Maffulli clinical observations to science. Tech 44. Flury M, Rickenbacher D, Schwyzer HK,
N, Denaro V: Histopathology of rotator Should Elbow Surg 2009;10:130-137. et al: Does pure platelet-rich plasma affect
cuff tears. Sports Med Arthrosc Rev 2011; postoperative clinical outcomes after
19:227-236. 33. Dierckman BD, Ni JJ, Karzel RP, Getelman arthroscopic rotator cuff repair? A
MH: Excellent healing rates and patient randomized controlled trial. Am J Sports
20. Moore DR, Cain EL, Schwartz ML, Clancy satisfaction after arthroscopic repair of Med 2016;44:2136-2146.
WG Jr: Allograft reconstruction for medium to large rotator cuff tears with a
massive, irreparable rotator cuff tears. Am J single-row technique augmented with bone 45. Pandey V, Bandi A, Madi S, et al: Does
Sports Med 2006;34:392-396. marrow vents. Knee Surg Sports Traumatol application of moderately concentrated
Arthrosc 2018;26:136-145. platelet-rich plasma improve clinical and
21. Walton JR, Bowman NK, Khatib Y, structural outcome after arthroscopic
Linklater J, Murrell GA: Restore 34. Milano G, Saccomanno MF, Careri S, repair of medium-sized to large rotator
orthobiologic implant: Not recommended Taccardo G, De Vitis R, Fabbriciani C: cuff tear? A randomized controlled trial. J
for augmentation of rotator cuff repairs. J Efficacy of marrow-stimulating technique Shoulder Elbow Surg 2016;25:
Bone Joint Surg Am 2007;89:786-791. in arthroscopic rotator cuff repair: A 1312-1322.
July 1, 2019, Vol 27, No 13 477
46. Jo CH, Shin JS, Shin WH, Lee SY, Yoon KS, the rotator cuff: A prospective, randomized, 58. Ellera Gomes JL, da Silva RC, Silla LM,
Shin S: Platelet-rich plasma for arthroscopic double-blinded study. Am J Sports Med Abreu MR, Pellanda R: Conventional
repair of medium to large rotator cuff tears: 2013;41:263-270. rotator cuff repair complemented by the aid
A randomized controlled trial. Am J Sports of mononuclear autologous stem cells. Knee
Med 2015;43:2102-2110. 53. Castricini R, Longo UG, De Benedetto M, Surg Sports Traumatol Arthrosc 2012;20:
et al: Platelet-rich plasma augmentation for 373-377.
47. Malavolta EA, Gracitelli ME, Ferreira Neto arthroscopic rotator cuff repair: A
AA, Assuncao JH, Bordalo-Rodrigues M, de randomized controlled trial. Am J Sports 59. Kim YS, Sung CH, Chung SH, Kwak SJ,
Camargo OP: Platelet-rich plasma in rotator Med 2011;39:258-265. Koh YG: Does an injection of adipose-
cuff repair: A prospective randomized study. derived mesenchymal stem cells loaded in
Am J Sports Med 2014;42:2446-2454. 54. Zumstein MA, Rumian A, Thelu CE, et al: fibrin glue influence rotator cuff repair
SECEC research grant 2008 II: Use of outcomes? A clinical and magnetic
48. Ruiz-Moneo P, Molano-Munoz J, Prieto E, platelet- and leucocyte-rich fibrin (L-PRF) resonance imaging study. Am J Sports Med
Algorta J: Plasma rich in growth factors in does not affect late rotator cuff tendon 2017;45:2010-2018.
arthroscopic rotator cuff repair: A healing: A prospective randomized
randomized, double-blind, controlled controlled study. J Shoulder Elbow Surg 60. Gerber C, Fuchs B, Hodler J: The results
clinical trial. Arthroscopy 2013;29:2-9. 2016;25:2-11. of repair of massive tears of the rotator
49. Jo CH, Shin JS, Lee YG, et al: Platelet-rich cuff. J Bone Joint Surg Am 2000;82:
55. Gumina S, Campagna V, Ferrazza G, et al: 505-515.
plasma for arthroscopic repair of large to Use of platelet-leukocyte membrane in
massive rotator cuff tears: A randomized, arthroscopic repair of large rotator 61. Gladstone JN, Bishop JY, Lo IK, Flatow EL:
single-blind, parallel-group trial. Am J cuff tears: A prospective randomized Fatty infiltration and atrophy of the rotator
Sports Med 2013;41:2240-2248. study. J Bone Joint Surg Am 2012;94: cuff do not improve after rotator cuff repair
50. Zhang Z, Wang Y, Sun J: The effect of 1345-1352. and correlate with poor functional
platelet-rich plasma on arthroscopic outcome. Am J Sports Med 2007;35:
56. Berebichez-Fridman R, Gomez-Garcia R, 719-728.
double-row rotator cuff repair: A clinical
Granados-Montiel J, et al: The holy grail of
study with 12-month follow-up. Acta
orthopedic surgery: Mesenchymal stem 62. Eliasberg CD, Dar A, Jensen AR, et al:
Orthop Traumatol Turc 2016;50:191-197.
cells-their current uses and potential Perivascular stem cells diminish muscle
51. Randelli P, Arrigoni P, Ragone V, Aliprandi applications. Stem Cells Int 2017;2017: atrophy following massive rotator cuff tears
A, Cabitza P: Platelet rich plasma in 2638305. in a small animal model. J Bone Joint Surg
arthroscopic rotator cuff repair: A Am 2017;99:331-341.
prospective RCT study, 2-year follow-up. J 57. Dominici M, Le Blanc K, Mueller I, et al:
Shoulder Elbow Surg 2011;20:518-528. Minimal criteria for defining multipotent 63. Klomps LV, Zomorodi N, Kim HM: Role
mesenchymal stromal cells. The of transplanted bone marrow cells in
52. Weber SC, Kauffman JI, Parise C, Weber International Society for Cellular Therapy development of rotator cuff muscle fatty
SJ, Katz SD: Platelet-rich fibrin matrix in position statement. Cytotherapy 2006;8: degeneration in mice. J Shoulder Elbow
the management of arthroscopic repair of 315-317. Surg 2017;26:2177-2186.
Research Article
The Application of Medicare Data

for Musculoskeletal Research in
the United States: A Systematic
Review
Abstract
Elham Mahmoudi, PhD Introduction: Musculoskeletal conditions disproportionately affect
Sunitha Malay, MPH the lives of aging adults. We aimed to examine the literature using
Medicare claims data in the United States for musculoskeletal surgical
Brianna L. Maroukis, BS
procedures.
Tiana Sarsour, BS Methods: Following the Preferred Reporting Items for Systematic
Kevin C. Chung, MD, MS Review and Meta-Analysis guidelines, we searched the PubMed and
Medline databases for peer-reviewed articles published between
1990 and 2015. We included the studies that (1) reported primary
Medicare claims data use, (2) involved musculoskeletal surgery, and
From the Department of Family (3) were original peer-reviewed studies. We abstracted the types of
Medicine (Dr. Mahmoudi), University
of Michigan Medical School, the surgical procedure and aims, and evaluated outcomes, and strengths
Department of Surgery (Ms. Malay and weaknesses of each included article. We assessed the quality of
and Ms. Maroukis), Section of Plastic included articles with Newcastle Ottawa Assessment Scale.
Surgery, University of Michigan
Medical School, the Michigan Health Results: The literature search returned 3,233 articles, of which 119
Science Undergraduate Research met our inclusion criteria. These studies focused on different
Academy (Ms. Sarsour), Office of outcomes: epidemiology and treatment variation (26), cost of care
Health Equity and Inclusion,
University of Michigan, Ann Arbor, MI, (15), hospital-level analyses (30), health outcomes (31), the validity
the University of Toledo, College of and accuracy of Medicare claims data (4), disparities in health care
Natural Sciences, Toledo, OH (Ms.
(10), and policy evaluation (3).
Sarsour), and the Section of Plastic
Surgery, Faculty Affairs, University of Discussion: Medicare claims data provide a unique way for
Michigan Medical School, Ann Arbor, researchers to study a nationally representative patient population
MI (Dr. Chung).
longitudinally. A significant limitation of using claims data has been a
Correspondence to Dr. Mahmoudi: lack of granularity on defining severity of a condition.
Mahmoudi@med.umich.edu
Level of Evidence: Therapeutic level III
Supported by the National Institute of
Arthritis and Musculoskeletal and Skin
Diseases of the National Institutes of
Health (2R01 AR047328-06), and a
Midcareer Investigator Award in
Patient-Oriented Research (2K24
O wing to the declining mortality
rate and aging of the pop-
ulation, musculoskeletal conditions
ditions affect individual’s activities
of daily living, use of health systems,
and direct and indirect cost of care.
AR053120-06) to Dr. Chung. The
content is solely the responsibility of
are on the rise. The World Health Direct cost of surgical treatment for
the authors and does not necessarily Organization called “2000 to 2010” these conditions varies. For example,
represent the official views of the the “Bone and Joint Decade” to raise the average incremental health care
National Institutes of Health. awareness about musculoskeletal expenditures associated with distal
J Am Acad Orthop Surg 2019;27: conditions. Osteoarthritis, rheuma- forearm and hip fractures among
e622-e632 toid arthritis, osteoporosis, and low Medicare beneficiaries were esti-
DOI: 10.5435/JAAOS-D-17-00297 back pain are a few common ex- mated to be $7,788 and $31,310,
amples of these common conditions, respectively.1 Orthopaedic-related
Academy of Orthopaedic Surgeons. leading to persistent pain and loss conditions disproportionately affect
of functioning. Musculoskeletal con- older adults. Owing to confounding
Elham Mahmoudi, PhD, et al
factors and inadequacy of medical tations, including studies on adverse lines.7 The initial search used the key
coding system in the United States, health events after surgery and cost of words “Medicare database” and
measuring health outcomes and care, using Medicare claims data. “surgery” and related terms for
complications for surgical treatment Medicare claims data were originally each. We expanded these terms to
is a challenging task. For example, developed for billing purposes and include corresponding medical sub-
Manley et al2 examined the associ- researchers should be aware of the ject headings. The search was limited
ation between hospital volume and inherent limitations of the data for to peer-reviewed articles published in
revision for total hip arthroplasty. research. For example, a lack of English after 1990 (Table 1). After
But their study had a limitation of detailed information regarding pa- eliminating duplicates, two re-
lack of information on pain and tients’ clinical, functional, and viewers separately screened the title
physical functioning. A study by Ong socioeconomic status has been fre- and abstracts of publications fol-
et al3 on the relationship between quently cited for missing con- lowed by full-text review to arrive at
procedure duration and revision of founders. Some of these limitations the final inclusive articles. We man-
total hip or knee arthroplasty also can be overcome by merging claims ually checked citations of articles
had a limitation of the lack of clinical data with other available data sets, that met our study criteria to identify
information. The authors did not such as census data, state inpatient relevant articles not found through
have information on surgical ap- data, the Area Resource File, or data the initial search.
proach or the type of repair, which from the American Hospital Asso-
could affect duration, complications, ciation. In this study, we aimed to Study Criteria
and the cost of treatment. determine the types of research
Publications were included if they
Although originally created for conducted in musculoskeletal sur-
met all the following criteria: (1)
payment and administrative pur- gery and to define the main advan-
studies reporting primary Medicare
poses, Medicare claims data possess tages and limitations of each type of
claims data use, (2) studies involving
longitudinal records of utilization research conducted. Medicare claims
musculoskeletal surgery, and (3)
and costs of health care services from data uniquely capture longitudinal
original peer-reviewed studies. We
office visits to hospitalization that records of diagnosis, surgical proce-
excluded all commentaries, abstracts,
make this data set useful for health dures, health care utilization, and
discussions, letters, reviews, edito-
service research among individuals spending for millions of adults aged
rials, and expert opinions, as well as
aged 65 years or older. Since 2010, 65 years and older throughout the
studies that did not use Medicare
under the Affordable Care Act, the United States. Our analytical evalu-
data. Discrepancies regarding article
Centers for Medicare and Medicaid ation of published musculoskeletal
inclusion/exclusion were resolved by
Services (CMS) have made efforts not research based on Medicare claims
consensus among authors who met at
only to be more transparent to the data will equip administrators and
regular intervals.
public but also to facilitate and investigators to expand their under-
encourage more investigators to use standing and potential use of the
claims data for research.4-6 Musculo- data in future. Data Extraction and Quality
skeletal surgeries are prevalent and Assessment
costly procedures. In addition, pro- The full texts of all final articles
viding a coordinated postacute care Methods included in the study were reviewed
(PAC) after these surgeries has been between August 1, 2016 and
problematic and needs further investi- Literature Search December 15, 2016. For each
gation. Despite its importance, little is We performed a systematic review of included article, we collected data
known about the range of research on the literature in PubMed and Medline regarding the author or authors,
surgical outcomes of musculoskeletal to identify all studies that used the journal, year of publication, type of
conditions or strengths and weak- Medicare database for research Medicare data file used, type of
nesses of these studies. related to musculoskeletal surgeries musculoskeletal condition evaluated,
This systematic review evaluates (Figure 1). We followed the Preferred analyses performed, study sample
the literature on musculoskeletal Reporting Items for Systematic re- characteristics, inclusion of control
surgical outcomes and their limi- views and Meta-analyses guide- groups, types of outcomes evaluated,
Ms. Sarsour or an immediate family member has stock or stock options held in Amgen, Norvartis, and Sanofi-Aventis. None of the following
authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or
institution related directly or indirectly to the subject of this article: Dr. Mahmoudi, Ms. Malay, Ms. Maroukis, and Dr. Chung.
July 1, 2019, Vol 27, No 13 e623
Musculoskeletal Research in the United States
Figure 1 reviewed publications reporting on

Medicare claims data increased by
642% (Figure 2). Because of the
heterogeneity of topics, we catego-
rized the articles into seven groups:
(1) epidemiological trends and treat-
ment variations, (2) cost, (3) hospital
evaluation, (4) health outcomes, (5)
evaluation of Medicare claims data,
(6) disparity, and (7) policy evalua-
tion (Figure 3). Quality assessment
results are available in Figure 4.
Epidemiologic Trends and

Treatment Variation Studies
Twenty-six studies focused on epi-
demiologic trends and variations in
surgical treatment patterns evaluated
conditions, such as fractures, shoul-
der conditions, ankle, hip, and knee
arthritis (see Table, Supplemental
Digital Content 1, http://links.lww.
com/JAAOS/A280). The main out-
comes of interest included frequency of
certain musculoskeletal diagnoses and
surgical procedures, providers’ spe-
cialty and variation in outcomes, and
frequency of complications, including
mortality, rehospitalization, and revi-
sion surgery. The main limitations
included an inability to define preex-
isting conditions, the imprecision of
the International Classification of
Diseases, Ninth Revision, Clinical
Modification (ICD-9) coding system,
and unknown health conditions and
health outcomes (Figure 5).
Five studies examined trends at the
Schematic flow of the literature search. patient level by examining the utili-
zation of different surgical techni-
advantages and limitations of Medi- in this systematic review was 5 ques, such as arthrodesis compared
care claims data, and specific measures (range between 5 and 8). Detailed with arthroplasty for ankle arthritis,
undertaken to overcome those limi- information regarding each study rates and outcomes of specific sur-
tations, if any. We also performed a included in this systematic review is geries, and surgical versus nonsurgi-
quality assessment of the included ar- available upon request. cal treatments. Specific outcomes
ticles using the Newcastle Ottawa examined included the length of stay
Assessment Scale. This tool assesses Results and complication rates. For example,
the quality of nonrandomized case- Bozic et al9 found no difference in
control and cohort studies and has a After screening for eligibility criteria, revision rates in total hip arthro-
maximum score of nine, which rep- 119 articles were included in the plasty patients across different
resents the highest quality.8 The qualitative analysis. Between 1990 bearing types; however, they found
minimum score for studies included and 2015, the number of peer- that metal-on-metal bearings were
Table 1
Search Algorithm
PubMed Search
(medicare data [tiab] or “medicare data set” [tiab] or “medicare database” [tiab] or “medicare claim” [tiab] or “medicare claims”
[tiab] or “MBSF” [tiab] or “MedPar” [tiab] or “Medicare Inpatient Prospective Payment System” [tiab] or “outpatient file” [tiab]
or “carrier file” [tiab] or “inpatient file” [tiab] or “outpatient files” [tiab] or “carrier files” [tiab] or “inpatient files” [tiab] or
“Medicare HOS” [tiab] or “Medicare Current Beneficiary Survey” [tiab] or “Medicare Current Beneficiary Surveys” [tiab] or
“Medicare Beneficiary Survey” [tiab] or “Medicare Current Beneficiaries Survey” [tiab] or (“medicare” [tiab] and “MCBS”
[tiab]) or (“medicare” [tiab] and “administrative data” [tiab]) or “public use file” [tiab] or “public use files” [tiab] or “medicare
enrollment database” [tiab] or (“medicare” [tiab] and “linked” [tiab])) AND (“surgery” [tiab] or “surgeries” [tiab] or “surgical”
[tiab] or “operation” [tiab] or “operations” [tiab] or “procedure” [tiab] or “procedures” [tiab] or “operative” [tiab] or “surgery
department, hospital” [mh] or “surgical procedures, operative” [mh]) AND “english” [la] AND “1990/01/01” [dp]: “3000” [dp]
MEDLINE Search
(((medicare adj1 (data* or claim*)) or MBSF or MedPar or (Medicare Inpatient Prospective Payment System) or (outpatient
file*) or (carrier file*) or (inpatient file*) or (Medicare HOS) or (Medicare Current Beneficiar* Survey*) or (medicare AND
MCBS) or (medicare enrollment data*) or (medicare administrative data*) or (medicare AND linked)).mp. AND ((surg* or
operat* or procedure*).mp. OR ’exp surgery department, hospital/or exp surgical procedures, operative/) AND english.la
AND limits: publication yr 1990-current.
associated with a higher risk of joint Figure 2

infection, mechanical loosening, and
deep vein thrombosis than other
bearing types. Mahomed et al10
concluded that the rates of primary
total hip arthroplasty were three to
six times higher than those of revi-
sion total hip arthroplasty. How-
ever, the 90-day postoperative
adverse outcomes were higher after
revision than after primary total hip
arthroplasty. They found that being
an older African-American male
patient and having low income were
associated with adverse outcomes.
Some of these studies linked Medi-
care claims with other data, such as
American Hospital Association sur-
vey, Rural-Urban Commuting area,
National Hospital Discharge Survey,
or the Dartmouth Atlas of
Musculoskeletal Care11 to obtain
more details on patients’ socioeco- Number of peer-reviewed publications on musculoskeletal surgical procedures
nomic status or providers’ density using Medicare claims data.
information. Finally, some studies
focused on patient and provider increased between 2000 and 2010. nomic status increased the likelihood
factors that influenced surgical Approximately 500,000 osteopo- to receive surgical treatment.
treatment,12,13 with some specifi- rotic compression fractures occur
cally focusing on physician specialty. annually among the elderly in the Cost-analysis Studies
Long et al14 noted that despite the United States, costing up to $600 Fifteen studies examined the cost of
lack of level-I evidence for long-term million per year. Zhong et al12 found musculoskeletal surgeries from the
effectiveness of vertebral augmenta- that among Medicare beneficiaries beneficiary or payer perspective (see
tion procedures, use of verte- with rheumatoid arthritis, being Table, Supplemental Digital Content 2,
broplasty and kyphoplasty had white and having higher socioeco- http://links.lww.com/JAAOS/A281).
July 1, 2019, Vol 27, No 13 e625
Figure 3 istics and capital loss as a result of a

condition or procedure, and inabil-
ity to measure any type of cost-
effectiveness and efficiency analysis
(Figure 6).
Belatti et al15 found that between
2000 and 2011, the utilization of
total joint arthroplasty grew sub-
stantially (27%) until 2005, after
which the growth became stagnant.
Owing to the new coding scheme,
despite a decrease in physician
reimbursement (22.3% per year),
increase in hospital reimbursement
was associated with complicated
cases (14.5% per year) versus
uncomplicated cases (20.65%).15
Another study reported significant
geographic variation in costs, par-
ticularly for costs of specific episodes
of care and total annual cost of care
Categories of peer-reviewed publications on musculoskeletal surgical among the 10 costly and common
procedures using medicare claims data.
conditions examined (ie, knee joint
or lower leg degeneration, joint
degeneration of back or neck).16
Figure 4
Cost-effectiveness analyses for the
treatment of knee osteoarthritis17,18
and vertebral compression fractures
were investigated. For example, Gu
et al19 found that the societal benefits
of hip fracture surgery outweigh its
direct costs, with an average savings
of around $66,000 per patient. Cost
comparisons, for example, between
previously insured and uninsured
Medicare beneficiaries revealed that
previously uninsured adults had
higher hospitalization rates and
therefore spent more when they were
eligible for Medicare.20 Annual per
person Medicare spending for pre-
viously insured versus uninsured was
$ 4,773 versus $ 5,796.20 Also, cost
Quality assessment of included articles. comparisons between 2-year postsur-
gical expenditures for vertebroplasty
and kyphoplasty showed lower ad-
These studies examined the mean ment charges, Medicare payments justed cumulative treatment costs for
health plan payment per hospital for the initial surgery, and any kyphoplasty, reflecting lower utili-
admission or the amount paid by morbidities up to 1 year after the zation of medical resources.21 Fi-
Medicare, with further subdivision surgery were also examined. The nally, Ong et al22 concluded that the
into professional services pay- main limitations included impreci- economic burden of revision surgery
ments. Payment per beneficiary and sion in cost measurement, a lack of for total hip and knee arthroplasties
hospital reimbursement and treat- information on patient character- was more than the initial surgery.
Belatti and Phisitkul23 estimated that Figure 5

between 2000 and 2011, economic
burden of revision of foot and ankle
surgery increased by 38%. Owing to
their main use as reimbursement
mechanism, Medicare claims data
provide an excellent source for cost
and utilization analyses.
Hospital-level Studies
Thirty studies investigated Medicare
claims data at the hospital level (see
Table, Supplemental Digital Content 3,
http://links.lww.com/JAAOS/A282).
They compared hospitals based on
their surgical complication rates and Limitations of studies using Medicare claims to evaluate epidemiologic trends
and treatment.
by assessing the effect of the creation
of new hospital arthroplasty pro-
grams on arthroplasty utilization.
The main limitations included diffi- Figure 6
culty in measuring the patient case
mix in each hospital, a lack of data
on important clinical outcomes, diffi-
culty in assessing provider’s volume, a
lack of data on Medicare Managed
Care beneficiaries, and potential mis-
coding errors (Figure 7).
Two studies explored factors
influencing hospital choice among
patients. For example, 74% of pa-
tients who had revision surgery for
total hip arthroplasty had the surgery
in a hospital with the same volume
category as they had for their primary
surgery (64% of this group chose the
same hospital for both surgeries).24 Limitations of studies using Medicare claims to evaluate costs.
Hospital ownership, ranking, vol-
ume, specialization, location, and
teaching status were also examined cedure volume and outcomes. Katz and fewer complications for complex
in relationship to utilization, cost of and colleagues26,27 found that for surgical procedures.
procedures performed, and out- total hip and knee arthroplasties,
comes (eg, mortality, readmission high-volume surgeons and hospitals
rates, length of stay). For example, were associated with low rates of Health Outcome Studies
Cram et al25 found no significant mortality and complications and Thirty-one studies primarily examined
differences in postoperative compli- high rate of patient satisfaction. health outcomes that include major
cations and hospital length of stay Manley et al2 found that patients complications postsurgery and read-
after total knee arthroplasty between who had hip arthroplasties per- mission for surgery-related complica-
top-ranked hospitals, non–top- formed by low-volume surgeons tions, mortality, and readmission (see
ranked hospitals, and hospitals (fewer than 25 procedures/yr) were Table, Supplemental Digital Content
ineligible for ranking. Finally, five at a greater risk of having a revision 4, http://links.lww.com/JAAOS/A283).
studies examined the association surgery 6 months later. Medicare The main complications included a
between hospital and surgeon pro- claims data show better outcomes lack of detailed outcome information,
July 1, 2019, Vol 27, No 13 e627
Figure 7 120 to 150 minutes had significantly

lower revision rates than those last-
ing less than 90 minutes (P = 0.008)
or more than 150 to 180 minutes
(P = 0.004).3 This indicates that
longer procedure duration results in
increased complications and affects
resource use and insurance reim-
bursement. George et al31 found that
patients who underwent total knee
arthroplasty experienced significant
improvements in activities of daily
living compared with those who
opted for nonsurgical treatment (P ,
0.010). This will serve as an indica-
Limitations studies using Medicare claims to evaluate hospitals. tor to declining rates of disability in
elderly along with the effects of other
public health measures, such as bet-
Figure 8 ter health habits.
Validity of Medicare Claims

Data
Four studies evaluated the validity
and accuracy of Medicare claims data
(see Table, Supplemental Digital Con-
tent 5, http://links.lww.com/JAAOS/
A284). These studies examined how
well Medicare data identified revision
of total hip arthroplasties and detected
surgical site infection; they also
examined the use of present-on-
admission (POA) codes for common
Limitations studies using Medicare claims to evaluate health outcomes. complications after total knee arthro-
plasty. The main limitations included
potential coding errors, a lack of
ambiguity and limitations of the ICD-9 lower for shoulder arthroplasty detailed medical information, and a
coding system, potential coding errors, compared with lower extremity ar- lack of data on Medicare Managed
and exclusion of the Medicare Man- throplasty (0.53% versus 1.2%; P , Care beneficiaries (Figure 9).
aged Care population from the data 0.001). Therefore, the current stan- Calderwood et al32 found that
(Figure 8). dard of care of no chemoprophylaxis claims-based surveillance was more
For example, Modhia et al28 found after upper extremity arthroplasty is accurate than usual infection control
that patients had lower readmission appropriate. Katz et al30 found that surveillance (specific to each hospi-
rates after decompression alone total hip arthroplasty revisions were tal) at detecting surgical site in-
compared with fusion along with most likely to happen within fections after total knee or total hip
decompression procedure for lumbar 18 months after the primary surgery. arthroplasty (1.8 to 4.7-fold in-
spinal stenosis (7.2% versus 9.7%). Thus, more efforts targeted to moni- crease). However, Katz et al33 used
Reduced readmissions are important tor patients within the 18 months of Medicare claims data to identify
and can improve the cost-effectiveness postoperative period can minimize revision hip arthroplasty, as opposed
of this surgical treatment. Eight studies subsequent revisions. Effect of pro- to primary hip arthroplasty on the
analyzed complications. Day et al29 cedure length on outcomes was as- contralateral hip. They found that
found that the incidence of venous sessed in two studies because total Medicare data correctly identified
thromboembolism after surgery was knee arthroplasty procedures lasting revision of the primary surgery only
71% of the time. Also, ICD-9 coding Figure 9

system does not specify the side of
the surgery and thus is not a reliable
source for studying the risk of revi-
sion after hip arthroplasty. Starting
from October 1, 2015, CMS adopted
the International Classification of
Diseases, Tenth Revision, Clinical
Modification (ICD-10) codes, which
provide bilateral information on
procedures performed. Cram et al34
determined that POA information
was completed by hospitals for only
60% to 75% of complications.
Introduced by the CMS in October
2007, POA codes are important
because they add significant informa-
Limitations of studies using Medicare claims to evaluate validity of coding,
tion for determining complications disparities, and health policy.
after total knee arthroplasty. Finally,
research shows that claims data lack
adequate sensitivity to be used for the longer (average, 1.6 and 1.7 days, Johanson et al40 observed a pro-
selection of patients based on certain respectively) to have surgery after gressive increase in the utilization of
diagnoses, such as rheumatoid hip fracture than Caucasians (aver- arthroscopy after Centers for Medi-
arthritis or osteonecrosis. A more age, 1.2 days), suggesting the need to care and Medicaid changed their
conservative approach is warranted monitor care among minorities with coverage policy of arthroscopic
when defining patients’ cohorts. hip fracture. Ecsarce and McGuire37 procedures. Less invasive and low-
observed that racial differences ex- cost arthroscopy had a 10.2% fail-
isted in the usage of specific medical ure rate after 1 year, indicating its
Disparity Analysis procedures and tests with African- effectiveness in delaying arthroplasty
Ten articles conducted disparity Americans having lower usage rates and minimizing costs associated with
analyses to examine racial/ethnic dis- than Caucasians. Skinner et al38 revision arthroplasty.40 Miller et al41
parities in the utilization of knee and found that large variation in knee evaluated that the efficiency of inte-
hip arthroplasties (see Table, Supple- arthroplasty existed based on sex, grated delivery systems (IDSs) for
mental Digital Content 5, http:// race, and region. Finally, Tsai et al39 quality and cost was beneficial for
links.lww.com/JAAOS/A284). The also noted that African-Americans ambulatory care and preventive serv-
main limitations included difficulty in compared with Caucasians had ices but not for inpatient surgery when
identifying Hispanic populations, a higher odds of 30-day readmission compared with similar Accountable
lack of data on patient characteristics, after surgery, which has implications Care Organizations. The quality of
such as smoking or body mass index, for quality and cost of care. Although care measured by surgical mortality,
potential misidentification of race/ Medicare is a publicly available in- postoperative complications, and re-
ethnicity, and a lack of data on surance program, our results indicate admissions and total episode cost for
socioeconomic and functional status wide racial/ethnic disparities in the patients treated in IDS and non-IDS
of patients (Figure 9). use of musculoskeletal surgeries. hospitals were similar. FitzGerald
Singh et al35 found that compared et al42 studied the effect of the Short
with Caucasians, African-Americans Stay Transfer Policy that discourages
had 40% lower probability of Policy Evaluation early discharge to PAC on joint ar-
undergoing knee and hip arthro- Three studies evaluated various pol- throplasty and hip fracture surgery.
plasties and 24% higher probability icy changes (see Table, Supplemental The study found an immediate
of readmission after surgery. These Digital Content 5, http://links.lww. increase followed by stabilization in
disparities have remained consistent com/JAAOS/A284). The main limi- the hospital length of stay for joint
from 1991 to 2008. Nguyen-Oghalai tations included a lack of data on arthroplasty and hip fracture sur-
et al36 observed that African- patient characteristics, health status, geries (0.20 and 0.17 days, respec-
Americans and Hispanics waited and socioeconomic status (Figure 9). tively) and a decrease in early
July 1, 2019, Vol 27, No 13 e629
discharge to PAC (4.3 and 3.0 per- on Medicare inpatient admissions searchers appreciating the unique
centage points, respectively).42 Short since October 1, 2007) to distinguish opportunity provided by the data.
Stay Transfer Policy implementation between chronic and acute con- Earlier work during the 1980s
reduced the financial incentive for ditions.29 Cost-analysis studies used focused more on the hospital reim-
hospitals to discharge patients early MarketScan data, survey data,20 or bursement system. By constructing
to PAC. industry surveys of implant prices to episodes of care, researchers have
indirectly determine the profitability been able to use the data for outcome
of the procedure.15 analyses. Although there is no uni-
Noted Limitations versal way of defining the sample
Most common limitations across all cohort, researchers have used a
seven categories were the potential in- Discussion combination of ICD-9 and Current
accuracy in medical coding and the lack Procedural Terminology (CPT) codes
of detailed clinical data, such as pain or Our systematic review examined all to define specific patient cohort and
functioning, fracture pathology, dis- published literature between January health outcomes. Administrative
ease and complication severity, reason 1990 and December 2015 on surgical claims data are not as granular as
for revision surgery or readmission, treatment of musculoskeletal con- clinical data. Thus, researchers have
patient satisfaction, surgical technique, ditions using Medicare claims data. had to use a variety of measures to
and type of bearing surface. There was This study is the first to systematically define health outcomes. Mortality
also a lack of laterality in ICD-9 codes, review the use of Medicare claims rate at certain time intervals after
which is particularly important for hip data for different musculoskeletal undergoing a certain procedure,
and knee arthroplasties, and an inabil- conditions in the United States. Other recurrence of the same procedure,
ity to capture concerns related to systematic reviews on musculo- infection, or 30-day rehospitalization
patient and provider decision mak- skeletal conditions have focused on were used in different studies to
ing. Specific limitations of cost-analysis clinical conditions rather than on the measure health outcomes. With the
studies were the discrepancy between use of secondary data, such as implementation of the ICD-10 cod-
charge and reimbursement data and Medicare claims data for different ing system, there will be more
significant geographic variations in types of studies related to musculo- opportunities to develop new out-
the cost of a procedure. Articles skeletal conditions. We identified come and complication measures.
examining racial/ethnic disparity the aims achieved by these studies The past few years have seen a surge
mentioned that the low number of and grouped them into seven cate- in using Medicare claims data to
Hispanics and African-Americans gories and then detailed the chal- evaluate policies, such as the value-
as compared with Caucasians, espe- lenges that these studies faced and based system, hospital quality sys-
cially in rural areas, might have whether they overcame the limi- tem, and preventive services. Use of
underpowered some analyses.36 tations of the data. Medicare claims data in research is
Researchers overcame some limi- The main limitation or challenge in expected to grow. Compared with
tations by linking Medicare data to our systematic review was the het- clinical trials, these data are less
other data sets such as American erogeneity of studies using Medicare expensive, population based, allow
Hospital Association data, or US claims data. We dealt with this chal- patients and hospitals to be examined
News & World Report, Dartmouth lenge by stratifying final included longitudinally, and are relatively
Atlas of Health Care, Rural Urban studies into seven groups based on reliable. Our review findings will
Area Commuting data, and Census their defined objectives. Although the serve as a source of reference for re-
data. Some clinical data, such as groups sometimes overlapped, for searchers seeking to examine Medicare
physical functioning, could be de- each group, we explained how the claims data on topics related to
termined by supplementing Medi- data were used, what measures were musculoskeletal conditions.
care claims data with surveys.29,31 defined as outcomes, and how com-
One study validated laterality through plication rates have been defined. This
medical records using a subset sam- heterogeneity in the use of Medicare References
ple.24 Another study collected data for claims shows the growing popularity
3 years before the index surgery to of this database in health services References printed in bold type are
determine whether previous surgery research. those published within the past 5
was performed, rather than relying Over the past two decades, the use years.
on medical codes. Authors also of claims data in health services 1. Kilgore ML, Morrisey MA, Becker DJ, et al:
suggested using POA codes (required research has evolved, with more re- Health care expenditures associated with
skeletal fractures among Medicare 13. Chung KC, Shauver MJ, Yin H, Kim HM, 26. Katz JN, Losina E, Barrett J, et al:
beneficiaries, 1999-2005. J Bone Miner Res Baser O, Birkmeyer JD: Variations in the Association between hospital and
2009;24:2050-2055. use of internal fixation for distal radial surgeon procedure volume and
fracture in the United States medicare outcomes of total hip replacement
2. Manley M, Ong K, Lau E, Kurtz SM: Effect population. J Bone Joint Surg Am 2011;93: in the United States Medicare
of volume on total hip arthroplasty revision 2154-2162. population. J Bone Joint Surg 2001;
rates in the United States Medicare 83-A:1622-1629.
population. J Bone Joint Surg Am 2008;90: 14. Long SS, Morrison WB, Parker L:
2446-2451. Vertebroplasty and kyphoplasty in the 27. Katz JN, Barrett J, Mahomed NN, Baron
United States: Provider distribution and JA, Wright RJ, Losina E: Association
3. Ong KL, Lau E, Manley M, Kurtz SM: guidance method, 2001-2010. AJR Am J between hospital and surgeon procedure
Effect of procedure duration on total hip Roentgenol 2012;199:1358-1364. volume and the outcomes of total knee
arthroplasty and total knee arthroplasty replacement. J Bone Joint Surg Am 2004;
survivorship in the United States Medicare 15. Belatti DA, Pugely AJ, Phisitkul P, 86-A:1909-1916.
population. J Arthroplasty 2008;23(6 suppl Amendola A, Callaghan JJ: Total joint
1):127-132. arthroplasty: Trends in medicare 28. Modhia U, Takemoto S, Braid-Forbes
reimbursement and implant prices. J MJ, Weber M, Berven SH:
4. CMS: Medicare Provider Utilization and Arthroplasty 2014;;29:1539-1544. Readmission rates after decompression
Payment Data: Outpatient. https://www. surgery in patients with lumbar
cms.gov/Research-Statistics-Data-and- 16. Reschovsky JD, Hadley J, O’Malley AJ, spinal stenosis among Medicare
Systems/Statistics-Medicare-Provider/ Landon BE: Geographic variations in the beneficiaries. Spine (Phila Pa 1976)
Medicare-Provider-Charge-Data/ cost of treating condition-specific episodes 2013;38:591-596.
Outpatient.html. Accessed December 1, of care among Medicare patients. Health
2015. Serv Res 2014;49:32-51. 29. Day JS, Ramsey ML, Lau E, Williams GR:
Risk of venous thromboembolism after
5. ÇMS: Medicare Provider Utilization and 17. Suter LG, Paltiel AD, Rome BN, et al: shoulder arthroplasty in the Medicare
Payment Data: Inpatient. https://www.cms. Placing a price on medical device population. J Shoulder Elbow Surg 2015;
gov/Research-Statistics-Data-and-Systems/ innovation: The example of total knee 24:98-105.
Statistics-Medicare-Provider/Medicare- arthroplasty. PLoS One 2013;8:e62709.
Provider-Charge-Data/Inpatient.html. 30. Katz JN, Wright EA, Wright J, et al:
Accessed December 1, 2015. 18. Losina E, Walensky RP, Kessler CL, et al: Twelve-year risk of revision after primary
Cost-effectiveness of total knee arthroplasty total hip replacement in the U.S. Medicare
6. CMS: CMS. Medicare Provider Utilization in the United States: Patient risk and population. J Bone Joint Surg Am 2012;94:
and Payment Data: Physician and Other hospital volume. Arch Intern Med 2009; 1825-1832.
Supplier. https://www.cms.gov/Research- 169:1113-1121.
Statistics-Data-and-Systems/Statistics- 31. George LK, Ruiz D Jr, Sloan FA: The
Trends-and-Reports/Medicare-Provider- 19. Gu Q, Koenig L, Mather RC III, Tongue J: effects of total knee arthroplasty on
Charge-Data/Physician-and-Other- Surgery for hip fracture yields societal physical functioning in the older
Supplier.html. Accessed December 1, 2015. benefits that exceed the direct medical costs. population. Arthritis Rheum 2008;58:
Clin Orthop Relat Res 2014;472: 3166-3171.
7. Moher D, Liberati A, Tetzlaff J, et al: 3536-3546.
Preferred reporting items for systematic 32. Calderwood MS, Ma A, Khan YM, et al:
reviews and meta-analyses: The PRISMA 20. McWilliams JM, Meara E, Zaslavsky AM, Use of Medicare diagnosis and procedure
statement. Ann Intern Med 2009;151: Ayanian JZ: Medicare spending for codes to improve detection of surgical site
W64-W69. previously uninsured adults. Ann Intern infections following hip arthroplasty,
Med 2009;151:757-766. knee arthroplasty, and vascular surgery.
8. Wells GSB, O’Connell D, peterson J, Welch Infect Control Hosp Epidemiol 2012;33:
V, Losos M, Tugwell P: The Newcastle– 21. Ong KL, Lau E, Kemner JE, Kurtz SM:
40-49.
Ottawa Scale (NOS) for Assessing the Two-year cost comparison of
Quality of Nonrandomised Studies in vertebroplasty and kyphoplasty for the 33. Katz JN, Wright EA, Baron JA, et al:
Meta-analyses. http://www.ohri.ca/ treatment of vertebral compression Predictive value of Medicare claims data for
programs/clinical_epidemiology/oxford. fractures: Are initial surgical costs identifying revision of index hip
asp. Accessed January 3, 2017. misleading? Osteoporos Int 2013;24: replacement was modest. J Clin Epidemiol
1437-1445. 2011;64:543-546.
9. Bozic KJ, Lau EC, Ong KL, Vail TP,
Rubash HE, Berry DJ: Comparative 22. Ong KL, Mowat FS, Chan N, Lau E, 34. Cram P, Bozic KJ, Callaghan JJ, Lu X, Li Y:
effectiveness of metal-on-metal and metal- Halpern MT, Kurtz SM: Economic burden Use of present-on-admission indicators for
on-polyethylene bearings in Medicare total of revision hip and knee arthroplasty in complications after total knee arthroplasty:
hip arthroplasty patients. J Arthroplasty Medicare enrollees. Clin Orthop Relat Res An analysis of Medicare administrative
2012;27(8 suppl):37-40. 2006;446:22-28. data. J Arthroplasty 2014;29:923-928.e2.
10. Mahomed NN, Barrett JA, Katz JN, et al: 23. Belatti DA, Phisitkul P: Economic burden of 35. Singh JA, Lu X, Rosenthal GE, Ibrahim S,
Rates and outcomes of primary and foot and ankle surgery in the US Medicare Cram P: Racial disparities in knee
revision total hip replacement in the United population. Foot Ankle Int 2014;35: and hip total joint arthroplasty: An
States medicare population. J Bone Joint 334-340. 18-year analysis of national Medicare
Surg Am 2003;85-A:27-32. data. Ann Rheum Dis 2014;73:
24. Katz JN, Wright EA, Wright J, et al: Choice 2107-2115.
11. Koval KJ, Lurie J, Zhou W, et al: Ankle of hospital for revision total hip
fractures in the elderly: What you get replacement. J Bone Joint Surg Am 2010; 36. Nguyen-Oghalai TU, Kuo YF, Wu H, et al:
depends on where you live and who you 92:2829-2834. The impact of race/ethnicity on
see. J Orthop Trauma 2005;19:635-639. preoperative time to hip stabilization
25. Cram P, Cai X, Lu X, Vaughan-Sarrazin procedure after hip fracture. South Med J
12. Zhong L, Chung KC, Baser O, Fox DA, MS, Miller BJ: Total knee arthroplasty 2010;103:414-418.
Yuce H, Waljee JF: Variation in rheumatoid outcomes in top-ranked and non-top-
hand and wrist surgery among medicare ranked orthopedic hospitals: An analysis of 37. Escarce JJ, McGuire TG: Changes in racial
beneficiaries: A population-based cohort Medicare administrative data. Mayo Clin differences in use of medical procedures and
study. J Rheumatol 2015;42:429-436. Proc 2012;87:341-348. diagnostic tests among elderly persons:
July 1, 2019, Vol 27, No 13 e631
1986-1997. Am J Public Health 2004;94: for Medicare beneficiaries by race and 41. Miller DC, Ye Z, Gust C, Birkmeyer JD:
1795-1799. site of care. Ann Surg 2014;259: Anticipating the effects of accountable care
1086-1090. organizations for inpatient surgery. JAMA
38. Skinner J, Zhou W, Weinstein J: The Surg 2013;148:549-554.
influence of income and race on total knee 40. Johanson NA, Kleinbart FA, Cerynik DL,
arthroplasty in the United States. J Bone Brey JM, Ong KL, Kurtz SM: Temporal 42. FitzGerald JD, Boscardin WJ, Hahn BH,
Joint Surg Am 2006;88:2159-2166. relationship between knee arthroscopy Ettner SL: Impact of the Medicare Short
and arthroplasty: A quality measure for stay transfer policy on patients undergoing
39. Tsai TC, Orav EJ, Joynt KE: Disparities joint care? J Arthroplasty 2011;26: major orthopedic surgery. Health Serv Res
in surgical 30-day readmission rates 187-191. 2007;42:25-44.

5 6242249775165472907 PDF

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

5 6242249775165472907 PDF

Diunggah oleh

Hak Cipta:

Format Tersedia

Research Article

A 46-year Analysis of Gender

Chang-Yeon Kim, MD, MS Introduction: Participation of women in medicine has increased

Copyright 2019 by the American

July 1, 2019, Vol 27, No 13 493

B. The number of articles and authors

494 Journal of the American Academy of Orthopaedic Surgeons

July 1, 2019, Vol 27, No 13 495

Proportion of female authors on a yearly basis from 1972 to 2018.

496 Journal of the American Academy of Orthopaedic Surgeons

Distribution of publication count: frequency (%)

July 1, 2019, Vol 27, No 13 497

Figure 3 In terms of research longevity,

498 Journal of the American Academy of Orthopaedic Surgeons

1 89.6% 75.8% 82.1% 68.8% 75.9% 65.6% 70.6% 63.7%

Analysis of the Total Digital Content 1, http://links.lww.

July 1, 2019, Vol 27, No 13 499

500 Journal of the American Academy of Orthopaedic Surgeons

July 1, 2019, Vol 27, No 13 501

Artistic Gymnastics Injuries;

Natasha Desai, MD Artistic gymnastics is a physically demanding sport that requires

A rtistic gymnastics is known for

July 1, 2019, Vol 27, No 13 459

460 Journal of the American Academy of Orthopaedic Surgeons

Diagram of a gymnast showing biomechanics of landing after dismount,

July 1, 2019, Vol 27, No 13 461

Figure 4 rest of approximately 10 to 12 weeks investigated that will decrease the

462 Journal of the American Academy of Orthopaedic Surgeons

July 1, 2019, Vol 27, No 13 463

Figure 7 Figure 8 Figure 9

Coronal MRI image demonstrating Arthroscopic image of the elbow

464 Journal of the American Academy of Orthopaedic Surgeons

Figure 10 Figure 11 Figure 12

Intraoperative image of the medial

July 1, 2019, Vol 27, No 13 465

Figure 13 Figure 14 Figure 15

AP radiograph of the wrist

466 Journal of the American Academy of Orthopaedic Surgeons

13. Christoforidou Α, Patikas D, Bassa E, et al: shoulder in elite female gymnasts. Am J

July 1, 2019, Vol 27, No 13 467

Business Modeling of Orthopaedic

Correspondence to Dr. Jacobs:

None of the following authors or any

e612 Journal of the American Academy of Orthopaedic Surgeons

July 1, 2019, Vol 27, No 13 e613

e614 Journal of the American Academy of Orthopaedic Surgeons

220 true positives and 1,384 true neg- Table 1

Analysis of the actual professional fees

July 1, 2019, Vol 27, No 13 e615

5 5 $16,125 $2,104 $218 $18,447 $3,689

29105 Long arm 101 17 118 86

CPT = Current Procedural Terminology, ED = emergency department

25605 Distal radius 200 20 220 91

CPT = Current Procedural Terminology, ED = emergency department

e616 Journal of the American Academy of Orthopaedic Surgeons

29105 Long arm 101 17 118 86 $95,739.51 $81,946.53

CPT = Current Procedural Terminology, ED = emergency department

25605 Distal radius 200 20 220 91 $178,291.30 $162,083.00

CPT = Current Procedural Terminology, ED = emergency department

July 1, 2019, Vol 27, No 13 e617

e618 Journal of the American Academy of Orthopaedic Surgeons

July 1, 2019, Vol 27, No 13 e619

e620 Journal of the American Academy of Orthopaedic Surgeons