Medical Manuscript

American Journal of Hospice

& Palliative Medicine®
Bereaved Parents’ Health Status During the 1-11
ª The Author(s) 2017
Reprints and permission:
First 6 Months After Their Child’s Death sagepub.com/journalsPermissions.nav
DOI: 10.1177/1049909117744188
journals.sagepub.com/home/ajh

Nancy Dias, PhD, RN, CNE1 , Debra Brandon, PhD, RN, CNS, FAAN2,
Joan E. Haase, PhD, RN, FAAN3, and Paula Tanabe, PhD, RN, FAEN, FAAN2

Abstract
Purpose: To examine bereaved parents’ physical, mental, and social health during the first 6 months after their child’s (<12 years)
death from a life-threatening illness. Background and Significance:: Bereaved parents have higher mortality and morbidity
rates when compared to nonbereaved parents. Acute illnesses, hospitalizations, and medication changes are highest in the first
6 months. An understanding of bereaved parents’ health risk indicators can help inform development of health promotion and
disease prevention measures. Methods and Analysis:: A prospective descriptive study examined 8 parent dyads. Parents
completed health surveys (Patient-Reported Outcomes Measurement Information System–global, social, and sleep; Brief
Symptom Inventory [BSI] 18), which are used to assess parents’ health at 3 and 6 months after their child’s death.
Demographic data included a medical history, hospital or emergency department visits, and smoking and alcohol intake.
Descriptive statistics were used to compare parents’ scores to US general population scores. Findings: Mothers’ and fathers’
physical, mental, and sleep health scores were typically within 1 to 2 standard deviations of the population norms. However, their
social health scores were as low as 3 standard deviations and all parents’ scores were below population norms. Four (25%) of the
16 parents had new diagnosis during the first 6 months. Based on the BSI-18, 3 parents had their scores above population
cutoffs, which warranted a need for further clinical evaluation. Conclusions: Health data highlight the “at-risk” health status of
bereaved parents. Further validation of these data is required to support the development of health promotion and disease
prevention programs.

Keywords
parental bereavement, bereaved parents, health status, health outcomes, pediatric bereavement care, bereavement programs,
child’s death, palliative and end-of-life

Introduction Hospitalizations and medication changes are also highest in the

According to the Centers for Disease Control and Prevention,
23 215 infants and an additional 9000 children aged between 1
and 14 years died in 2014.1 The death of a child can completely
transform the lives of the parents and can negatively impact
their health.2-4 However, there are discrepancies in the litera-
ture regarding the health of parents after their child’s death.
Death rates are highest among parents in the first 3 years after a
child’s death. The mortality rates for bereaved parents is 2 to 4
times higher when compared to nonbereaved parents.5 In con-
trast, some studies indicate that bereaved parents’ mortality and
morbidity rates are no different from the general population of
parents.6 Mortality rates of bereaved parents monitored up to
19 years after a child’s death showed no increase when com-
pared to nonbereaved parents.6
Similar discrepancies are documented on bereaved parents’
morbidity data. Higher incidences of physical and mental con-
ditions such as cancer, cardiac illnesses, anxiety, depression,
and post-traumatic stress symptoms were reported in bereaved
parents when compared to nonbereaved parents. 7-12
first 6 months following a child’s death.13 At 5 years after a
child’s death, mental distress and trauma is 2 to 3 times higher
when compared to the general population.14 Long-term nega-
tive health outcomes are documented as late as 35 years after a
child’s death.15 Moreover, parents experience a decrease in
their social interactions and some experience negative changes
in relationships with others including their spouse.13,16 Alter-
nately, some studies reported no differences in the risk of
stroke,17 myocardial infarction,10 cancer survival,6,18 and
1
College of Nursing, Department of Nursing Science, East Carolina University,
Greenville, NC, USA
2
Duke University School of Nursing, Durham, NC, USA
3
School of Nursing, Indiana University–Purdue University at Indianapolis,
Indianapolis, IN, USA
Corresponding Author:
Nancy Dias, PhD, RN, CNE, East Carolina University, 4165 F Health Science
Building, Greenville, NC 27858, USA.
Email: diasn17@ecu.edu
2 American Journal of Hospice & Palliative Medicine® XX(X)
frequency/duration of hospitalization19 between bereaved and
nonbereaved parents. Despite these discrepancies, most of the
research indicates that all aspects of a parent’s health including
their physical, mental, and social health are affected after a
child’s death.
Several factors contribute to the discrepancies in research
on bereaved parents’ health: variations in study methods (ret-
rospective or prospective), sample sizes (small to large
population-based studies), deceased child’s age (newborn to
adult child), time since death (3 months to 35 years), parent’s
age (young to older adult), and psychometric properties of
assessment tools (standardized established tool or self-report
survey).20 In addition, most studies that report no difference
between the bereaved and nonbereaved parents were con-
ducted outside the United States including Denmark, 10,12
Sweden,21 and Israel.3,6,22 These discrepancies may reflect
the morbidity and mortality of the different geographical
locations and the cultural attitudes related to death and
dying in that region. These variations continue to fuel the
debate on whether or not a child’s death influences bereaved
parents’ health.
In the United States, there are limited and conflicting data
on the health of bereaved parents after losing a child. Bire-
nbaum and colleagues found bereaved parents’ health is not
affected after their child’s death.23 In contrast, Youngblut et al
found bereaved parents’ health may be “at risk” after losing
their child, specifically they had changes in behaviors such as
eating, sleeping, smoking, drinking alcohol, as well as changes
in body weight, blood pressure, and blood sugar that could
serve as health risk indicators in the first year following a
child’s death.13
To better understand the bereaved parents’ health affects
after their child’s death, it is essential to gather enough data
to address the following unanswered questions: When do
adverse changes in health become evident in these parents?
Can assessment tools identify at-risk parents? What is the
appropriate time to administer the assessment tools? Is it
feasible for parents to fill out these assessment tools, con-
sidering their emotional status after their child’s death? This
information can help determine the need for and the timing
of health screening, so that multidisciplinary intervention
programs with a focus on health promotions and disease
preventive measures for bereaved parents can be designed
and tested.
The purpose of this study was to assess bereaved parents’
physical, sleep, mental, and social health with a goal to identify
early health risk indicators during the first 6 months after their
child’s death. To understand which aspects of bereaved par-
ents’ health is most affected, this study used established and
quantifiable health survey outcome measures recommended by
the National Institute of Health, the Patient-Reported Out-
comes Measurement Information System (PROMIS) scales to
assess all aspects of bereaved parents’ physical, mental, and
social health. Additionally, since there is ample evidence that
bereaved parents sleep is negatively impacted,24,25 we also
included a sleep assessment scale (PROMIS, sleep). To
evaluate the time frames for changes in bereaved parents’
health, we measured health risk indicators across 2 time points
(3 and 6 months) after their child’s death. This time frame is the
most stressful time after their child’s death13,26 and will pro-
vide data on the early indicators of poor health, which would
allow for early assessment and intervention for parents exhibit-
ing high-risk criteria. Screening programs and early interven-
tion for bereaved parents will guide the design of bereavement
programs with a focus on health promotion and disease
prevention, thus improving long-term health outcomes of
bereaved parents.26
Methods
Design
We used a prospective case-based study approach to assess
bereaved parents’ health and bereavement challenges during
the first 6 months after their child’s death. Each case
included parent dyads (mother and father) of children below
12 years who had died of any life-threatening illness. Indi-
vidual interviews were conducted, and parents were also
asked to complete 5 health surveys and a demographic form.
Interviews were done either through home visits or via a
telephone call. Demographic data and health surveys were
administered via an e-mail link. Additionally, a paper-based
health survey was also administered. A paper format was
used instead of digital due to copyright restrictions for that
specific survey. Details of all health surveys are presented
in the Measures section. This article presents all health sur-
vey data that assessed the parents’ physical, sleep, mental,
and social health.
Sample and Setting
Approval for the study was obtained from the institutional
review board for human subjects research. Informed written
consent was obtained from all participants prior to data
collection.
Five cases were recruited through the palliative care pro-
gram of an academic tertiary medical center in the southeast
region of the United States. This palliative care program offers
a comprehensive, family-centered palliative and end-of-life
care for children with complex chronic or life-threatening con-
ditions. Parents for additional 3 cases were recruited by word of
mouth through nursing colleagues who personally knew the
bereaved parents. While an attempt was made to recruit using
support groups (Compassionate Friends, Kinder Mourn, and
Kids Path), only 1 parent responded to a letter sent by the
support group coordinator on behalf of the Principal Investiga-
tor, but was unreachable for subsequent follow-up.. Thus, no
participants were recruited using this method. Finally, a total of
8 cases were recruited within 9 months.
Eligible participants were parents 18 years old who spoke
and read English and had experienced the death of a child
below 12 years within the last 2 months from a life-
Dias et al
threatening illness. Parents were excluded if they were <18
years of age at the time of their child’s death because of dif-
ferent developmental needs of adolescent parents. Parents
whose child died following a traumatic injury were also
excluded, because traumatic deaths are associated with a higher
risk of complicated parental grief given the guilt associated
with parents’ actual or perceived role in the child’s death.27,28
In addition, parents who experienced a fetal loss were excluded
because of key differences in the grief needs of parents who
experience fetal loss.29,30
We used a convenience sampling strategy to recruit eligible
parents. Parents who met the study eligibility criteria were
contacted via either a letter or phone call from a palliative care
team member. The letter included a toll-free opt-out telephone
number. To further assess parents’ willingness to participate in
the study, parents who did not opt out were contacted by phone
by the first author, 2 to 3 weeks after the letter was sent. Simi-
larly, for parents who received phone calls by the palliative
team member, their willingness to participate in the study was
assessed during that phone call. Parents who were recruited by
word of mouth were contacted by an individual known to them
to determine their willingness to participate in the study. Par-
ents who expressed their willingness to participate or learn
more about the study were then contacted by the first author
via a phone call to provide details of the study, obtain a verbal
assent to send study material including consent form for their
review, and obtain a potential date for data collection. To
obtain child demographics and confirm study eligibility, verbal
assent was obtained from parents who were recruited through
the palliative care program to view their child’s medical
record about illness-specific data such as diagnoses, complica-
tions, and length of illness. For parents recruited via word of
mouth, information about the child’s illness was obtained from
the parents.
A total of 17 eligible parent dyads were contacted. Among
these, 8 parent dyads (8 mothers and 8 fathers) agreed to par-
ticipate, enrolled in the study, and completed data collection at
both time points. The recruitment rate was 47%, with a 100%
retention rate. Of the 9 parent dyads who did not participate, 7
were not reachable or did not respond to the telephone call. Of
the remaining 2, 1 parent said that it would be too stressful to
participate and another stated that the spouse was not willing to
participate.
Measures
Demographic data included information about parent’s age,
gender, race, and income.
Self-reported medical history included illness diagnosis (eg,
diabetes or hypertension) before and 6 months after their
child’s death, hospital admissions and emergency department
visits, smoking and alcohol use, and use of bereavement
services.
Self-Health Perception Scale (SHPS): Parent’s health was
assessed based on a visual line scale 0 to 10, with 0 indicating
worst health and 10 the best health possible. Parents identified
3
a number, which best depicted their perception of their current
overall health.
Validated Health Surveys
Patient-Reported Outcomes Measurement Information System
was used to measure parent’s global, sleep, and social health,
and the Brief Symptom Inventory 18 (BSI-18) was used to
assess their physical, sleep, mental, and social health.
Patient-Reported Outcomes Measurement Information Sys-
tem: All PROMIS surveys have established psychometric
properties (Buysse et al, 2010; Castel et al, ref).31,32 Patient-
Reported Outcomes Measurement Information System–global
(version: 1.0) is a 10-item measure that assesses overall phys-
ical and mental health, including pain and fatigue. All PROMIS
scale items are rated on a 5-point Likert scale. According to the
PROMIS scoring manual (2014), the social and sleep scales, T-
score of 50 is the average of the calibrated sample of people
with chronic illnesses and the T-score is provided with an error
term (standard error), and for the PROMIS–global, a T-score of
50 is the average for the general population in the United
States. All these scales have a standard deviation (SD) of 10.
A higher PROMIS T-score indicates better health when com-
pared to the general population. The PROMIS–social (version:
8a) assesses the individual’s perceived ability to participate in
social activities. The PROMIS–sleep impairment scale (ver-
sion-8a) includes 8 items that assess quality and ability to sleep.
The BSI is an 18-item tool and is a multidimensional mea-
sure of mental distress that assesses specific psychiatric illness
such as anxiety, depression, and somatization (Asner-Self,
Schreiber, and Marotta, 2006; Galdon et al, 2008) conditions
predominantly observed in this population.33,34 The Cronbach
a for BSI is between 0.71 and 0.84 (Merport and Recklitis,
2012) with modest predictive and discriminate validity (Mea-
chen, Hanks, Millis, and Rapport, 2008).35,36 For the BSI-18,
the paper version of the answer sheet was used to record the
value (0-4) for each question in the corresponding line. The
lines were arranged in 3 columns to correspond to the 3 BSI
subscales: (1) somatization, (2) depression, and (3) anxiety.
The total values of each column were added to obtain raw
scores for each subscale, and the Global Severity Index raw
score was obtained by adding the sum of all subscales. The raw
scores of each participant were plotted on the profile for com-
munity norms based on their gender (father ¼ male profile,
mother ¼ female profile) to obtain the T-scores.
Data Collection
Data collection included both interview data and health survey
data and occurred at 3 and 6 months after a child’s death either
in person at their home or via a telephone call. All parents
completed health surveys prior to their qualitative interview.
Parents who met in person with the first author first completed
the PROMIS surveys on an electronic device (iPAD or laptop)
connected to parent-specific REDCap electronic survey link or
the BSI-18 (paper version). The SHPS was completed after all
4 American Journal of Hospice & Palliative Medicine® XX(X)

Table 1. Parent and Child Demographics.

Dyad Parent Age Race Child’s Age, Diagnosis Living Children Household Income Education Job Status

1 Mother 36 White 7 years, congenital illness 2 30 000-60 000 Associate degree Home maker
Father 39 White 2 Bachelors Employed
2 Mother 35 Asian 2 months, prematurity 1 60 000-100 000 Masters Out of work
Father 35 Asian 1 Masters Employed
3 Mother 33 White 9 months, cancer 1 Masters Employed
Father 39 White 1 >100 000 Masters Employed
4 Mother 30 Asian 6 months, congenital illness 0 Masters Home maker
Father 33 Asian 0 >100 000 Masters Employed
5 Mother 37 White 5.7 years, cancer 1 Masters Employed
Father 49 White 1 >100 000 <2 years of college Employed
6 Mother 35 White 1 week, congenital illness 1 30 000-60 000 Professional degree Home maker
Father 38 White 1 Bachelors Employed
7 Mother 38 White 2.6 years, congenital illness 1 60 000-100 000 Bachelors Employed
Father 41 White 2 > 100 000 Bachelors Employed
8 Mother 37 Black 11.9 years, cancer 1 60 000-100 000 <2 years of college Employed
Father 33 Black 1 <2 years of college Employed

the health surveys were completed. The BSI-18 and the SHPS
responses were then transferred to the REDCap database after
the home visit was completed.
For parents who had telephone interviews, PROMIS health
survey REDCap link was sent via e-mail. On the scheduled
interview day, parents first completed the PROMIS health sur-
veys, followed by the BSI-18 paper version and SHPS. Parents
then orally reported responses to the BSI-18 and SHPS for
entry into the REDCap database.
At the second time point (6 months), the REDCap link was
e-mailed a week prior to the scheduled interview to allow
parents to complete the survey prior to the interview. If parents
did not complete the measures prior to the scheduled interview,
they were given time before their interview to complete the
survey. No identifiable data were recorded on the paper ver-
sions to mainatian particpant privacy.
Data Analysis
Both PROMIS and BSI-18 health survey raw scores were con-
verted to T-scores based on the T-score conversion table pro-
vided for each scale. The differences between each family’s
father’s and mother’s score were calculated and summarized at
each time point. Each parent’s change from 3 to 6 months was
also calculated and summarized separately for each gender.
Cliff delta, which ranges from 1 to þ1, was used as a non-
parametric measure of effect size. A positive Cliff delta
indicated that the father’s score was more often larger than
the mother’s score, and a negative Cliff delta indicated the
mother’s score was more often larger than the father’s score.
The difference between the 6 months to 3 months change in the
father’s score and the change in the mother’s score was also
calculated and summarized. Using the change data for father
and mother, Cliff delta was calculated as the measure of effect
size for the difference in changes.
For the SHPS, individual parent scores at each time point
and a difference from 3 months to 6 months are summarized.
Additionally, the nonparametric Spearman rank correlation
coefficient (r) was used to estimate the associations between
the self-health perception with the PROMIS and BSI subscales.
Data summaries and correlation coefficients were generated
using SAS/STAT software (version 9.4 of the SAS System for
Windows; Copyright # 2002-2012; SAS Institute Inc, Cary,
North Carolina). The SAS and all other SAS Institute Inc prod-
uct or service names are registered trademarks or trademarks of
SAS Institute Inc. Cliff delta and the 95% confidence intervals
were calculated using R version 3.1.3 (2015-03-09; Copyright
# 2015).
Results
Table 1 shows the participant and child demographics. Overall,
parents were within the ages of 30 to 50 and had some racial
diversity including Caucasians (n ¼ 10), African Americans
(n ¼ 2), and Asians (n ¼ 4). The children’s age ranged from 48
hours after birth up to 12 years of age at the time of death. All
children (n ¼ 8) were diagnosed with a life-threatening illness
including cancer (n ¼ 3), congenital illness (n ¼ 4), and pre-
maturity (n ¼ 1). All the parents had at least 1 year of college
education, and 6 of the 8 parent dyads had a combined house-
hold income of 60 000 or greater. Most of the parents (n ¼ 13)
were employed with no change in job status at the 3- and
6-month time points. Two mothers were homemakers and
1 mother who was employed prior to her child’s death was not
working up to 6 months after her child’s death. At the 6-month
interview, this mother stated that she was soon going to take a
new job. Most of the parents had at least 1 living child (n ¼ 14).
Self-Reported Medical History
Based on the self-reported medical history data, 4 (25%) of the
16 parents (8 mothers and 8 fathers) had new diagnosis within
the 6-month period after their child’s death. One mother and 1
father reported prediabetes and anxiety as new diagnoses at the
Dias et al 5

3-month time point and 2 mothers reported anxiety and sleep

10 3 Months disorder as new diagnoses at the 6-month time point. One father
9 6 Months was referred to a psychiatrist for further evaluation, but was not
8 medically diagnosed with any condition. Self-reported smok-
7 ing was unchanged, but 2 parents reported an increase in alco-
6 hol consumption after their child’s death. The mother
5 diagnosed with sleep disorder reported that she was taking
4 medications to aid her sleep.
3 Self-Health Perception Scale: Most parents (n ¼ 14, 87.5%)
2 rated their overall health at 7 or 8 (see Figure 1) at both time
1 points. While 43.75% of the parents (mothers ¼ 4, fathers ¼ 3)
0
had an improved score at 6 months, the scores of more fathers
Score than mothers were reduced (n ¼ 4, 25%; mothers ¼ 1, fathers ¼
Mother_1

Mother_2

Mother_3

Mother_4

Mother_5

Mother_6

Mother_7

Mother_8
Father_1

Father_2

Father_3

Father_4

Father_5

Father_6

Father_7

Father_8
3) and more mothers than fathers had stable scores at both time
points (n ¼ 5, 31.25%; mothers ¼ 3, fathers ¼ 2).

Figure 1. Parents self-health perception scores.

Health Surveys
Table 2 summarizes the descriptive statistics of the PROMIS
and BSI-18 health surveys for the mothers and fathers and the

Figure 2. Parents Patient-Reported Outcomes Measurement Information System (PROMIS) score means at 3 and 6 months.
6 American Journal of Hospice & Palliative Medicine® XX(X)

Table 2. Summary of Parents Health Survey Scores and the Score Changes From 3 to 6 Months.

Scores Change From T1 to T2

Scale Parent Gender Time Mean (SD) Median (Min, Max) Mean (SD) Median (Min, Max)

PROMIS–physical T-score Mother T1 51.0 (4.2) 52.5 (42.3, 54.1)

T2 49.9 (4.5) 50.8 (42.3, 54.1) 1.2 (5.1) 0 (9.2, 8.5)
Father T1 50.4 (5.7) 50.9 (39.8, 57.7)
T2 50.5 (6.9) 49.3 (42.3, 61.9) 0.2 (5.2) 0.4 (6.4, 6.4)
PROMIS–mental T-score Mother T1 46.6 (6.1) 47.1 (36.3, 53.3)
T2 46.6 (5.7) 48.3 (38.8, 53.3) 0.0 (2.9) 1.2 (4.7, 2.5)
Father T1 44.1 (6.3) 45.8 (33.8, 53.3)
T2 45.4 (6.3) 44.7 (38.8, 59.0) 1.4 (4.0) 2.4 (4.7, 5.7)
PROMIS–sleep T-score Mother T1 45.9 (7.2) 47.2 (35.1, 56.1)
T2 48.7 (11.5) 52.0 (30.0, 60.3) 2.8 (6.5) 4.3 (5.3, 10.4)
Father T1 51.3 (7.5) 50.3 (41.4, 66.3)
T2 51.5 (7.4) 53.5 (41.4, 60.3) 0.2 (5.1) 1.5 (8.9, 5.6)
PROMIS–social T-score Mother T1 39.3 (6.8) 42.1 (29.7, 48.0)
T2 37.5 (5.6) 38.5 (25.9, 44.0) 1.8 (6.5) 3.4 (9.9, 8.8)
Father T1 35.3 (5.5) 36.5 (25.9, 43.0)
T2 38.0 (5.4) 38.9 (25.9, 44.0) 2.7 (6.6) 0.9 (5.4, 14.3)
BSI–somatization T-score Mother T1 50.8 (5.5) 50.0 (41.0, 59.0)
T2 48.1 (6.8) 49.0 (41.0, 59.0) 2.6 (6.3) 0 (14, 5)
Father T1 50.3 (7.3) 49.0 (42.0, 61.0)
T2 46.6 (7.7) 42.5 (42.0, 64.0) 3.6 (10.0) 4 (18, 16)
BSI–depression T-score Mother T1 56.1 (8.9) 57.5 (45.0, 70.0)
T2 55.1 (8.8) 56.0 (40.0, 67.0) 1.0 (6.5) 0.5 (11, 6)
Father T1 57.6 (9.0) 60.0 (45.0, 70.0)
T2 52.1 (11.4) 49.0 (41.0, 66.0) 5.5 (8.0) 5.5 (21, 4)
BSI–anxiety T-score Mother T1 52.3 (8.1) 52.5 (38.0, 66.0)
T2 50.0 (4.7) 48.0 (45.0, 59.0) 2.3 (7.2) 1.5 (14, 7)
Father T1 54.5 (8.0) 50.0 (48.0, 66.0)
T2 46.6 (5.7) 47.5 (39.0, 57.0) 7.9 (5.9) 9 (17, 0)
BSI–global score T-score Mother T1 54.1 (6.6) 53.0 (47.0, 62.0)
T2 51.6 (7.4) 51.0 (39.0, 63.0) 2.5 (6.3) 1 (14, 5)
Father T1 56.6 (7.0) 56.0 (48.0, 67.0)
T2 49.0 (10.6) 48.5 (36.0, 62.0) 7.6 (7.6) 4.5 (20, 1)
Self-Health Perception Scale T-score Mother T1 7.3 (0.5) 7.0 (7.0, 8.0)
T2 7.4 (1.1) 8.0 (5.0, 8.0) 0.1 (1.4) 0.5 (3, 1)
Father T1 7.6 (0.7) 8.0 (6.0, 8.0)
T2 7.5 (0.5) 7.5 (7.0, 8.0) 0.1 (0.9) 0 (1, 1)
Abbreviations: BSI, Brief Symptom Inventory; PROMIS, Patient-Reported Outcomes Measurement Information System; SD, standard deviation.

score changes from T1 to T2. (See Figures 2 and 4 for box plots
of mothers’ and fathers’ health scores at 3 and 6 months) On the
PROMIS health measures, all parents exhibited health scores
below the population mean for at least 1 area of health status.
Social health was the most affected, with 100% of parents’
social health scores below the population mean; the lowest
score was 25.9, which is 3 SD below population mean (score
¼ 50, SD ¼ 10). Physical health scores were the least affected,
yet around 40% parents had scores below population means.
On the BSI-18, 1 mother had a subscale score for depres-
sion above the cutoff score of 63 at both 3 and 6 months and a
global score of 63 at 6 months, thus warranting further clinical
evaluation. Two mothers scored above 63 on at least 1 sub-
scale at 3 months. Regarding fathers, 1 scored above 63 on the
global score at 3 months and 1 scored above 63 on at least 1
subscale at both the time points. Most parent’s scores (mothers
¼ 43.75, fathers ¼ 37.5) did not indicate the need for further
mental health evaluation for specific disease diagnosis at either
time points.
Within-group comparisons (Figure 3) include interpreta-
tions of the direction of average within-gender changes.
Higher PROMIS scores indicate better health and higher BSI
scores indicate worse health, that is, an increase in PROMIS
scores and a decrease in BSI scores from time 1 to time 2
indicate an improvement in health. On average, mothers
showed improvement in PROMIS–sleep t-scores, BSI soma-
tization t-score, BSI depression t-score, BSI anxiety t-score,
and BSI global score t-score, though none of these improve-
ments were statistically significant as seen by the inclusion of
0 in the 95% confidence intervals. Mothers on average had
worsening PROMIS–physical t-score and the PROMIS–social
t-score. These confidence intervals also included 0 and
declines were not statistically significant changes as P values
were <.05. Fathers, on average, improved in the PROMIS–
Dias et al
Figure 3. Parents Brief Symptom Inventory 18 (BSI-18) score means at 3 and 6 months.
mental t-score, the PROMIS–social t score, the BSI somatiza-
tion t-score, the BSI depression t-score, the BSI t anxiety
score, and the BSI global score t-score. Only the 95% confi-
dence interval of the BSI global score t-score lies completely
below 0, indicating fathers’ BSI global score was a statisti-
cally significant improvement.
Between-group comparisons, Table 3 summarize the differ-
ence in scores between fathers and mothers, including the Cliff
delta and 95% confidence interval. All 95% confidence inter-
vals were 0, indicating the lack of a significant difference
between father’s and mother’s scores. Nevertheless, the point
estimate of the effect sizes can be used to design a larger study.
At 3 months, the PROMIS–physical, mental, social scores and
the BSI somatization subscale score had positive Cliff delta,
indicating that, in general, fathers’ scores were larger (ie, bet-
ter) than mothers’ scores. However, none of these differences
were statistically significant, since the corresponding 95% con-
fidence intervals included a 0. Correspondingly, at 6 months,
the PROMIS–mental, sleep scores, the BSI depression, anxiety
subscale scores, and the SHPS had positive Cliff delta.
7
Similar to the difference between fathers and mothers at 3
months, these effect sizes were not statistically significant at
the a ¼ .05 level. Spearman correlations between the health
surveys (PROMIS and BSI-18) and SHPS were not strong (see
Table 4) as the Spearman r coefficients were not closer to 1 or
1. Thus, parents’ self-assessment of their health did not
strongly associate with the objective health survey scores and
was not statistically significant.
Discussion
This is the first prospective study to evaluate bereaved parents’
physical, mental, and sleep health prospectively during the
most stressful time frame (first 6 months) after their child’s
death using standardized health surveys, which allow compar-
ison of resulting scores to the general US population health data
(T-scores). Previous research assessed bereaved parents’ health
during the first 13 months after their child’s death, evaluated
grief, mental health, and self-reported physical health.13,26 The
importance of prospective and early assessment (3 and
8 American Journal of Hospice & Palliative Medicine® XX(X)
Figure 4. Parents health score mean changes from 3 to 6 months.
6 months) in this study cannot be overestimated because it
allows for the early identification of potential health risks and
thus early intervention.
This study supports the findings of previous studies that
bereaved parents’ health is susceptible to illnesses. Based on
the self-reported medical history, it is significant to note that
out of the 16 participants in this study, 4 participants had new
diagnosis within the first 6 months. These findings underscore
that bereaved parents’ health is “at risk.”13 Additionally, the
self-health report data highlight other related health issues
such as increase in habits such as smoking or alcohol con-
sumption and sleep deprivation that can lead to chronic ill-
nesses and therefore deserve attention too. Thus, including
assessment of change in habits or behaviors would be a nec-
essary step toward health promotion and disease prevention.
Moreover, the lack of correlation between the objective health
surveys and parents’ self-health perception indicates that
bereaved parents may not be able to objectively assess their
health due to their focus on their grief.37 Consequently, par-
ents may delay accessing the health-care system until pre-
sented with specific health condition, which emphasizes the
need for the health-care system to be proactive and invest in
health promotion and preventive services for bereaved
parents.
Consistent with other studies, our results indicate bereaved
parents have at least 1 health concern during the first 6 months
following their child’s death. The sleep and social health data
from this study confirm findings from a qualitative study,
which identified bedtime was the most difficult time for
bereaved parents and that parents’ social interactions with fam-
ily, friends, and community in general has negatively
affected.25 Additionally, the potential lack of sleep38 and
decline in the quality of social relationships39 may be precur-
sors that affect other health outcomes including physical and
mental health. Comparisons of 3 and 6 months’ health scores
identified that mothers had trends toward worsening in their
physical and social health, also indicating a need for further
examination of the relationship of social health to parents’
physical health.
Comparison between mothers’ and fathers’ health concurred
with prior studies, in which fathers fared better than mothers at
6 months.19 Although the Cliff delta (an indication of effect sizes)
on the PROMIS–sleep scale, the BSI scale, and the SHPS were all
positive, indicating that the change in the father’s score was, in
general, larger than the change in the mother’s scores. Although
these scores were not statistically significant, at the a ¼ .05 level,
these estimated effect sizes will be particularly useful for inform-
ing the sample size and power calculation for a future study.
Dias et al 9

Table 3. Summary of the Difference in Scores Between Fathers and Mothers.

Difference Between Father and Mother T1 to T2 Change Between Father and Mother

Cliff Median Cliff

Scale Time Mean (SD) Median (Min, Max) Delta 95% CI Mean (SD) (Min, Max) Delta 95% CI

PROMIS–physical T1 0.6 (6.2) 1.6 (7.9, 11.8) 0.268 (0.623 to 0.856) 1.3 (8.6) 1 (11.8, 12.8) 0.125 (0.868 to 0.790)
T-score T2 0.7 (6.9) 0 (6.4, 12.8) 0.018 (0.756 to 0.740)
PROMIS–mental T1 2.5 (4.3) 2.4 (9.7, 4.7) 0.678 (0.258 to 0.958) 1.3 (5.8) 1.2 (7.2, 9.7) 0.607 (0.973 to 0.626)
T-score T2 1.1 (6.7) 0 (12.2, 8.2) 0.196 (0.614 to 0.805)
PROMIS–sleep T1 5.3 (8.4) 8.7 (7.5, 16.4) 0.857 (0.998 to 0.657) 2.6 (9.1) 2.7 (6.4, 7.8) 0.411 (0.669 to 0.933)
T-score T2 2.8 (11.5) 1.1 (7.5, 24.2) 0.161 (0.646 to 0.798)
PROMIS–social T1 4.0 (8.4) 5.1 (17.1, 6.6) 0.661 (0.608 to 0.980) 4.5 (10.5) 7.9 (14.2, 7.1) 0.875 (0.999 to 0.762)
T-score T2 0.5 (7.9) 0 (12.6, 15.2) 0.089 (0.745 to 0.654)
BSI–somatization T1 0.5 (8.1) 1.5 (11, 11) 0.071 (0.716 to 0.779) 1.0 (11.8) 2 (23, 16) 0.464 (0.610 to 0.937)
T-score T2 1.5 (5.6) 0.5 (12, 5) 0.196 (0.823 to 0.645)
BSI–depression T1 1.5 (8.7) 3.5 (13, 13) 0.482 (0.915 to 0.468) 4.5 (11.3) 7 (19, 12) 0.607 (0.633 to 0.974)
T-score T2 3.0 (7.7) 4 (15, 8) 0.357 (0.558 to 0.880)
BSI–anxiety T- T1 2.3 (10.4) 4 (18, 14) 0.196 (0.845 to 0.685) 5.6 (9.2) 4.5 (22, 5) 0.964 (0.864 to 0.999)
score T2 3.4 (7.1) 3.5 (13, 9) 0.696 (0.494 to 0.978)
BSI–global score T1 2.5 (8.1) 2.5 (11, 17) 0.661 (0.970 to 0.469) 5.1 (8.8) 4 (18, 8) 0.946 (0.838 to 0.999)
T-score T2 2.6 (8.3) 0.5 (12, 8) 0.304 (0.529 to 0.838)
Self-Health T1 0.3 (0.8) 0.5 1, 1) 0.643 (0.971 to 0.526) 0.2 (1.4) 0 (2, 2) 0.286 (0.630 to 0.869)
Perception T2 0.1 (1.1) 0 (1, 2) 0.071 (0.733 to 0.793)
Scale T-score
Abbreviations: BSI, Brief Symptom Inventory; CI, confidence interval; PROMIS, Patient-Reported Outcomes Measurement Information System; SD, standard
deviation.

Table 4. Correlations of Health Survey With Self-Health Perception altruistic after their child’s death and are willing to tell their
Scores. story to help other bereaved parents.41 Most of the parents in
this study also acknowledged their altruistic motivations to
Outcome Time, months Spearman r P Value
share their stories with the intent to help other bereaved par-
BSI–anxiety 3 0.25 .3458 ents. Although support groups may be an alternative recruit-
6 0.23 .3883 ment venue, it also may provide access to a biased sample that
BSI–depression 3 0.10 .7028 is already reaching out for help. Therefore, using the hospital-
6 0.24 .3756 based palliative care program was intentional to reach out to
BSI–somatization 3 0.20 .4659 all parents and provide them with an opportunity to tell their
6 0.02 .9279
story and assess their health.
BSI–global health 3 0.23 .3846
6 0.19 .4731 Access to a hospitalized pediatric palliative care program,
PROMIS–mental 3 0.10 .7007 however, provides parents of terminally ill children an opportu-
6 0.25 .3415 nity to foster a relationship with a palliative team during the
PROMIS–physical 3 0.34 .1991 child’s illness. This relationship can potentially have a positive
6 0.45 .0783 effect on parent’s experience during the child’s end of life and
PROMIS–sleep 3 0.28 .2864 consequentially have a positive influence on their grief and
6 0.35 .1884
health outcomes after their child’s death.42 Ten of the 16 parti-
PROMIS–social 3 0.41 .1177
6 0.41 .1174 cipants in this study were recruited through the palliative care
program and possibly benefitted from the relationship developed
Abbreviations: BSI, Brief Symptom Inventory; PROMIS, Patient-Reported with the palliative care team during the child’s illness. Although
Outcomes Measurement Information System.
offering pediatric palliative care services by pediatric hospitals is
on the rise, there are still many facilities that do not offer these
Recruitment for this study was a long process, because it services.43 This population may have different predeath experi-
was difficult to reach out to parents immediately after their ences, leading to different postdeath grief and health outcomes.
child died. Health-care providers served as gatekeepers for this This study has both strengths and limitations. Strengths
sensitive group of people, which while well intended can hin- include the prospective and longitudinal design, early assess-
der obtaining the important information needed to understand ment at 3 months, use of validated health measures, 100% reten-
the needs, challenges, and health concerns of this population at tion, and comparisons between both fathers and mothers.
the most critical time.40 The 100% retention rate in this study Limitations include a small sample size; however, the case study
supports the assumption that bereaved parents become design is suitable for smaller sample size studies. In addition, the
10
sample was not representative of racial demographic. Despite
the limitations, this study provides foundational data about
bereaved parents’ health to guide future longitudinal studies that
address the limitation of a small sample size and a short time
frame (first 6 months). Additional objective data that assess the
trajectory of bereaved parents over longer time frames are essen-
tial to address concerns regarding bereaved parents’ health.
Conclusion
Our study provides essential baseline empirical data to confirm
that bereaved parents are indeed at risk of health problems. It is
imperative that bereaved parents are made aware of the poten-
tial health consequences and are provided with tools to monitor
their health. Bereaved parents’ health merits the attention of the
health-care system, specifically improvement in parental
bereavement care services, including early risk assessment and
interventions that target health promotion and prevention.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for
the research, authorship, and/or publication of this article: I received
grants from the following institutes, American Nursing Foundation
and National Association of Indian Nurses of America.
ORCID iD
Nancy Dias, PhD, RN, CNE http://orcid.org/0000-0001-5918-1457
References
1. Kochanek KD, Murphy SL, Xu J, Tejada-Vera B. Deaths: final
data for 2014. Natl Vital Stat Rep. 2016;65(4):1-122.
2. Rostila M, Saarela J, Kawachi I. The forgotten griever: a nation-
wide follow-up study of mortality subsequent to the death of a
sibling. Am J Epidemiol. 2012;176(4):338-346.
3. Van Ommeren M, Levav I. Mortality in parents after death of a
child. Lancet. 2003;361(9370):1747; author reply 1748.
4. Werthmann J, Smits LJ, Li J. Parental mortality rates in a western
country after the death of a child: assessment of the role of the
child’s sex. Gend Med. 2010;7(1):39-46.
5. Harper M, O’Connor RC, O’Carroll RE. Increased mortality in
parents bereaved in the first year of their child’s life. BMJ Support
Palliat Care. 2011;1(3):306-309.
6. Levav I, Friedlander Y, Kark JD, Peritz E. An epidemiologic
study of mortality among bereaved parents. N Engl J Med.
1988;319(8):457-461.
7. Barry LC, Kasl SV, Prigerson HG. Psychiatric disorders among
bereaved persons: the role of perceived circumstances of death and
preparedness for death. Am J Geriatr Psychiatry. 2002;10(4):447-457.
8. Lannen PK, Wolfe J, Prigerson HG, Onelov E, Kreicbergs UC.
Unresolved grief in a national sample of bereaved parents:
impaired mental and physical health 4 to 9 years later. J Clin
Oncol. 2008;26(36):5870-5876.
American Journal of Hospice & Palliative Medicine® XX(X)
9. Stroebe M, Schut H, Stroebe W. Health outcomes of bereavement.
Lancet. 2007;370(9603):1960-1973.
10. Li J, Hansen D, Mortensen PB, Olsen J. Myocardial infarction in
parents who lost a child: a nationwide prospective cohort study in
Denmark. Circulation. 2002;106(13):1634-1639.
11. Kreicbergs U, Valdimarsdóttir U, Onelöv UVE, Henter J, Stei-
neck G. Anxiety and depression in parents 4-9 years after the loss
of a child owing to a malignancy: a population-based follow-up.
Psychol Med. 2004;34(8):1431-1441.
12. Li J, Johansen C, Hansen D, Olsen J. Cancer incidence in parents
who lost a child: a nationwide study in Denmark. Cancer. 2002;
95(10):2237-2242.
13. Youngblut JM, Brooten D, Cantwell GP, del Moral T, Tota-
pally B. Parent health and functioning 13 months after infant
or child NICU/PICU death. Pediatrics. 2013;132(5):
e1295-1301.
14. Murphy SA, Johnson LC, Wu L, Fan JJ, Lohan J. Bereaved par-
ents’ outcomes 4 to 60 months after their children’s deaths by
accident, suicide, or homicide: a comparative study demonstrat-
ing differences. Death Stud. 2003;27(1):39-61.
15. Rogers CH, Floyd FJ, Seltzer MM, Greenberg J, Hong J. Long-
term effects of the death of a child on parents’ adjustment in
midlife. J Fam Psychol. 2008;22(2):203-211.
16. Chan CL, Wang CW, Ho AH, et al. Symptoms of posttraumatic
stress disorder and depression among bereaved and non-bereaved
survivors following the 2008 Sichuan earthquake. J Anxiety Dis-
ord. 2012;26(6):673-679.
17. Li J, Johnsen SP, Olsen J. Stroke in parents who lost a child: a
nationwide follow-up study in Denmark. Neuroepidemiology.
2003;22(3):211-216.
18. Li J, Johansen C, Olsen J. Cancer survival in parents who lost a
child: a nationwide study in Denmark. Br J Cancer. 2003;88(11):
1698-1701.
19. Li J, Laursen TM, Precht DH, Olsen J, Mortensen PB. Hospitali-
zation for mental illness among parents after the death of a child.
N Engl J Med. 2005;352(12):1190-1196.
20. Hendrickson KC. Morbidity, mortality, and parental grief: a
review of the literature on the relationship between the death of
a child and the subsequent health of parents. Palliat Support Care.
2009;7(1):109-119.
21. Rostila M, Saarela J, Kawachi I. Mortality in parents following
the death of a child: a nationwide follow-up study from Sweden. J
Epidemiol Community Health. 2011;66(10):927-933.
22. Levav I, Kohn R, Iscovich J, Abramson JH, Tsai WY, Vigdoro-
vich D. Cancer incidence and survival following bereavement.
Am J Public Health. 2000;90(10):1601-1607.
23. Birenbaum LK, Stewart BJ, Phillips DS. Health status of bereaved
parents. Nurs Res. 1996;45(2):105-109.
24. Murphy S, Shevlin M, Elklit A. Psychological consequences of
pregnancy loss and infant death in a sample of bereaved parents. J
Loss Trauma. 2014;19(1):56-69.
25. Dias N, Docherty S, Brandon D. Parental bereavement: looking
beyond grief. Death Stud. 2017;41(5):318-327.
26. Hawthorne DM, Youngblut JM, Brooten D. Parent spirituality, grief,
and mental health at 1 and 3 months after their infant’s/child’s death
in an intensive care unit. J Pediatr Nurs. 2016;31(1):73-80.
Dias et al
27. Kovarsky RS. Loneliness and disturbed grief: a comparison of
parents who lost a child to suicide or accidental death. Arch Psy-
chiatr Nurs. 1989;3(2):86-96.
28. Wijngaards-de Meij L, Stroebe M, Schut H, et al. Couples at risk
following the death of their child: predictors of grief versus
depression. J Consult Clin Psychol. 2005;73(4):617-623.
29. Lang A, Fleiszer AR, Duhamel F, Sword W, Gilbert KR, Corsini-
Munt S. Perinatal loss and parental grief: the challenge of ambiguity
and disenfranchised grief. Omega (Westport). 2011;63(2):183-196.
30. Rand CS, Kellner KR, Revak-Lutz R, Massey JK. Parental beha-
vior after perinatal death: twelve years of observations. J Psycho-
som Obstet Gynaecol. 1998;19(1):44-48.
31. Buysse DJ, Yu L, Moul DE, et al. Development and validation of
patient-reported outcome measures for sleep disturbance and
sleep-related impairments. Sleep. 2010;33(6):781-792.
32. Castel LD, Williams KA, Bosworth HB, et al. Content validity in
the promis social health domain: a qualitative analysis of focus
group data. Quality of Life Research?: An International Journal
of Quality of Life Aspects of Treatment, Care and Rehabilitation.
2008;17(5):737-749. http://doi.org/10.1007/s11136-008-9352-3.
33. Asner-Self KK, Schreiber JB, Marotta SA. A cross-cultural anal-
ysis of the brief symptom inventory-18. Cultural Diversity and
Ethnic Minority Psychology. 2006;12(2):367-375. http://dx.doi.
org/10.1037/1099-9809.12.2.367.
34. Galdón MJ1, Durá E, Andreu Y, et al. Psychometric properties of
the brief symptom inventory-18 in a spanish breast cancer sample.
J Psychosom Res. 2008;65(6):533-9. doi: 10.1016/j.jpsychores.
2008.05.009. Epub 2008 Oct 28.
11
35. Merport A1, Recklitis CJ Does the brief symptom inventory-18
case rule apply in adult survivors of childhood cancer? Compar-
ison with the symptom checklist-90. J Pediatr Psychol. 2012;
37(6):650-9. doi: 10.1093/jpepsy/jss050. Epub 2012 Mar 26.
36. Meachen SJ1, Hanks RA, Millis SR, et al. The reliability and
validity of the brief symptom inventory-18 in persons with trau-
matic brain injury. Arch Phys Med Rehabil. 2008;89(5):958-65.
doi: 10.1016/j.apmr.2007.12.028.
37. Gilmer MJ, Foster TL, Vannatta K, et al. Changes in parents after
the death of a child from cancer. J Pain Symptom Manage. 2012;
44(4):572-582.
38. Miles MS. Emotional symptoms and physical health in bereaved
parents. Nurs Res. 1985;34(2):76-81.
39. Song J, Floyd FJ, Seltzer MM, Greenberg JS, Hong J. Long-term
effects of child death on parents’ health-related quality of life: a
dyadic analysis. Fam Relat. 2010;59(3):269-282.
40. Butler AE, Hall H, Copnell B. Ethical and practical realities of
using letters for recruitment in bereavement research. Res Nurs
Health. 2017;40(4):372-377.
41. Gaab E. Altruism in families facing the death of a child: how do
they do it? J Paediatr Child Health. 2012;48(12):1056-1057.
42. van der Geest IM, Darlington AS, Streng IC, Michiels EM, Pieters
R, van den Heuvel-Eibrink MM. Parents’ experiences of pediatric
palliative care and the impact on long-term parental grief. J Pain
Symptom Manage. 2014;47(6):1043-1053.
43. Feudtner C, Womer J, Augustin R, et al. Pediatric palliative care
programs in children’s hospitals: a cross-sectional national
survey. Pediatrics. 2013;132(6):1063-1071.