doi:10.1111/jog.12732 J. Obstet. Gynaecol. Res. Vol. 41, No.

9: 13131325, September 2015

Accuracy of visual inspection with acetic acid and with

Lugols iodine for cervical cancer screening: Meta-analysis

Liang Qiao1, Bo Li1, Mei Long1, Xiao Wang1, Anrong Wang1 and Guonan Zhang2
Departments of 1Cancer Prevention and Treatment and 2Gynecologic Oncology, Sichuan Cancer Hospital and Institute and
Sichuan Cancer Prevention and Treatment Center, Chengdu, China

Abstract
The aim of this review was to provide an updated summary estimation of the accuracy of visual inspection with
acetic acid (VIA) and with Lugols iodine (VILI) in detecting cervical cancer and precancer. Studies on VIA/VILI
accuracy were eligible in which VIA/VILI was performed on asymptomatic women who all underwent conr-
matory testing of histology, combination of colposcopy and histology, or combination of multiple screening tests,
colposcopy and histology, to detect cervical intraepithelial neoplasia grade 2 or worse (CIN2+ or CIN3+). A bivar-
iate model was tted to estimate the accuracy of VIA/VILI and provide estimates of heterogeneity. Subgroup
analysis was used to investigate the source of heterogeneity. A total of 29 studies on VIA and 19 studies on VILI
were included nally in the meta-analysis. The summary sensitivity and specicity of VIA for CIN2+ were 73.2%
(95%CI: 66.580.0%) and 86.7% (95%CI: 82.990.4%), respectively, and those for VILI were 88.1% (95%CI: 81.5
94.7%) and 85.9% (95%CI: 81.790.0%), respectively. VIA and VILI were both more sensitive in detecting more se-
vere outcome, although there was a slight loss in specicity. Apparent heterogeneity existed in sensitivity and
specicity for both VIA and VILI. High sensitivity of both VIA and VILI for CIN2+ was found when a combina-
tion of colposcopy and histology was used as disease conrmation. VIA, VILI, even a combination of them in par-
allel, could be good options for cervical screening in low-resource settings. Signicant differences in sensitivity
between different gold standards might provide a proxy for optimization of ongoing cervical cancer screening
programs.
Key words: bivariate model, cervical cancer, meta-analysis, screening, visual inspection.

Introduction and that after the application of Lugols iodine (VILI),

Cervical cancer is the fourth most common cancer in
women, and the seventh overall, with an estimated
528 000 new cases and 266 000 deaths in 2012. A total of
84% of new cases and 86% of deaths occurred in the less
developed regions.1 Regularly repeated Pap smear
screening linked with treatment has prevented develop-
ment of cervical cancer in millions of women in the de-
veloped countries, but not in most developing
countries because of the lack of well-equipped and
resourced health-care services.2,3 Visual inspection of
the cervix after application of 35% acetic acid (VIA)
have been widely evaluated as alternative screening
tests in many developing countries, because they are
simple, inexpensive and feasible. Due to the drawbacks
of subjectivity and reliance on detection of the transfor-
mation zone, there is wide variability in the reported ac-
curacy of these visual inspection methods,4,5 therefore a
meta-analysis of VIA and VILI test qualities and a corre-
sponding heterogeneity investigation would be useful.
Several system reviews and meta-analyses have been
conducted previously,68 but in most of them, the true
disease verication was done using biopsy taken only
from colposcopically suspect lesions. Those studies

Received: November 6 2014.

Accepted: March 14 2015.
Reprint request to: Professor Guonan Zhang, Department of Gynecologic Oncology, Sichuan Cancer Hospital and Institute and Sichuan
Cancer Prevention and Treatment Center. No. 55, Section 4, Peoples South Road. Chengdu 610041, China. Email: zhanggn@hotmail.com

2015 Japan Society of Obstetrics and Gynecology 1313

L. Qiao et al.
using a reference standard of multiple screening tests (in-
cluding HPV test) and blind biopsy even in the absence
of colposcopically detected lesions were not included.
In addition, the statistical approaches used in most of
the previous meta-analyses did not take into account
within-study sampling error and additional unexplained
heterogeneity between studies. I2 statistic was widely
used to measure the extent of heterogeneity, which was
not recommended by the Cochrane Diagnostic Test
Accuracy (DTA) Working Group because it did not
account for heterogeneity explained by phenomena such
as positivity threshold effects, and will overestimate the
degree of heterogeneity observed.9,10 Thus, we con-
ducted a meta-analysis using a statistically rigorous hier-
archical model and based upon more comprehensive
studies with diverse reference standards for true disease
verication to provide updated average estimates of the
accuracy of VIA and VILI, respectively, in detecting cer-
vical intraepithelial neoplasia grade 2 or worse (CIN2+)
or CIN3 or worse (CIN3+) in asymptomatic women, as
well as estimates of the heterogeneity, and investigated
the source of heterogeneity.
Methods
Databases including PubMed, EMBASE, Cochrane
Library, China National Knowledge Infrastructure
(CNKI), VIP and WANFANG up to December 2013
were searched using the following: ((((visual inspec-
tion) AND (acetic acid)) OR VIA) OR (((visual
inspection) AND (Lugols iodine)) OR VILI)) AND
(cervix dysplasia OR cervix neoplasms OR cervical
intraepithelial neoplasia OR cervical neoplasia OR
cervical cancer OR cervical carcinoma OR CIN) AND
(screening). The reference lists of those retrieved
articles were then checked to obtain relevant studies
not identied in the database search.
The study selection criteria for eligibility were as fol-
lows: (i) a focus on the accuracy of VIA or VILI testing
in apparently healthy, asymptomatic women; (ii) refer-
ence standard histology alone, or a combination of col-
poscopy and histology (abnormal colposcopy should
have histologic conrmation), or combination of multi-
ple screening tests, colposcopy and histology (abnormal
colposcopy and/or positive on one or more screening
tests [e.g. liquid-based cytology, HPV etc.] should have
histologic conrmation); (iii) disease threshold CIN2 or
CIN3; (iv) verication of the screening test using the
reference standard so that sufcient data could be
obtained to complete all four cells of a 2 2 table. No
1314
language, publication date, or publication status restric-
tions were imposed.
Among studies based on the same population, the one
with the latest data, the most comprehensive and valid
analysis was eventually retained. Studies with missing or
unclear key information on test accuracy were excluded.
Any reviews, comments, letters were also excluded.
Two investigators independently screened each re-
cord. The title and abstract of each citation were screened
rst, and the full text was screened second.
The Cochrane version of Quality Assessment of Diag-
nostic Accuracy Studies (QUADAS), which includes 11
items, was used to assess the quality of each included
study.11 Each item related to a single aspect of quality
could be judged as yes, no, or unclear. The quality
score for yes was dened as 1, and that for no or un-
clear was dened as 0.
Two reviewers used a systematic review data extrac-
tion form with the following entries: rst authors name,
year of study publication, country, study period, age
range of the study population, study design, size of study
population, screener, place of screening, gold standard,
disease threshold, and QUADAS indicators. The number
of true positives (TP), false positives (FP), false negatives
(FN), and true negatives (TN) were also extracted from
each included study. The two reviewers reached consen-
sus each time there was a discrepancy in data collection.
If more than one study sample was included in a single
report and data for each sample were provided sepa-
rately, we treated the sample as if they had been pre-
sented in individual studies. Missing key information
on study design or test accuracy was requested from
the authors by email. If the authors did not respond, their
studies were excluded from the meta-analysis.
The summary sensitivity and specicity for VIA or
VILI were estimated directly using the bivariate model,
which was a hierarchical model recommended by the
Cochrane DTA Working Group.9 The bivariate approach
ts a two-level model with independent binomial distri-
butions for the TP and TN conditional on the sensitivity
and specicity in each study, and a bivariate normal
model for the logit transformations of sensitivity and
specicity between studies. Given that the sensitivity
and specicity in each study are assumed to have a bi-
variate normal distribution across studies, the possibility
of correlation between them can be incorporated.12
Threshold effects could be explored if the covariance be-
tween logit sensitivity and specicity was estimated to
be statistically signicantly negative, which represented
the trade-off in sensitivity and specicity as the test pos-
itivity threshold across studies varied.9
2015 Japan Society of Obstetrics and Gynecology

For the comparison of VIA and VILI, a binary covari-
ate for test type was included in the model based on all
available studies to investigate whether the expected
sensitivity and/or specicity differed between the tests.
Individual and summary estimates of sensitivity and
specicity and summary receiver operating characteris-
tic (SROC) curve were plotted on an SROC graph. The
95% condence region around the pooled estimates
was included, as was 95% prediction region, which gave
an indication of between-study heterogeneity.
Heterogeneity was evaluated statistically using the
variance of logit transformed sensitivity and specicity
and graphically by the prediction region in ROC space.
Where heterogeneity is high, the value of the corre-
sponding variance parameter is far away from 0, and
the 95% prediction region is much larger than the 95%
condence region.10
Study level covariates of gold standard, screener, place
of screening, size of study population, and the summary
QUADAS score were added to the hierarchical model in-
dividually to explore subgroup heterogeneity for VIA
and VILI. Size of study population was aggregated by
median. Summary QUADAS score was divided into
two categories of 11 and <11.
Forest plots and SROC plot were output by Review
Manager (RevMan), version 5.2. The bivariate model
was tted using Proc NLMIXED in SAS, version 9.2.
Results
A total of 1990 records were identied using the search
strategy after removing the duplicates, among which
1828 records were excluded on the basis of title or ab-
stract. Of the 162 full texts assessed, 143 were excluded
due to irrelevant content (n = 15), inappropriate popula-
tion source (n = 18), inappropriate reference standard
(n = 1), partial verication (n = 34), repetitive analysis
or publication based on the same database (n = 17), miss-
ing or unclear key information on test accuracy (n = 8),
and other literature types such as review, comment, let-
ter and so on (n = 50). Finally, 19 articles were included
in the quantitative synthesis. Of the 19 articles, one re-
ported results respectively from 11 separate centers,
and one reported two sets of results for test performed
by doctor and nurse, respectively, in the same popula-
tion. We treated those separate results as if they were ob-
tained from different studies, therefore 29 studies on
VIA, and 19 studies on VILI were included in the meta-
analysis (Fig. 1). Finally, we performed meta-analysis of
studies in which VIA was used as the screening test
2015 Japan Society of Obstetrics and Gynecology
Accuracy of VIA/VILI for cervical cancer
(CIN2 threshold, n = 29; CIN3 threshold, n = 16); and
in which VILI was used as the screening test (CIN2
threshold, n = 19; CIN3 threshold, n = 12). The main
characteristics of all integrated studies for VIA and VILI
are listed in Table 1.
The forest plots show the variation of the sensitivity
and specicity of VIA and VILI to detect CIN2+ or
CIN3+ (Fig. 2). There appeared to be greater variability
in estimated sensitivity than specicity across studies.
Estimates of sensitivities of both tests for CIN2+ were
made with less certainty than those for CIN3+.
The sensitivity and specicity for all tests and out-
comes are summarized in Table 2. The summary sensi-
tivity and specicity of VIA for CIN2+ and CIN3+
were 73.2% (95%CI: 66.580.0%) and 86.7% (95%CI:
82.990.4%), 80.6 (95%CI: 73.887.4%) and 85.3 (95%CI:
81.389.3%), respectively, and those for VILI were
88.1% (95%CI: 81.594.7%) and 85.9% (95%CI: 81.7
90.0%), 92.4% (95%CI: 86.498.3%) and 84.7% (95%CI:
81.288.1%), respectively. The two tests were both more
sensitive in detecting more severe outcome, although
there was a slight loss in specicity. For the outcome
CIN2+, the summary sensitivity of VILI was statistically
signicantly higher than that of VIA (P = 0.003). In con-
trast, the overall specicity of VILI was not signicantly
different from that of VIA (P = 0.438). Similar differences
in comparison of the two tests for the outcome CIN3+
were also found.
The variance of logit-transformed estimates in Table 2
and the size of the prediction region on the SROC plot in
Figure 3 indicated that the magnitude of the heterogene-
ity was evident in both sensitivity and specicity for any
test and any outcome. The P-values of covariance
between logit sensitivity and specicity for all tests and
outcomes indicated that no threshold effect existed
(P > 0.05).
Subgroup analysis was performed for the outcome of
CIN2+. It showed a statistical improvement in sensitiv-
ity, without loss of specicity as the true disease
verication was done using a combination of colposcopy
and histology compared with histology alone or a com-
bination of multiple tests, colposcopy and histology for
both VIA and VILI testing. VIA performed by physician
and nurse had statistically signicantly less sensitivity
than that performed by the other three kinds of
screeners. The statistically signicant highest specicity
for VIA was noted in the setting of hospital and primary
health center, and the lowest in the setting of primary
health center alone. The number of women screened
and summary QUADAS score had no effect on the accu-
racy of VIA and VILI (Table 3).
1315
L. Qiao et al.
Discussion
Unlike those traditional non-hierarchical models used
extensively in previous meta-analyses of DTA,68 the
present bivariate model preserved the 2-D nature of the
original data and took into account both study size and
heterogeneity beyond chance between studies.32
The pooled VIA specicity of CIN2+ determined in
the present meta-analysis (86.7%) was higher than the
specicity reported by Sritipsukho and Thaweekul
(77.2%),8 but lower than that reported by Sauvaget
et al. (92%),7 and similar to those reported by Arbyn
et al. (84.7%)18 and Chen et al. (87%).6 In contrast, the
present pooled VIA results were less sensitive for CIN2
+ than those previously reported (73.2% vs 74.8%, 80%,
79.2% and 77%).68,18 The discrepancies observed among
meta-analyses are likely to be due to the differences in
statistical methods and the study inclusion criteria. Nev-
ertheless, comparing the conventional cytology results
reported in previous meta-analyses, in which the disease
threshold was also CIN2 and all women also underwent
conrmatory testing, the present VIA sensitivity for
CIN2+ was found to be higher, but had lower specicity
(59% and 93% for sensitivity and specicity, respectively,
1316
Figure 1 Flow Diagram of Study
Selection.
reported by Chen et al.6; 53% and 96% by Nanda et al.33).
The difference was also found in studies in which con-
ventional cytology was concurrently evaluated with
VIA.13,22,26 Considering the strengths such as affordabil-
ity, feasibility, and minimal infrastructure needed,4 there
is a high potential for VIA to be an alternative test to cy-
tology. In additional, VIA and VILI are both more sensi-
tive in detecting CIN3+ than CIN2+, and without
obvious loss in specicity, consistent with the Arbyn et al.
study.18 In the present meta-analysis, for any outcome,
VILI was more sensitive and as specic compared with
VIA, consistent with other published meta-analyses for
these two tests.6,18 This suggests that VILI could be used
as an adjunctive test to VIA with expectation of gain in
sensitivity and without apparent loss in specicity,
which is advisable when individual women seek screen-
ing or early diagnosis.4 Muwonge et al. also determined
the potential benets for the combined test compared
with VIA alone.34
There was wide variation in accuracy parameters of
VIA and VILI in the present individual studies, espe-
cially in sensitivity for outcome CIN2+, even after
restricting participant source, reference standard, disease
threshold, and verication extent. Heterogeneity was
2015 Japan Society of Obstetrics and Gynecology
 Table 1 Studies on cervical cancer screening using VIA and VILI
First author Pub. Country Period Age Study No. Screener Place of Gold Disease QUADAS
year range design Screened screening standard threshold score
(years)
VIA
University 1999 Zimbabwe 19951997 2555 Cross- 2130 Nurse Primary Colposcopy CIN2 11
of sectional health and histology
Zimbabwe/ center
JHPIEGO13
Belinson14 2001 China 1999 3545 Cross- 1997 Physician Field Histology CIN2/3 11
sectional clinic
Rodriguez- 2002 Mexico 1998 1945 Cross- 376 Physician Hospital Histology CIN2 9
Reyes15 sectional
Ngelangel16 2003 Philippines 19982001 2565 Cross- 3316 Physician Hospital Colposcopy CIN2 11
sectional and nurse and and histology
primary
health

2015 Japan Society of Obstetrics and Gynecology

center
Sangwa- 2006 Congo 20032004 30+ Cross- 1528 Nurse Primary Colposcopy CIN2 11
Lugoma17 sectional health and histology
center
Sangwa- 2006 Congo 20032004 30+ Cross- 1528 Physician Primary Colposcopy CIN2 10
Lugoma17 sectional health and histology
center
Arbyn18 2008 Mali 19992003 2564 Cross- 5552 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 Congo 19992003 2564 Cross- 6935 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 Guinea 19992003 2564 Cross- 8627 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Jaipur) 19992003 2564 Cross- 5786 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Kolkata 1) 19992003 2564 Cross- 5894 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Kolkata 2) 19992003 2564 Cross- 8080 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Mumbai) 19992003 2564 Cross- 4004 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Niamey) 19992003 2564 Cross- 2534 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 Burkina Fasso 1999-2003 25-64 Cross- 2051 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India (Trivandrum 1) 19992003 2564 Cross- 4457 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
(Continues)
Accuracy of VIA/VILI for cervical cancer

1317
 Table 1 (Continued)

1318
First author Pub. Country Period Age Study No. Screener Place of Gold Disease QUADAS
year range design Screened screening standard threshold score
(years)
L. Qiao et al.

Arbyn18 2008 India (Trivandrum 2) 19992003 2564 Cross- 4759 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Qiao19 2008 China 2007 3054 Cross- 2388 Nurse Hospital Multiple CIN2/3 11
sectional tests,
colposcopy
and histology
Li20 2008 China 2004 3049 Cross- 2432 NA NA Multiple CIN2 9
sectional tests,
colposcopy
and histology
Li21 2009 China 20042005 1529 Cross- 1819 Physician NA Multiple CIN2 10
sectional tests,
colposcopy
and histology
Murillo22 2010 Colombia 20072008 2559 Cross- 4957 Nurse NA Multiple CIN2 10
sectional tests,
colposcopy
and histology
Muwonge23 2010 Angola 20022006 2559 Cross- 8849 Nurse Hospital Colposcopy CIN2 11
sectional and and histology
primary
health
center
Ngoma24 2010 Tanzania 20022007 2559 Cross- 10374 Nurse Hospital Colposcopy CIN2 11
sectional and and histology
primary
health
center
Cremer25 2011 El Salvador 20072009 50+ Cross- 578 Physician Hospital Histology CIN2 9
sectional and nurse
Fei26 2011 China 2009 2565 Cross- 859 Physician NA Multiple CIN2/3 8
sectional tests,
colposcopy
and histology
Dasgupta27 2012 India 20062009 NA Cross- 4873 Physician Hospital Histology CIN2 10
sectional
Deodhar28 2012 India 20062007 3049 Cross- 5519 Nurse Field Multiple CIN2/3 11
sectional clinic tests,
colposcopy
and histology
(Continues)

2015 Japan Society of Obstetrics and Gynecology

 Table 1 (Continued)
First author Pub. Country Period Age Study No. Screener Place of Gold Disease QUADAS
year range design Screened screening standard threshold score
(years)
Hao29 2012 China 2011 2565 Cross- 890 NA NA Histology CIN2/3 10
sectional
Hu30 2012 China 2010 2855 Cross- 1100 NA Hospital Multiple CIN2 10
sectional tests,
colposcopy
and histology
VILI
Li31 2006 China 2004 1959 Cross- 734 Physician NA Multiple CIN2 11
sectional tests,
colposcopy
and histology
Sangwa- 2006 Congo 20032004 30+ Cross- 1528 Nurse Primary Colposcopy CIN2 10
Lugoma17 sectional health and histology
center

2015 Japan Society of Obstetrics and Gynecology

Sangwa- 2006 Congo 20032004 30+ Cross- 1528 Physician Primary Colposcopy CIN2 9
Lugoma17 sectional health and histology
center
Arbyn18 2008 Mali 19992003 2564 Cross- 5552 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 Congo 19992003 2564 Cross- 6934 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 Guinea 19992003 2564 Cross- 8627 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Jaipur) 19992003 2564 Cross- 5786 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Kolkata 2) 19992003 2564 Cross- 8076 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Mumbai) 19992003 2564 Cross- 4003 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Niamey) 19992003 2564 Cross- 2534 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 B-Fasso 19992003 2564 Cross- 2051 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Trivandrum 1) 1999-2003 25-64 Cross- 4457 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Arbyn18 2008 India(Trivandrum 2) 19992003 2564 Cross- 4759 Health Field Colposcopy CIN2/3 11
sectional worker clinic and histology
Li20 2008 China 2004 3049 Cross- 2432 NA NA Multiple CIN2 8
sectional tests,
(Continues)
Accuracy of VIA/VILI for cervical cancer

1319
1320
L. Qiao et al.

Table 1 (Continued)
First author Pub. Country Period Age Study No. Screener Place of Gold Disease QUADAS
year range design Screened screening standard threshold score
(years)
Colposcopy
and histology
Muwonge23 2010 Angola 20022006 2559 Cross- 8842 Nurse Hospital Colposcopy CIN2 11
sectional and and histology
primary
health
center
Ngoma24 2010 Tanzania 20022007 2559 Cross- 10367 Nurse Hospital Colposcopy CIN2 11
sectional and and histology
primary
health
center
Fei26 2011 China 2009 25-65 Cross- 859 Physician NA Multiple CIN2/3 8
sectional tests,
colposcopy
and histology
Deodhar28 2012 India 20062007 3049 Cross- 5519 Nurse Field Multiple CIN2/3 11
sectional clinic tests,
colposcopy
and histology
Hu30 2012 China 2010 2855 Cross- 1100 NA Hospital Multiple CIN2 9
sectional tests,
colposcopy
and histology
CIN, cervical intraepithelial neoplasia; NA, not available; QUADAS, quality assessment of diagnostic accuracy studies; VIA, visual inspection after application of acetic acid; VILI, visual
inspection after application of Lugols iodine.

2015 Japan Society of Obstetrics and Gynecology

 Accuracy of VIA/VILI for cervical cancer

Figure 2 Coupled forest plots for visual inspection after application of acetic acid (VIA) and visual inspection after application of
Lugols iodine (VILI) for detecting cervical intraepithelial neoplasia (CIN)2+ or CIN3+.

further conrmed by the variance parameters quantita-
tively and by the prediction ellipse graphically, in which
threshold effects did not exist (according to the test of co-
variance parameter). Considering the impact of subjec-
tivity on the accuracy of VIA and VILI, relative factors
were included to explore the apparent heterogeneity, in
which screener and place of screening were proxies for
screener competence, size of study population was con-
sidered as a proxy for accumulated experience of the as-
sessors, and gold standard was related to the strength of

2015 Japan Society of Obstetrics and Gynecology 1321

L. Qiao et al.

Table 2 Summary estimates of VIA and VILI for CIN2+ and CIN3+ using a bivariate random effects model
Test Outcome Sensitivity Variance logit Specicity Variance logit Covariance between logit
(95%CI) (sensitivity) (95%CI) (specicity) sensitivity and specicity
VIA CIN2+ 0.732 (0.6650.800) 0.702 0.867 (0.8290.904) 0.709 0.212
CIN3+ 0.806 (0.7380.874) 0.475 0.853 (0.8130.893) 0.356 0.143
VILI CIN2+ 0.881 (0.8150.947) 1.472 0.859 (0.8170.900) 0.495 0.048
CIN3+ 0.924 (0.8640.983) 1.279 0.847 (0.8120.881) 0.165 0.033
VIA vs VILI in sensitivity for CIN2+ (P = 0.003); VIA vs VILI in sensitivity for CIN3+ (P = 0.011). CIN, cervical intraepithelial neoplasia; VIA,
visual inspection after application of acetic acid; VILI, visual inspection after application of Lugols iodine.

Figure 3 Summary receiver operating characteristic curve of () visual inspection after application of acetic acid (VIA) and ()
visual inspection after application of Lugols iodine (VILI) for detection of underlying (a) cervical intraepithelial neoplasia
(CIN)2+ or (b) CIN3+. () Summary point; () 95% condence region; (- - -) 95% prediction region.

true disease verication. Also, summary QUADAS
score, representing the whole screening methodology
quality, was also included. We did not perform subgroup
analysis for the outcome of CIN3+, however, because of
the small number of included studies, and the homoge-
neity of the study level covariates among those studies.
The source of heterogeneity for the outcome CIN3+
may be investigated through individual level covariates.
Again, limited by the small number of included studies,
we could not include all the covariates simultaneously in
the hierarchical model for the outcome CIN2+. It was
possible to include only one covariate at a time to per-
form the subgroup analysis.
For the outcome CIN2+, on subgroup analysis, disease
conrmation using histology alone or a combination of
multiple tests, colposcopy and histology had the conser-
vative sensitivity of both VIA (61.3%) and VILI (63.5%)
testing compared with the combination of colposcopy
and histology (79.5% and 92.1%, respectively). Possible
explanations for the high sensitivity in disease verica-
tion for the combination of colposcopy and histology
could be the high correlation between the visual inspec-
tion methods and colposcopy. The sensitivity of visual
inspection methods may be overestimated if
colposcopically-directed biopsy and visual inspection
miss similar small lesions35 that could be identied on
histology alone or a combination of multiple tests, col-
poscopy and histology. Pretorius et al. also conrmed
that the sensitivity of colposcopy-directed biopsy for
CIN2+ was only 57% when multiple random biopsies
were taken from all tested women.36 Even though VIA
performed by both physician and nurse had a statisti-
cally signicantly lower sensitivity than when per-
formed by the other three kinds of screeners, of note,

1322 2015 Japan Society of Obstetrics and Gynecology

 Accuracy of VIA/VILI for cervical cancer

Table 3 Estimates of test accuracy for CIN2+

Covariates VIA VILI
Sensitivity (95%CI) Specicity (95%CI) Sensitivity (95%CI) Specicity (95%CI)
Gold standard
Histology/multiple tests, 0.613 (0.5000.726) 0.870 (0.8140.927) 0.635 (0.4250.845) 0.880 (0.8100.950)
colposcopy and histology
Colposcopy and histology 0.795 (0.7330.858) 0.864 (0.8140.913) 0.921 (0.8810.961) 0.851 (0.8010.901)
No. women screened
4000 0.673 (0.5620.768) 0.840 (0.7730.890) 0.807 (0.6230.914) 0.849 (0.7680.905)
>4000 0.777 (0.6880.846) 0.890 (0.8380.927) 0.912 (0.8320.956) 0.866 (0.8050.910)
Summary QUADAS score
11 0.762 (0.6820.827) 0.870 (0.8180.908) 0.906 (0.8320.950) 0.860 (0.8050.901)
11 0.657 (0.5140.776) 0.860 (0.7800.914) 0.758 (0.4970.909) 0.855 (0.7520.920)
Screener
Health worker 0.801 (0.7140.866) 0.862 (0.7940.911) 0.918 (0.8420.959) 0.855 (0.7920.901)
Nurse 0.721 (0.5890.823) 0.897 (0.8250.941) 0.893 (0.7270.964) 0.860 (0.7550.924)
Physician 0.739 (0.5890.848) 0.788 (0.6560.878) 0.777 (0.4630.934) 0.812 (0.6600.906)
Physician and nurse 0.298 (0.1030.611) 0.927 (0.7980.976)
Setting
Field clinic 0.784 (0.6920.854) 0.854 (0.8000.895) 0.901 (0.8300.944) 0.855 (0.7980.898)
Primary health center 0.821 (0.6150.929) 0.666 (0.4770.814) 0.839 (0.5330.960) 0.755 (0.5460.887)
Hospital 0.637 (0.4330.802) 0.868 (0.7800.924) 0.853 (0.3630.983) 0.846 (0.5910.954)
Hospital and primary health center 0.746 (0.5260.886) 0.954 (0.9050.979) 0.972 (0.8840.994) 0.900 (0.7780.958)
Histology/ multiple tests, colposcopy and histology vs colposcopy and histology in sensitivity for VIA (P = 0.0057). Histology/ multiple tests,
colposcopy and histology vs colposcopy and histology in sensitivity for VILI (P = 0.0014). Physician and nurse vs health worker in sensitivity for
VIA (P = 0.0028); physician and nurse vs nurse in sensitivity for VIA (P = 0.0162); physician and nurse vs physician in sensitivity for VIA
(P = 0.0138). Primary health center vs eld clinic in specicity for VIA (P = 0.018), primary health center vs hospital in specicity for VIA
(P = 0.0215), primary health center vs hospital and primary health center in specicity for VIA (P<0.001). Hospital and primary health center
vs Field clinic in specicity for VIA (P = 0.0062), Hospital and primary health center vs hospital in specicity for VIA (P = 0.0251). CIN, cervical
intraepithelial neoplasia; QUADAS, quality assessment of diagnostic accuracy studies; VIA, visual inspection after application of acetic acid; VILI,
visual inspection after application of Lugols iodine.

only two studies reported this kind of screener, and the
corresponding subgroup summary sensitivity had wide
condence interval (0.1030.611), so the conclusion that
there was signicant difference in sensitivities between
screeners should be made with caution. In contrast, there
was no statistically signicant difference in VIA and
VILI performed by health worker, nurse, and physician.
This suggests that trained health workers and nurses can
be effective alternative to physicians for cervical cancer
screening using VIA or VILI testing.37 In the present
study, VIA had the statistically signicantly highest
specicity in the setting of hospital and primary health
center, and the lowest in the primary health center alone.
Even though there was no statistically signicant differ-
ence in sensitivity of VIA between settings, the primary
health center had the highest and the setting of hospital
and primary health center had moderate sensitivity. It
seems that VIA performed in the former is more likely
to identify precancerous and cancerous lesions, and
conducted in the latter is more likely to minimize over-
treatment. Most of the covariates were not associated
with the performance of VILI, perhaps because the num-
ber of studies of VILI was lower than that of VIA.
Conclusion
VIA and VILI have the advantage of correctly identify-
ing cervical precancerous and cancerous lesions. Despite
lower specicity, the two tests, even a combination of
them in parallel, could be good options for cervical
screening in low-resource settings. Signicant difference
in sensitivities of VIA/VILI between different gold
standards might provide a proxy for optimization of
ongoing cervical cancer screening programs. VIA and
VILI need to be evaluated in more settings, and more
detailed information is required in order to explore the
source of heterogeneity.
Disclosure
None declared.

2015 Japan Society of Obstetrics and Gynecology 1323

L. Qiao et al.
References
1. Ferlay J, Soerjomataram I, Ervik M et al. GLOBOCAN 2012 v1.0,
Cancer incidence and mortality worldwide: IARC CancerBase
No. 11 [Internet]. Lyon: International Agency for Research on
Cancer, 2013. [Cited 26 May 2014.] Available from URL:
http://globocan.iarc.fr
2. Sankaranarayanan R, Budukh AM, Rajkumar R. Effective
screening programmes for cervical cancer in low- and middle-
income developing countries. Bull World Health Organ 2001;
79: 954962.
3. International Agency for Research on Cancer, World Health
Organization. IARC Handbooks on Cancer Prevention: Cervix
Cancer Screening. Lyon: International Agency for Research on
Cancer, 2005.
4. Sankaranarayanan R, Nessa A, Esmy PO, Dangou JM. Visual
inspection methods for cervical cancer prevention. Best Pract
Res Clin Obstet Gynaecol 2012; 26: 221232.
5. Sankaranarayanan R, Gafkin L, Jacob M, Sellors J, Robles S. A
critical assessment of screening methods for cervical neoplasia.
Int J Gynaecol Obstet 2005; 89: S4S12.
6. Chen C, Yang Z, Li Z, Li L. Accuracy of several cervical screen-
ing strategies for early detection of cervical cancer: A meta-
analysis. Int J Gynecol Cancer 2012; 22: 908921.
7. Sauvaget C, Fayette JM, Muwonge R, Wesley R,
Sankaranarayanan R. Accuracy of visual inspection with acetic
acid for cervical cancer screening. Int J Gynaecol Obstet 2011;
113: 1424.
8. Sritipsukho P, Thaweekul Y. Accuracy of visual inspection with
acetic acid (VIA) for cervical cancer screening: A systematic re-
view. J Med Assoc Thai 2010; 93: S254S261.
9. Macaskill P, Gatsonis C, Deeks JJ, Harbord RM, Takwoingi Y.
Chapter 10: Analysing and presenting results. In: Deeks JJ,
Bossuyt PM, Gatsonis C (eds) Cochrane Handbook for Systematic
Reviews of Diagnostic Test Accuracy, version 1.0. London:
Cochrane Collaboration, 2010. [Cited 23 December 2010.]
Available from URL: http://srdta.cochrane.org/
10. Bossuyt P, Davenport C, Deeks J, Hyde C, Leeang M, Scholten
R. Chapter 11: Interpreting results and drawing conclusions. In:
Deeks JJ, Bossuyt PM, Gatsonis C (eds) Cochrane Handbook for
Systematic Reviews of Diagnostic Test Accuracy, version 0.9.
London: Cochrane Collaboration, 2013. [Cited 13 December
2013.] Available from URL: http://srdta.cochrane.org/
11. Reitsma JB, Rutjes AWS, Whiting P, Vlassov VV, Leeang
MMG, Deeks JJ. Chapter 9: Assessing methodological quality.
In: Deeks JJ, Bossuyt PM, Gatsonis C (eds) Cochrane Handbook
for Systematic Reviews of Diagnostic Test Accuracy, version 1.0.0.
London: Cochrane Collaboration, 2009. [Cited 27 October
2009.] Available from URL: http://srdta.cochrane.org/
12. Chu H, Cole SR. Bivariate meta-analysis of sensitivity and spec-
icity with sparse data: A generalized linear mixed model ap-
proach. J Clin Epidemiol 2006; 59: 13311332.
13. University of Zimbabwe/JHPIEGO Cervical Cancer Project.
Visual inspection with acetic acid for cervical-cancer screening:
Test qualities in a primary-care setting. Lancet 1999; 353:
869873.
14. Belinson J, Qiao YL, Pretorius R et al. Shanxi Province Cervical
Cancer Screening Study: A cross-sectional comparative trial of
multiple techniques to detect cervical neoplasia. Gynecol Oncol
2001; 83: 439444.
1324
15. Rodriguez-Reyes ER, Cerda-Flores RM, Quinez-Prez JM,
Velsco-Rodrguez V, Corts-Gutirrez EI. Acetic acid test: A
promising screening test for early detection of cervical cancer.
Anal Quant Cytol Histol 2002; 24: 134136.
16. Ngelangel CA, Limson GM, Cordero CP, Abelardo AD, Avila
JM, Festin MR. Acetic-acid guided visual inspection vs.
cytology-based screening for cervical cancer in the Philippines.
Int J Gynaecol Obstet 2003; 83: 141150.
17. Sangwa-Lugoma G, Mahmud S, Nasr SH et al. Visual inspec-
tion as a cervical cancer screening method in a primary health
care setting in Africa. Int J Cancer 2006; 119: 13891395.
18. Arbyn M, Sankaranarayanan R, Muwonge R et al. Pooled anal-
ysis of the accuracy of ve cervical cancer screening tests
assessed in eleven studies in Africa and India. Int J Cancer
2008; 123: 153160.
19. Qiao YL, Sellors JW, Eder PS et al. A new HPV-DNA test for
cervical-cancer screening in developing regions: A cross-
sectional study of clinical accuracy in rural China. Lancet Oncol
2008; 9: 929936.
20. Li L, Li LY, Qiao ZQ et al. The value of visual inspection with
acetic acid and Lugols iodine in the screening of cervical cancer
in the rural areas of China. Shi Yong Ai Zheng Za Zhi 2008; 23:
599604 (in Chinese).
21. Li N, Shi JF, Franceschi S et al. Different cervical cancer screen-
ing approaches in a Chinese multicentre study. Br J Cancer
2009; 100: 532537.
22. Murillo R, Luna J, Gamboa O et al. Cervical cancer screening
with naked-eye visual inspection in Colombia. Int J Gynaecol
Obstet 2010; 109: 230234.
23. Muwonge R, Manuel Mda G, Filipe AP, Dumas JB, Frank MR,
Sankaranarayanan R. Visual screening for early detection of cer-
vical neoplasia in Angola. Int J Gynaecol Obstet 2010; 111: 6872.
24. Ngoma T, Muwonge R, Mwaiselage J, Kawegere J, Bukori P,
Sankaranarayanan R. Evaluation of cervical visual inspection
screening in Dar es Salaam, Tanzania. Int J Gynaecol Obstet
2010; 109: 100104.
25. Cremer M, Conlisk E, Maza M et al. Adequacy of visual inspec-
tion with acetic acid in women of advancing age. Int J Gynaecol
Obstet 2011; 113: 6871.
26. Fei HL, Cheng YF, Cheng XD et al. Evaluation of ve screening
methods for an early detection of cervical cancer and its precan-
cerous lesions in Zhejiang province. Zhonghua Yi Xue Za Zhi
2011; 91: 309312 (in Chinese).
27. Dasgupta S, Bhattacharya S. Is visual inspection with acetic acid
better than cervical cytology to screen women 40 years of age
for carcinoma cervix? A cross-sectional study on proportion of
screen-positive women (by VIA and cervical cytology) having
CIN II/III lesion on cervical biopsy: Difference between two
age groups and among screening methods. Arch Gynecol Obstet
2012; 285: 17311736.
28. Deodhar K, Sankaranarayanan R, Jayant K et al. Accuracy of
concurrent visual and cytology screening in detecting cervical
cancer precursors in rural India. Int J Cancer 2012; 131:
E954E962.
29. Hao YF. Comparison of three methods for screening cervical
cancer in Southeast Shandong Province. Zhongguo Re Dai Yi
Xue 2012; 12: 11101112 (in Chinese).
30. Hu YH. The Contrast Research pf Three Joint Screening Strategies
For Cervical Cancer. Nanchang: Nanchang University, 2012
(in Chinese).
2015 Japan Society of Obstetrics and Gynecology

31. Li N, Ma CP, Sun LX et al. Evaluation on the visual inspection
with Lugols iodine in cervical cancer screening program.
Zhonghua Liu Xing Bing Xue Za Zhi 2006; 27: 1518 (in Chinese).
32. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM,
Zwinderman AH. Bivariate analysis of sensitivity and specic-
ity produces informative summary measures in diagnostic re-
views. J Clin Epidemiol 2005; 58: 982990.
33. Nanda K, McCrory DC, Myers ER et al. Accuracy of the
Papanicolaou test in screening for and follow-up of cervical cy-
tologic abnormalities: A systematic review. Ann Intern Med
2000; 132: 810819.
34. Muwonge R, Walter SD, Wesley RS et al. Assessing the gain in
diagnostic performance when two visual inspection methods
2015 Japan Society of Obstetrics and Gynecology
Accuracy of VIA/VILI for cervical cancer
are combined for cervical cancer prevention. J Med Screen
2007; 14: 144150.
35. Pretorius RG, Kim RJ, Belinson JL, Elson P, Qiao YL. Ination of
sensitivity of cervical cancer screening tests secondary to corre-
lated error in colposcopy. J Low Genit Tract Dis 2006; 10: 59.
36. Pretorius RG, Zhang WH, Belinson JL et al. Colposcopically di-
rected biopsy, random cervical biopsy, and endocervical curet-
tage in the diagnosis of cervical intraepithelial neoplasia II or
worse. Am J Obstet Gynecol 2004; 191: 430434.
37. Sherigar B, Dalal A, Durdi G, Pujar Y, Dhumale H. Cervical
cancer screening by visual inspection with acetic acid: Interob-
server variability between nurse and physician. Asian Pac J
Cancer Prev 2010; 11: 619622.
1325