www.nature.

com/scientificreports

OPEN Comparative accuracy of cervical

cancer screening strategies
in healthy asymptomatic women:
a systematic review and network
meta‑analysis
Teruhiko Terasawa1*, Satoyo Hosono2, Seiju Sasaki3, Keika Hoshi4, Yuri Hamashima5,
Takafumi Katayama6 & Chisato Hamashima7

To compare all available accuracy data on screening strategies for identifying cervical intraepithelial
neoplasia grade ≥ 2 in healthy asymptomatic women, we performed a systematic review and network
meta-analysis. MEDLINE and EMBASE were searched up to October 2020 for paired-design studies
of cytology and testing for high-risk genotypes of human papillomavirus (hrHPV). The methods
used included a duplicate assessment of eligibility, double extraction of quantitative data, validity
assessment, random-effects network meta-analysis of test accuracy, and GRADE rating. Twenty-seven
prospective studies (185,269 subjects) were included. The combination of cytology (atypical squamous
cells of undetermined significance or higher grades) and hrHPV testing (excepting genotyping for
HPV 16 or 18 [HPV16/18]) with the either-positive criterion (OR rule) was the most sensitive/least
specific, whereas the same combination with the both-positive criterion (AND rule) was the most
specific/least sensitive. Compared with standalone cytology, non-HPV16/18 hrHPV assays were more
sensitive/less specific. Two algorithms proposed for primary cytological testing or primary hrHPV
testing were ranked in the middle as more sensitive/less specific than standalone cytology and the
AND rule combinations but more specific/less sensitive than standalone hrHPV testing and the OR rule
combination. Further research is needed to assess these results in population-relevant outcomes at
the program level.

Cervical cancer is the fourth most frequently diagnosed cancer and fourth most common cause of cancer-specific
mortality in women, with a worldwide estimated prevalence of 570,000 cases and 311,000 associated deaths
in ­20181,2. Observational studies have clearly demonstrated a reduction in the invasive cancer incidence and
mortality in well-organized screening programs using cervical cytological testing that have been ­implemented3.
Moreover, randomized controlled trials (RCTs) of well-screened populations have shown that strategies incor-
porating testing for high-risk human papillomavirus (hrHPV) subtypes, which are the central etiological agents
of cervical cancer ­pathogenesis4, were, in aggregate, associated with a reduction in the invasive cancer incidence
relative to that shown by cytological screening a­ lone5. Therefore, current guidelines recommend three primary
screening options: cytological testing alone, standalone hrHPV testing, and cytological + hrHPV combination
testing (co-testing)6–10. However, subsequent management strategies for women with positive primary testing
are complex. Although specific triage and/or follow-up testing algorithms for primary cytology and co-testing11

1
Section of General Internal Medicine, Department of Emergency and General Internal Medicine, Fujita Health
University School of Medicine, 1‑98 Dengakugakubo, Kutsukakecho, Toyoake, Aichi 470‑1192, Japan. 2Division of
Cancer Screening Assessment and Management, Center for Public Health Science, National Cancer Center, Tokyo,
Japan. 3Center for Preventive Medicine, St. Luke’s International Hospital Affiliated Clinic, Tokyo, Japan. 4Center
for Public Health Informatics, National Institute of Public Health, Wako, Japan. 5Department of Population Health
Science, Bristol Medical School, University of Bristol, Bristol, UK. 6Department of Statistics and Computer Science,
College of Nursing Art and Science, University of Hyogo, Hyogo, Japan. 7Department of Nursing, Faculty of
Medical Technology, Teikyo University, Tokyo, Japan. *email: terasawa@fujita-hu.ac.jp

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

and for primary hrHPV ­testing9 have been proposed, the evidence base to improve patient-important outcomes
with these algorithms is immature.
The comparative effectiveness of alternative screening strategies should be based on a comprehensive assess-
ment of benefits and harms. Given the low incidence and mortality due to cervical cancer in high-income
countries and the challenges associated with conducting de novo large and long-term RCTs, decision modeling
is an alternative realistic option to better understand the theoretical utility of the screening ­options12. In this
regard, comprehensive synthesis of the screening accuracy, a key model parameter of cytological and hrHPV
testing and their available combination algorithms reported in rigorously conducted paired-design studies, is a
valuable intermediate step. However, recent meta-analyses have focused on either standalone cytological and/
or hrHPV t­ esting13–15 or a comparison of cytological testing with a specific combination algorithm not proposed
­ nly16.
in guidelines o
For those studies that assessed the diagnostic accuracy of selected and different pairs of tests of interest and
their combination algorithms, network meta-analysis of diagnostic test accuracy studies is a useful approach
that can compare all the assessed tests and combination algorithms in a single a­ nalysis17. The current study
aimed to perform network meta-analysis to quantitatively compare and rank the cross-sectional accuracy of all
reported screening algorithms based on cytological and hrHPV testing. We specifically focused on the compara-
tive accuracy of guideline-proposed combination algorithms by examining data derived from primary studies
of healthy asymptomatic women that addressed verification bias because such bias is commonly observed in
cancer screening accuracy studies.

Methods
This extended systematic review is based on an update evidence review conducted for revision of the Japanese
Guidelines for Cervical Cancer S­ creening18,19. Although the complete evidence review was planned before analy-
sis, no protocol was registered for this extended review. This report followed PRISMA guidelines for diagnostic
test accuracy (PRISMA-DTA)20 and did not require ethics review or patient consent.

Search strategy.  We searched OVID MEDLINE and EMBASE for publications between January 1, 1992,
and October 14, 2020, with no language restrictions. The search strategies are detailed in the Supplementary
methods. Complementarily, the reference lists of eligible studies and relevant review articles were also screened
for other appropriate studies.

Study eligibility.  Three paired reviewers independently double screened the first 3000 abstracts in a cali-
bration phase. The same reviewers single screened the remaining abstracts. Two reviewers independently deter-
mined the eligibility of potential full-text articles, with discrepancies adjudicated by a third reviewer. Only fully
paired-design screening studies of cytology and hrHPV testing, either opportunistic or organized screening,
aimed at detecting cervical intraepithelial neoplasia ≥ grade 2 (CIN2+) in healthy asymptomatic women were
eligible for inclusion. We included all studies that performed either routine colposcopy-directed biopsy or col-
poscopy and selective biopsy in all screened women to verify target lesions along with studies that performed
either of the colposcopy methods among women with protocol-specified screening results and statistical cor-
rections for data from unverified samples. In studies that analyzed both eligible and ineligible populations, only
those with relevant and extractable data were included. In case of multiple publications, we included the publica-
tion with the largest sample size (see Supplementary methods for more details).

Data extraction.  One reviewer extracted descriptive data, which were independently confirmed by another
reviewer. Next, two reviewers independently extracted numerical data, with discrepancies resolved by consensus.
We preferred cross-tabulated count data over reported accuracy estimates when both data types were extractable
(see Supplementary methods for more details).

Operationalization.  Cytology results were standardized according to the Bethesda ­system21,22 if other clas-
sification systems had been used. For studies that used both conventional and liquid-based cytology tests (CC
and LBC, respectively), we favored LBC data over CC data; we jointly analyzed both smear preparation methods.
Operationally, hrHPV assays were categorized into four groups: hybridization with signal amplifications of
DNA (e.g., Hybrid Capture 2 [HC2], Qiagen, Gaithersburg, MD), polymerase chain reaction (PCR) of DNA
from ≥ 13 hrHPV genotypes, amplification of E6/E7 viral messenger RNA (mRNA), and assays identifying DNA
or RNA of genotypes, either HPV16 or HPV18 or both (HPV16/18)23. For mRNA-based genotyping, since the
genotype HPV45 was additionally targeted with HPV16 and HPV18 (HPV16/18/45), we adopted these results.
HC2 positivity was defined as ≥ 1.0 relative light units. We did not assess point-of-care testing platforms (e.g.,
careHPV, Qiagen, Gaithersburg, MD).
We operationally categorized combination tests as follows: (i) combination algorithms based on the OR rule
(women with either test positive were categorized as screening positive while women with both tests negative
as screening negative) or the AND rule (women with both tests positive were categorized as screening positive
while women with at least 1 negative test as negative)24; (ii) thresholds for cytological testing as, e.g., undeter-
mined significance or worse grades (≥ ASCUS), or low- or high-grade squamous intraepithelial lesions or worse
grades (≥ LSIL or ≥ HSIL, respectively); and (iii) hrHPV assays (Table 1)6–9,25. As cross-sectional representation of
guidelines-proposed algorithms, we assessed two specific strategies: “ ≥ LSIL OR [hrHPV AND ASCUS]”, which
classified only women with cytologic testing ≥ LSIL, or both by cytologic testing ASCUS and hrHPV testing
positive as screening positive; and “HPV16/18(/45) OR [hrHPV AND ≥ ASCUS]”, which classified only women

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 2

Vol:.(1234567890)
www.nature.com/scientificreports/

Methods FDA-approved systems

Standalone or component tests
Cytology
A specified cytological grade, ASCUS, ASCH, LSIL, or HSIL (and
SurePath (TriPath Imaging, Inc., Burlington, NC) and ThinPrep
 Conventional or liquid-based cytology higher grades if indicated) defined by the Bethesda system is
(Cytyc Corp, Marlborough, MA)
used as the positive criterion
hrHPV assays
Nucleic acid hybridization with signal amplification or cleavage- Hybrid Capture 2 (Qiagen, Gaithersburg, MD); Cervista HPV
 Signal amplifications
based signal amplification of ≥ 13 hrHPV DNA HR (Hologic, Madison, WI)
 PCR-based tests PCR of DNA from ≥ 13 hrHPV genotypes Cobas HPV test (Roche Molecular Systems, Pleasanton, CA)
 mRNA-based tests Amplification of E6/E7 viral mRNA Aptima (Hologic, Bedford, MA)
Cobas HPV test (Roche Molecular Systems, Pleasanton, CA);
Genotyping assays identifying DNA or RNA of HPV genotypes
 HPV 16/18(/45) Aptima HPV16 and HPV18/45 (Hologic, Bedford, MA); Cervista
either 16 or 18 or both
HPV 16/18 (Hologic, Madison, WI)
Specific algorithms Test operation system Definition of screening strategy positive
Combination tests
Women with both positive tests as screening positive and women
AND rule (e.g., hrHPV AND ≥ ASCUS) Co-testing or serial testing
with ≥ 1 negative test as screening negative
Women with ≥ 1 positive test as screening positive and women
OR rule (e.g., hrHPV OR ≥ ASCUS) Co-testing only
with both negative tests as screening negative
Co-testing or primary cytologic testing with reflex to hrHPV
Cytologic testing ≥ LSIL, or both by cytologic testing ASCUS and
 ≥ LSIL OR [hrHPV AND ASCUS] testing by a joint guideline published in 2012 among ACS,
hrHPV testing positive
ASCCP, and ASCP
HPV genotypes 16 or 18 positive, or both cytologic test-
Co-testing or primary hrHPV testing with HPV genotyping with
HPV16/18(/45) OR [hrHPV AND ≥ ASCUS] ing ≥ ASCUS and hrHPV testing positive for non-16/18 hrHPV
reflex to cytologic testing proposed by the SGO 2015
genotypes

Table 1.  Operational categorizations of cytological testing, assays for hrHPV testing, and their combination
algorithms. ACS American Cancer Society, ASCCP American Society for Colposcopy and Cervical Pathology,
ASCH atypical squamous cells, cannot exclude HSIL, ASCP American Society for Clinical Pathology, ASCUS
atypical squamous cells of undetermined significance, BPR both-positive rule, DNA deoxyribonucleic acid,
EPR either-positive rule, hrHPV high-risk human papillomavirus, HPV human papillomavirus, HR high risk,
HSIL high-grade squamous intraepithelial lesion, LSIL low-grade squamous intraepithelial lesion, mRNA
messenger ribonucleic acid, SGO Society for Gynecologic Oncology.

with HPV genotypes 16 or 18 (or 45) positive, or both cytologic testing ≥ ASCUS and hrHPV testing positive
for non-16/18(/45) hrHPV genotypes as screening positive (Table 1).

Quality assessment.  Paired independent reviewers double rated the validity of a study using a risk of bias
tool for comparative diagnostic accuracy studies (QUADAS-C)26, an extension to the existing Quality Assess-
ment of Diagnostic Accuracy Studies 2 ­tool27. Discrepancies were resolved via consensus. Operationally, a study
was defined to have low risk of verification bias only when all screened samples had been histologically verified.

Data synthesis and statistical analysis.  The primary outcome was sensitivity and specificity for detect-
ing CIN2+. We used their relative risk values for and absolute differences in (Δ) sensitivity and specificity for
any paired alternative screening algorithms (e.g., a standalone test vs. a combination algorithm) as measures of
comparative accuracy.
Between-study heterogeneity was assessed visually by using crosshair plots of sensitivity and specificity esti-
mates in the receiver operating characteristic (ROC) ­space28. We calculated the average sensitivity and specificity
estimates and their derived relative and Δ sensitivity and specificity values with their corresponding 95% credible
intervals (CrIs) by using an arm-based, two-stage hierarchical, Bayesian bivariate random-effects network meta-
analysis ­model29. Credible regions for the average estimates were constructed by using the standard m ­ ethod30.
For comparison, we also calculated average sensitivity and specificity estimates separately by using the standard
bivariate meta-analysis model for diagnostic ­accuracy31. Hierarchical summary ROC (HSROC) curves were
derived on the basis of the estimated p­ arameters32.
We performed study-level univariable meta-regression for the following prespecified binary predictors
when ≥ 10 studies were available: study location (countries ranked as “very high human development” by the
Human Development Index ­201733 vs. those that were not), study design (histology-based vs. colposcopy-based
verification), and type of sample collectors (physicians vs. nonphysicians). Scarce data on young individuals
(< 30 years old) precluded meta-regression based on age. Complete details of the methodology, model fitting,
choice of prior distributions for parameters assessed, and operational definitions used in sensitivity analyses are
provided in the Supplementary methods.
We used the Grading of Recommendation Assessment, Development, and Evaluation (GRADE) ­tool34 to
assess the certainty of evidence and focused on the comparisons among cytological testing (≥ ASCUS) alone,
standalone hrHPV assays, and the guideline-proposed combination algorithms. For calculating false negatives

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

First author and publication

year Country Enrollment year N Target age Cytologic test HPV assay
Histology-based studies
Belinson (2001)36–38 China 1999 1997 35–45 LBC HC2
Cárdenas-Turanzas (2008)39 USA; Canada 1998–2005 957  ≥ 30 CC HC2
Hovland (2010)40 Congo 2003 313 25–60 CC; LBC PCR
Histology-based correction studies
Schneider (2000)41 Germany 1996–1998 4761 18–70 CC PCR
Kulasingam (2002)42,43 USA 1997–2000 4075 18–50 LBC HC2
Bigras (2005)44 Switzerland 2002–2004 13,842 17–93 LBC HC2
Mayrand (2007)45,46 Canada 2002–2004 10,154 30–69 CC HC2
Li (2009)47 China 2004–2005 2562 15–59 LBC HC2
Castle (2011)48,49 USA 2008–2009 41,026 25–93 LBC HPV16/18
Mahmud (2012)50 Congo 2003–2004 1528  ≥ 30 CC HC2
Sangrajrang (2017)51,52 Thailand 2014–2015 5046 30–60 CC PCR; mRNA; HPV16/18
Kurokawa (2018)53 Japan 2015–2016 7585 25–69 LBC PCR; HPV16/18
Colposcopy-based studies
Blumenthal (2001)54,55 Zimbabwe 1995–1997 2073 25–55 CC HC2
Coste (2003)56,57 France 1999–2000 1324 18– CC; LBC HC2
Sankaranarayanan (2004)58 India 1999–2003 18,085 25–65 CC HC2
Qiao (2008)59 China 2007 2388 30–54 LBC HC2
McAdam (2010)60 Vanuatu 2006 494 30–50 CC HC2
Quincy (2012)61 Nicaragua ND 245 25–60 LBC HC2
Colposcopy-based correction studies
Cuzick (2003)62 UK 1998–2001 10,358 30–60 CC HC2
Petry (2003)63 Germany 1998–2000 7908  ≥ 30 CC HC2
Gravitt (2010)64 India 2005–2007 2331  ≥ 25 CC HC2
Moy (2010)65 China 2003–2006 9057 30–54 LBC HC2
Monsonego (2011)66 France 2008–2009 4429 20–65 LBC HC2; mRNA
Ferreccio (2013)67 Chile 2009–2010 8265 25–64 CC HC2
Agorastos (2015)68 Greece 2011–2013 3993 25–55 LBC PCR
Iftner (2015)69 Germany ND 9451 30–60 LBC HC2; mRNA; HPV16/18
Wu (2017)70 China 2015 11,064 21–65 LBC PCR

Table 2.  Study, participant, and screening test characteristics. CC conventional cytology, HC2 Hybrid
Capture 2, HPV human papillomavirus, LBC liquid-based cytology, mRNA messenger ribonucleic acid, PCR
polymerase chain reaction.

(FNs) and false positives (FPs), we assumed a healthy screening population of 1,000 women in which 20 are
CIN2 + (i.e., a prevalence of 2%)13.
We did not evaluate funnel-plot asymmetry because the required tests did not permit valid assessment of the
extent and impact of missing s­ tudies20. All analyses were performed by using WinBUGS 1.4.3 (MRC Biostatistics
Unit, Cambridge, UK) and Stata/SE 16.1 (Stata Corp, College Station, TX)35. We estimated the probability that
the true value (i.e., posterior distribution) of relative sensitivity or specificity was ≥ 1 (or ≤ 1) as a measure of
superiority of a test over a comparator test. A conventional, frequentist, two-tailed P-value of 0.05 corresponds
to a Bayesian posterior probability of 0.025, which we considered to be the threshold of statistical significance.

Results
Study selection.  Our literature search identified 15,488 citations, of which 27 prospective studies reported
in 35 publications corresponding to 185,269 women were included for the meta-analysis (Supplementary
Fig. S1)36–70. Supplementary material provides a list of excluded studies.

Characteristics of included studies.  All included studies had a prospective design, and 14 studies (52%)
were from high-income countries (Table 2). The average age of study participants ranged from 25 to 47 years.
Data on type of sample collectors was available for 20 studies (74%), with physician collectors in 14 studies
and nonphysician providers, typically trained nurses or midwives, in 6 studies. Thirteen studies had used only
CC, and 12 had adopted only LBC, whereas two other studies had used both CC and LBC (Table 2). Of the
four available hrHPV testing subgroups, HC2 was the most commonly reported hrHPV assay (assessed in 20
studies), whereas six studies assessed PCR-based tests, four genotyped for HPV16/18, and three used mRNA-
based tests, of which also genotyped for HPV16/18/45. Data on one or more combination algorithm(s) were
available in 19 studies (reported in 20 publications; 70%). The most commonly assessed combinations were

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 4

Vol:.(1234567890)
www.nature.com/scientificreports/

≥HSIL OR [HC2 AND ≥ASCUS]

≥HSIL
HC2 OR ≥HSIL
HC2 AND ≥HSIL ≥LSIL

≥LSIL OR [HC2 AND ASCUS] PCR

≥LSIL OR [PCR AND ASCUS]

HC2 AND ≥LSIL

HC2 OR ≥LSIL PCR OR ≥ASCUS

HC2 AND ≥ASCUS ≥ASCUS HPV16/18 OR

[PCR AND ≥ASCUS]

HC2 OR ≥ASCUS PCR AND ≥ASCUS

≥ASCH HPV16/18/45 OR
[mRNA AND ≥ASCUS]

HC2 mRNA AND ≥ASCUS

mRNA
mRNA OR ≥ASCUS
HPV16/18
HPV16/18 AND ≥ASCUS
HPV16/18 OR ≥ASCUS

Figure 1.  Network of eligible comparisons of cervical cancer screening algorithms. The line thickness is
proportional to the number of studies comparing the linked pair of screening algorithms. The size of each node
is proportional to the number of study participants. ASCH atypical squamous cells cannot exclude high-grade
lesion, ASCUS atypical squamous cells of undetermined significance, HC2 Hybrid Capture 2, HPV16/18(/45)
genotyping for HPV types 16 or 18 (or 45), HSIL high-grade squamous intraepithelial lesion, LBC liquid-based
cytology, LSIL low-grade squamous intraepithelial lesion, mRNA messenger ribonucleic acid, PCR polymerase
chain reaction.

HC2 AND ≥ ASCUS, which were reported in 10 studies. Reference standards were used for all participants with
routine colposcopy-directed biopsy in three s­tudies36,39,40 and colposcopy and selective biopsy in six studies
(Table 2)54,56,58–61. Other studies performed statistical corrections for data from unverified samples based on the
verified samples with colposcopy-directed biopsy in nine ­studies41,42,44,45,47,48,50,51,53 and colposcopy and selective
biopsy in nine s­ tudies62–70. See Supplementary results and Supplementary Tables S1–S3 for more details on study,
test, and reference standard characteristics.

Risk of bias.  Although the studies were predominantly well conducted, their designs varied substantially,
and several sources of bias were observed (Supplementary Fig. S2), such as lack of blinding of the colposcopists
or grading pathologists to the screening results. Additionally, verification bias could not be ruled out in studies
that did not perform histological evaluation of all samples.

Topology of direct comparisons of alternative screening algorithms.  Figure 1 shows the network
of compared algorithms available from the 27 studies, and Supplementary Table S4 shows the numbers of studies
and participating women contributing to each comparison. From 25 screening strategies, 300 pairwise compari-
sons are theoretically constructable. However, the 27 studies provided 337 contrast data (median 6 [min–max,
1–55] contrasts per study) on only 123 unique pairwise comparisons (41% of all theoretically constructable
contrasts). A comparison was based on a median of two studies (min–max, 1–14), and only 18 (15%) of 123
comparisons were based on five or more studies. The three most common comparisons were derived from stud-
ies that assessed HC2 and ≥ ASCUS; that is, the comparisons on standalone HC2 vs. standalone ≥ ASCUS (14
studies; 84,330 women), ≥ ASCUS alone vs. HC2 OR ≥ ASCUS (10 studies; 53,337 women), and HC2 alone vs.
HC2 OR ≥ ASCUS (10 studies; 53,303 women).

Sensitivity and specificity.  The sensitivity estimates varied substantially across studies with broad confi-
dence intervals (CIs); the specificity values also varied although their CIs were narrow (Supplementary Fig. S3).
Large between-study heterogeneity was visually noted in studies of HC2, all thresholds of cytological testing,
and their combinations. These results were also reflected in large credible and predictive regions of the average
sensitivity and specificity in the separately performed standard bivariate meta-analyses (Supplementary Fig. S4).
Although data points were limited, heterogeneity was less prominent in PCR and PCR-based combinations. See
Supplementary Fig. S5 for the average estimates of screening accuracy based on the standard meta-analysis.

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

Screening strategies Sensitivity (CrI) Ranking (CrI) Best P Specificity (CrI) Ranking (CrI) Best P

PCR OR ≥ASCUS 1.000 (0.994 to 1.000) 1 (1 to 1) 1.00 0.846 (0.753 to 0.907) 24 (20 to 25) 0.00
HC2 OR ≥ASCUS 0.968 (0.937 to 0.984) 2 (2 to 4) 0.00 0.850 (0.794 to 0.893) 24 (22 to 25) 0.00
mRNA OR ≥ASCUS 0.957 (0.832 to 0.991) 3 (2 to 8) 0.00 0.877 (0.750 to 0.940) 22 (16 to 25) 0.00
PCR 0.941 (0.872 to 0.976) 4 (2 to 8) 0.00 0.874 (0.810 to 0.918) 22 (18 to 25) 0.00
HC2 OR ≥LSIL 0.929 (0.870 to 0.963) 5 (3 to 8) 0.00 0.879 (0.828 to 0.917) 22 (19 to 24) 0.00
HC2 OR ≥HSIL 0.904 (0.824 to 0.950) 7 (4 to 10) 0.00 0.902 (0.858 to 0.935) 19 (16 to 22) 0.00
HC2 0.884 (0.821 to 0.926) 8 (5 to 10) 0.00 0.906 (0.872 to 0.932) 19 (16 to 22) 0.00
mRNA 0.872 (0.690 to 0.955) 8 (4 to 13) 0.00 0.921 (0.864 to 0.955) 17 (13 to 22) 0.00
HPV16/18/45 OR [mRNA AND ≥ASCUS] 0.844 (0.456 to 0.966) 9 (4 to 20) 0.00 0.941 (0.878 to 0.972) 15 (10 to 21) 0.00
≥LSIL OR [PCR AND ≥ASCUS] 0.839 (0.670 to 0.943) 9 (4 to 12) 0.00 0.913 (0.849 to 0.951) 18 (15 to 23) 0.00
HPV16/18 OR ≥ASCUS 0.748 (0.570 to 0.876) 11 (9 to 15) 0.00 0.928 (0.875 to 0.959) 16 (13 to 20) 0.00
≥LSIL OR [HC2 AND ASCUS] 0.699 (0.542 to 0.826) 13 (9 to 17) 0.00 0.961 (0.938 to 0.975) 11 (8 to 15) 0.00
HPV16/18 OR [PCR AND ≥ASCUS] 0.696 (0.515 to 0.834) 13 (11 to 18) 0.00 0.952 (0.920 to 0.971) 13 (10 to 16) 0.00
mRNA AND ≥ASCUS 0.657 (0.190 to 0.921) 14 (7 to 25) 0.00 0.963 (0.922 to 0.985) 11 (4 to 16) 0.00
≥HSIL OR [HC2 AND ≥ASCUS] 0.612 (0.445 to 0.765) 15 (11 to 20) 0.00 0.974 (0.958 to 0.984) 7 (6 to 11) 0.00
≥ASCUS 0.611 (0.499 to 0.710) 15 (12 to 18) 0.00 0.943 (0.922 to 0.959) 14 (12 to 17) 0.00
PCR AND ≥ASCUS 0.578 (0.380 to 0.746) 17 (13 to 23) 0.00 0.966 (0.945 to 0.980) 10 (6 to 13) 0.00
≥ASCH 0.568 (0.445 to 0.675) 17 (14 to 21) 0.00 0.971 (0.953 to 0.981) 8 (6 to 12) 0.00
HC2 AND ≥ASCUS 0.534 (0.384 to 0.668) 19 (14 to 22) 0.00 0.979 (0.968 to 0.986) 5 (4 to 8) 0.00
HPV16/18 0.520 (0.317 to 0.719) 20 (13 to 24) 0.00 0.966 (0.942 to 0.980) 10 (5 to 13) 0.00
≥LSIL 0.520 (0.392 to 0.631) 20 (16 to 23) 0.00 0.976 (0.966 to 0.983) 6 (5 to 9) 0.00
HC2 AND ≥LSIL 0.461 (0.305 to 0.611) 22 (17 to 24) 0.00 0.985 (0.976 to 0.991) 4 (3 to 5) 0.00
HPV16/18 AND ≥ASCUS 0.370 (0.133 to 0.626) 23 (16 to 25) 0.00 0.989 (0.975 to 0.995) 3 (1 to 6) 0.07
≥HSIL 0.346 (0.216 to 0.497) 24 (21 to 25) 0.00 0.994 (0.990 to 0.996) 2 (1 to 3) 0.43
HC2 AND ≥HSIL 0.345 (0.183 to 0.519) 24 (21 to 25) 0.00 0.994 (0.989 to 0.997) 2 (1 to 3) 0.50

0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

Figure 2.  Average sensitivity and specificity and ranking of standalone tests and combination algorithms for
cervical cancer screening for detecting CIN2+. Point estimates (blue squares) and CrIs (extending lines) are
presented (ordered by the ranking of each test/combination’s sensitivity). See Table 1 for the definition of each
strategy. ASCH atypical squamous cells cannot exclude high-grade lesion, ASCUS atypical squamous cells of
undetermined significance, CrI 95% credible interval, HC2 Hybrid Capture 2, HPV16/18(/45) genotyping for
HPV types 16 or 18 (or 45), HSIL high-grade squamous intraepithelial lesion, LBC liquid-based cytology, LSIL
low-grade squamous intraepithelial lesion, mRNA messenger ribonucleic acid, PCR polymerase chain reaction.

Figure 2 provides the average accuracy estimates and ranking estimated through the network meta-analysis.
Overall, the combinations with the OR rule of hrHPV and cytological testing were most sensitive and least spe-
cific, whereas combinations with the AND rule of hrHPV and cytological testing were most specific and least
sensitive. The rankings estimated in the network meta-analysis reflected the trade-off between sensitivity and
specificity by altering the thresholds; lowering the thresholds of cytological testing (e.g., from ≥ HSIL to ≥ ASCUS)
led to higher sensitivity but at the cost of reduced specificity, and tightening the thresholds increased specificity at
the cost of reduced sensitivity. This behavior resulted in average estimates and rankings for tests or combination
algorithms relying on few studies (e.g., HPV16/18- and mRNA-based combinations assessed in only one study
each), which were inconsistent with the standard meta-analysis.
In the network meta-analysis, PCR OR ≥ ASCUS was most sensitive (1.0; CrI: 0.994–1.0; probability of best
sensitivity: 1.0) but was one of the two least specific screening algorithms (0.846; CrI: 0.753–0.907). In contrast,
standalone ≥ HSIL and HC2 AND ≥ HSIL were the two most specific (respectively, 0.994 [CrI: 0.990–0.996;
probability of best specificity: 0.43] and 0.994 [CrI: 0.989–0.997]; probability of best specificity: 0.50) but were
the two least sensitive (respectively, 0.346 [95% CrI: 0.216–0.497] and 0.345 [CrI: 0.183–0.519]) algorithms.

Comparative accuracy.  Supplementary Figure S6, Supplementary Tables S5 and S6, respectively, summa-
rize the average relative sensitivity and specificity and ΔFNs and ΔFPs estimated based on a population of 1000
healthy women, with a 2% prevalence of CIN2+, across all possible paired comparisons of available standalone
tests and combination algorithms.

Comparative accuracy of standalone tests.  For cytological testing, the average relative estimates of
screening accuracy reflected the effect of altering the thresholds (Fig. 3a, Supplementary Table S7). For exam-
ple, ≥ ASCUS was more sensitive than ≥ LSIL (relative sensitivity: 0.86 [CrI: 0.69–0.97; Bayesian P(≥ 1) < 0.001])
but less specific than ≥ LSIL (relative specificity: 1.03 [CrI: 1.05–1.02; Bayesian P(≤ 1) < 0.001]). Two studies that
directly compared the alternative smear preparation methods showed identical sensitivity and specificity for CC
and LBC for each threshold (Supplementary Fig. S7).
HPV16/18 was more specific but less sensitive than the other hrHPV assays (Fig. 3a, Supplementary Table S7).
For example, for comparing HPV16/18 with HC2, the relative specificity was 1.06 [CrI: 1.10–1.04; Bayesian
P(≤ 1) < 0.001] and relative sensitivity was 0.59 [CrI: 0.36–0.81; Bayesian P(≥ 1) < 0.001]. Among HC2, PCR-
based tests, and mRNA-based tests, data were limited as to whether a specific hrHPV assay was more sensitive
or specific than any other. For example, although the PCR-based tests appeared more sensitive but less specific
than HC2, the CrIs for the relative accuracy crossed 1, the null value (i.e., the relative sensitivity of PCR vs. HC2

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 6

Vol:.(1234567890)
www.nature.com/scientificreports/

a 1.0
b 1.0
HC2 OR ≥ASCUS
PCR HC2 OR ≥LSIL
HC2 HC2 OR ≥HSIL
HC2
mRNA

0.8 0.8

≥LSIL OR [HC2 AND ASCUS]

≥ASCUS
0.6 ≥ASCUS
0.6 ≥HSIL OR [HC2 AND ≥ASCUS]
Sensitivity

Sensitivity
≥ASCH
HPV16/18 HC2 AND ≥ASCUS
≥LSIL
HC2 AND ≥LSIL

0.4 0.4
≥HSIL HC2 AND ≥HSIL

0.2 0.2

0 0
1.0 0.8 0.6 1.0 0.8 0.6
Specificity Specificity

c 1.0
PCR OR ≥ASCUS d 1.0
mRNA OR ≥ASCUS
PCR

mRNA
≥LSIL OR [PCR AND ASCUS] HPV16/18/45 OR [mRNA AND ≥ASCUS]
0.8 0.8
HPV16/18 OR ≥ASCUS
HPV16/18 OR [PCR AND ≥ASCUS]
mRNA AND ≥ASCUS

0.6 ≥ASCUS
0.6 ≥ASCUS
Sensitivity

Sensitivity

PCR AND ≥ASCUS

HPV16/18

0.4 0.4
HPV16/18 AND ≥ASCUS

0.2 0.2

0 0
1.0 0.8 0.6 1.0 0.8 0.6
Specificity Specificity

Figure 3.  Network meta-analysis of standalone tests and combination algorithms for cervical cancer screening
for detecting CIN2+. Average sensitivity and specific and their 95% credible regions for (a) standalone cytology
or hrHPV testing, (b) HC2-based combination algorithms, (c) PCR-based combination algorithms (including
PCR-based genotyping for HPV16/18), and (d) mRNA-based combination algorithms (including mRNA-based
genotyping for HPV16/18/45). Graded colors (black, dark gray, gray, and light gray) indicate cytology with a
specific threshold, red indicates HC2, blue indicates PCR-based tests, green indicates HPV16/18, and magenta
indicates mRNA-based tests. Triangles and diamonds represent standalone hrHPV testing and cytology,
respectively. Circles and squares represent combinations based on the OR-rule and the AND-rule, respectively.
For combination algorithms (b–d), standalone component hrHPV testing and cytology (≥ ASCUS) are also
presented as reference. See Table 1 for the definition of each strategy. ASCH atypical squamous cells cannot
exclude high-grade lesion, ASCUS atypical squamous cells of undetermined significance, HC2 Hybrid Capture
2, HPV16/18(/45) genotyping for HPV types 16 or 18 (or 45), HSIL high-grade squamous intraepithelial lesion,
LBC liquid-based cytology, LSIL low-grade squamous intraepithelial lesion, mRNA messenger ribonucleic acid,
PCR polymerase chain reaction.

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

Index (for specificity) and comparator (for sensitivity) tests or combination algorithms
Index and
comparator tests HPV16/18/45 HPV16/18
or combination OR [mRNA  ≥ LSIL OR [PCR  ≥ LSIL OR [HC2 OR [PCR
algorithms PCR HC2 mRNA AND ≥ ASCUS] AND ASCUS] AND ASCUS] AND ≥ ASCUS]  ≥ ASCUS
Index (for sensitivity) and comparator (for specificity) tests or combination algorithms
0.94 (0.87 to 1.02) 0.93 (0.73 to 1.04) 0.90 (0.49 to 1.05) 0.89 (0.72 to 1.01) 0.75 (0.58 to 0.89) 0.74 (0.56 to 0.89) 0.65 (0.54 to 0.75)
PCR –
[0.06] [0.10] [0.12] [0.04] [< 0.001] [< 0.001] [< 0.001]
1.04 (0.99 to 1.11) 0.99 (0.79 to 1.10) 0.96 (0.52 to 1.11) 0.95 (0.76 to 1.08) 0.79 (0.63 to 0.93) 0.79 (0.59 to 0.95) 0.69 (0.58 to 0.79)
HC2 –
[0.08] [0.42] [0.35] [0.26] [0.001] [0.003] [< 0.001]
1.05 (0.99 to 1.13) 1.02 (0.96 to 1.06) 0.97 (0.53 to 1.24) 0.97 (0.77 to 1.24) 0.81 (0.62 to 1.05) 0.81 (0.59 to 1.05) 0.71 (0.58 to 0.89)
mRNA –
[0.05] [0.26] [0.41] [0.37] [0.049] [0.049] [0.005]
HPV16/18/45
1.07 (1.00 to 1.16) 1.04 (0.97 to 1.08) 1.02 (0.96 to 1.08) 1.00 (0.77 to 1.82) 0.84 (0.62 to 1.54) 0.84 (0.60 to 1.50) 0.73 (0.58 to 1.33)
OR [mRNA –
[0.02] [0.10] [0.23] [0.49] [0.19] [0.19] [0.09]
AND ≥ ASCUS]
 ≥ LSIL OR [PCR 1.04 (0.97 to 1.12) 1.01 (0.94 to 1.05) 0.99 (0.92 to 1.06) 0.97 (0.91 to 1.04) 0.84 (0.64 to 1.08) 0.84 (0.65 to 0.97) 0.73 (0.59 to 0.92)
–
AND ASCUS] [0.10] [0.39] [0.40] [0.18] [0.08] [< 0.001] [0.004]
 ≥ LSIL OR [HC2 1.10 (1.05 to 1.18) 1.06 (1.03 to 1.09) 1.04 (1.00 to 1.11) 1.02 (0.98 to 1.09) 1.05 (1.01 to 1.13) 1.00 (0.72 to 1.33) 0.88 (0.71 to 1.10)
–
AND ASCUS] [< 0.001] [< 0.001] [0.02] [0.16] [0.007] [0.49] [0.11]
HPV16/18
1.09 (1.04 to 1.16) 1.05 (1.02 to 1.09) 1.03 (0.99 to 1.09) 1.01 (0.97 to 1.08) 1.04 (1.01 to 1.11) 0.99 (0.96 to 1.02) 0.88 (0.71 to 1.16)
OR [PCR –
[< 0.001] [0.004] [0.06] [0.29] [< 0.001] [0.25] [0.16]
AND ≥ ASCUS]
1.08 (1.03 to 1.15) 1.04 (1.02 to 1.07) 1.02 (0.99 to 1.08) 1.00 (0.97 to 1.07) 1.03 (0.99 to 1.10) 0.98 (0.96 to 1.00) 0.99 (0.97 to 1.02)
 ≥ ASCUS –
[< 0.001] [< 0.001] [0.10] [0.45] [0.06] [0.04] [0.23]

Table 3.  Comparative accuracy of guideline-proposed combination algorithms. Above the diagonal line

(formed by cells with an en dash) represents relative sensitivity (95% CrI) [probability that relative sensitivity
is ≥ 1] and below the diagonal line represents relative specificity (95% CrI) [probability that relative specificity
is ≤ 1]. For relative sensitivity, the rows and columns, respectively, represent the index (the test of interest) and
comparator (the test in comparison) tests or combination algorithms. For relative sensitivity, the columns
and rows, respectively, represent the index and comparator tests or combination algorithms. ASCUS atypical
squamous cells of undetermined significance, CrI credible interval, HC2 Hybrid Capture 2, HPV16/18(/45)
genotyping for HPV types 16 or 18 (or 45), HSIL high-grade squamous intraepithelial lesion, LBC liquid-based
cytology, LSIL low-grade squamous intraepithelial lesion, mRNA messenger ribonucleic acid, PCR polymerase
chain reaction.

was 1.06 [CrI: 0.98–1.15]; Bayesian P(≤ 1) = 0.06) and relative specificity of HC2 vs. PCR was 1.04 [CrI: 0.99–1.11;
Bayesian P(≤ 1) = 0.08]).
Compared with standalone cytological testing irrespective of the thresholds, all standalone hrHPV assays
other than HPV16/18 were more sensitive but less specific in general (Fig. 3a, Supplementary Table S7). In con-
trast, the accuracy of HPV16/18 was comparable to cytological testing. For example, the relative specificity for
comparing ≥ LSIL with HPV16/18 was 1.0 [CrI: 0.68–1.62; Bayesian P(≥ 1) = 0.50] and relative specificity was
1.01 (CrI: 1.00–1.03; Bayesian P(≤ 1) = 0.10).

Comparative accuracy among combination algorithms based on specific hrHPV assays.  The
ROC plots of the average accuracy estimates and their credible regions reflected the effect of altering the thresh-
olds in combined cytological testing (i.e., lower thresholds with increased sensitivity and decreased specificity,
and higher thresholds with increased specificity and decreased sensitivity) and the effect of combination meth-
ods (i.e., the OR rule with increased sensitivity and decreased specificity, and the AND rule with increased speci-
ficity and decreased sensitivity) across the subgroups based on alternative hrHPV assays (Fig. 3b–d). Among
45 pairwise comparisons based on cytology, HC2, and their combinations, most (40 [89%] for sensitivity and
42 [93%] for specificity) showed a significant difference, reflecting the effect of the thresholds and combina-
tion methods (Fig. 3b, Supplementary Table S8). Similarly, among 36 pairwise comparisons based on cytology,
PCR-based tests, and their combinations, 28 (78%) for sensitivity and 27 (75%) for specificity showed a signifi-
cant difference (Fig. 3c, Supplementary Table S9). In contrast, 10 pairwise comparisons based on mRNA-based
combinations (Fig. 3d, Supplementary Table S10), only five (50%) and four (40%) contrasts for sensitivity and
specificity, respectively, were significantly different.

Comparative accuracy and GRADE assessment of guideline‑proposed combination algo‑

rithms.  Data on the guideline-proposed algorithms are available for HC2 and PCR-based tests on “≥ LSIL
OR [hrHPV AND ASCUS]” and for mRNA-based tests and PCR-based tests on “HPV16/18(/45) OR [hrHPV
AND ≥ ASCUS]”. Table 3 summarizes the comparative accuracy, and Supplementary Table S11 and Table 4 show
the GRADE summary of findings on specific tests or combination algorithms and their comparisons, respec-
tively.
In general, the proposed algorithms were less sensitive but more specific than the standalone component
hrHPV assays. However, only HC2-based “≥ LSIL OR [hrHPV AND ASCUS]” and PCR-based “HPV16/18 OR
[hrHPV AND ≥ ASCUS]” were significantly less sensitive (the average relative sensitivity ranged from 0.74 to
0.79; Bayesian P(≥ 1) ranged from < 0.001 to 0.003) and more specific (the average relative specificity ranged from

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 8

Vol:.(1234567890)
www.nature.com/scientificreports/

Table 4.  The GRADE summary of findings table for comparative evidence. Above the diagonal line (formed
by cells with an em dash) represents the number of the difference in (Δ) FNs (95% CrI) and below the diagonal
line represents Δ FPs (95% CrI). For Δ FPs, the rows and columns, respectively, represent the index (the test of
interest) and comparator (the test in comparison) tests or combination algorithms. For Δ FNs, the columns and
rows, respectively, represent the index and comparator tests or combination algorithms. Results are based on a
healthy screening population of 1000 women in which 20 are CIN2+ (2%). ASCUS atypical squamous cells of
undetermined significance, CIN2+, CrI cervical intraepithelial neoplasia grade 2 or higher grades; 95% credible
interval, FN false negative, FP false positive, GRADE Grading of Recommendations Assessment, Development
and Evaluation, HC2 Hybrid Capture 2, HPV16/18(/45) genotyping for HPV types 16 or 18 (or 45), HSIL high-
grade squamous intraepithelial lesion, LBC liquid-based cytology, LSIL low-grade squamous intraepithelial
lesion, mRNA messenger ribonucleic acid, PCR polymerase chain reaction, TN true negative, TP true positive.

1.04 to 1.10; Bayesian P(≤ 1) ranged from < 0.001 to 0.004). These results suggested that the proposed algorithms,
compared with their standalone component hrHPV tests, decreased by an average of 44 to 88 FPs but increased
4 to 5 more FNs (very low to low certainty of evidence).
In contrast, the proposed algorithms were in general equally specific but more sensitive than stan-
dalone ≥ ASCUS. However, only PCR-based “LSIL OR [hrHPV AND ASCUS]” was significantly less sensitive
than ≥ ASCUS alone (the relative sensitivity = 0.73 [CrI: 0.59–0.92; Bayesian P(≥ 1) = 0.004]; four more FNs [CrI:
1–7]; very low certainty of evidence), but evidence as to whether this combination was more specific or less
specific than ≥ ASCUS alone was insufficient (relative sensitivity = 0.98 [CrI: 0.96–1.00; Bayesian P(≥ 1) = 0.04]).
Comparative evidence across alternative guideline-proposed algorithms was generally limited. PCR-based
“LSIL OR [hrHPV AND ASCUS]” was significantly more specific and less specific than “HPV16/18 OR [hrHPV
AND ≥ ASCUS]” (relative sensitivity; 1.04 [CrI: 1.01–1.11]; Bayesian P(≤ 1) < 0.001]; 37 fewer FPs [CrI: 6–92] and
relative specificity: 0.84 [CrI: 0.65–0.97]; Bayesian P(≥ 1) < 0.001; three more FNs [CrI: 1–6]; very low certainty
of evidence). Although only HC2-based “LSIL OR [hrHPV AND ASCUS]” was more specific than PCR-based
“LSIL OR [hrHPV AND ASCUS]” (relative specificity: 1.05 [CrI: 1.01–1.13]; Bayesian P(≤ 1) = 0.007; 46 fewer FPs
[CrI: 7–105]; very low certainty of evidence) across-hrHPV assays, comparative data on the guideline-proposed
algorithms were insufficient.

Meta‑regression and sensitivity analyses.  Due to data paucity, meta-regression was undertaken for
only HC2, cytological testing, and their OR combination separately. Although high-income countries (vs. non-
high-income countries) for sensitivity of HC2 and sample collection by physicians (vs. nonphysician collectors)
for sensitivity and specificity of ≥ ASCUS were associated with higher estimates, these covariates were no longer
associated with higher (or lower) sensitivity or specificity in their combination, HC2 OR ≥ ASCUS (Supplemen-
tary Fig. S8).
The sensitivity analysis using the model with a common correlation parameter across tests yielded results
comparable to those of the main analysis based on the model with test-specific correlation parameters (Supple-
mentary Table S12). Relaxing threshold constraints yielded results not compliant with the expected threshold
effects in two specific thresholds for cytological testing (≥ LSIL and ≥ ASCH) and unstable results with wide CrIs

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

for sensitivity in four combination algorithms (i.e., mRNA AND ≥ ASCUS, HPV16/18 AND ≥ ASCUS, HPV16/18
OR ≥ ASCUS, “ ≥ LSIL OR [PCR AND ASCUS]”, and “≥ HSIL OR [HC2 AND ≥ ASCUS]”) regardless of whether
correlation parameters were separately assumed or not; all of these tests, except for ≥ LSIL, depended on only a
few primary studies. With lower deviance information criterion estimates, the models with threshold constraints
were deemed to be better-fitting than the models without threshold constraints; however, the differences were < 5,
suggesting no definitively preferred model.

Discussion
To the best of our knowledge, this is the first network meta-analysis that has comprehensively compared and
ranked the cross-sectional screening accuracy of standalone cytology or hrHPV testing with combination algo-
rithms for detecting CIN2+. Importantly, this analysis is based on published accuracy estimates from fully
paired-design comparative accuracy studies that addressed verification bias. First, our network meta-analysis
confirmed and quantified the theoretically expected gain in and trade-off of screening performance when com-
bining two t­ ests24, that is, the combinations with the OR rule (i.e., either test positive) of hrHPV and cytological
testing were most sensitive and least specific, whereas combinations with the AND rule (i.e., both test positive)
of hrHPV and cytological testing were most specific and least sensitive. Second, our network meta-analysis con-
firmed that the guideline-proposed combination algorithms, HC2-based “≥ LSIL OR [hrHPV AND ASCUS]”
and PCR-based “HPV16/18 OR [hrHPV AND ≥ ASCUS]” appeared to compensate the shortcomings of the
two component tests if used as standalone, which, though expected theoretically, had never been quantitatively
synthesized. Specifically, these proposed algorithms were not as sensitive but more specific than the component
standalone hrHPV testing. Similarly, these proposed algorithms appeared equally specific but more sensitive than
standalone ≥ ASCUS, though definitive conclusions could not be made due to limited comparative data. Third,
sparse, insufficient comparative evidence precluded reliable assessment of the comparative accuracy across these
alternative guideline-proposed algorithms.
Effectiveness of screening should be assessed as a whole program consisting of a set of a­ ctivities71. Since the
ultimate goal is to maximize participant-relevant benefits and simultaneously minimize harms, accuracy of
testing is, though an important measure, only an intermediate parameter. As already elucidated in the previous
meta-analyses13,14, which is congruent with our results, standalone testing for hrHPV using an assay other than
HPV 16/18 genotyping, if all screen-positive women underwent colposcopy, would identify more women with
CIN2+ than cytological testing alone but at the cost of more healthy women misclassified as CIN2+. The OR rule
combinations, the most sensitive group of strategies found in our meta-analysis, if used for primary co-testing
(i.e., performing both tests concurrently), would further increase the number of healthy women misclassified
as CIN2+ while identifying only a few more women with CIN2+. The consequences of such FP results include
unnecessary colposcopy, triage, or repeat testing with cytology, hrHPV, or other tests. Although infections with
hrHPV, and HPV16/18 in particular, carry a higher risk of progression than positive c­ ytology72–75, immediate
incremental costs and psychological burden incurred due to increased false-positive results may not be justified
in low risk screening settings as only a fraction of the identified CIN2+ lesions detected through standalone
hrHPV testing or its combinations progress to invasive cancer; the others actually carry a moderate chance of
­regression76. The AND rule combinations, the most specific group of strategies identified in our meta-analysis,
may substantially minimize FPs and their negative consequences. However, sensitivity is lower than cytology
alone (≥ ASCUS), potentially leading to unignorably large numbers of FNs depending on the prevalence of
CIN2+ in a screened population.
As interim recommendations, several protocols for triage and/or repeat testing followed by colposcopy for
screen-positive women have been proposed by professional societies. “≥ LSIL OR [hrHPV AND ASCUS]” and
“HPV16/18 OR [hrHPV AND ≥ ASCUS]” were cross-sectional representations for two such protocols, respec-
tively, proposed for positive primary cytological t­ esting11 and primary hrHPV t­ esting9. Our meta-analysis found
that the accuracy of these combination algorithms were generally ranked in the middle, being more sensitive
and less specific than standalone cytology (≥ ASCUS) and the AND rule combinations but more specific and less
sensitive than standalone hrHPV testing and the OR rule combination. We also quantified how each combination
algorithm increased or decreased the number of FNs and FPs relative to those of another specific standalone
test or combination, which is a strength of our study results. However, any benefits and harms associated with
specific screening tests or combinations should be formally assessed at the whole program level along with its
necessary resources and ­costs71.
We focused on cross-sectional accuracy of initial screening tests or combinations and their immediate conse-
quences. Our accuracy-based arguments necessarily lack long-term outcomes. Given the chance of r­ egression76,
the results based on our cross-sectional approach may be only relevant in populations with a low participation
rate of follow-up testing. Additionally, the positive criteria we adopted for the estimation of accuracy do not
necessarily represent the optimal indications of colposcopy in real-life practice; rather the criteria included the
joint indications of any additional intervention; i.e., triage and/or repeat testing, colposcopy, and immediate
direct treatments jointly. In this regard, a recent expert consensus statement proposed individualized risk-based
management decisions based on the combinations of the available screening ­results77.
Colposcopy-directed biopsy is an imperfect test even for routine biopsies on normal-appearing ­sites78 and
more so for colposcopy and selective ­biopsy79. Despite the theoretical superiority of verification bias-corrected
accuracy estimates over naïvely calculated estimates, these corrections are not error-free. Given the complex
mechanisms of missing ­verification80 and limitations in inverse probability ­weighting81, bias may not necessarily
have been corrected in the right direction. In addition, the effect of the excluded observations due to unsatisfac-
tory or missing test results, even though the reported proportions were not substantial, could be unpredictably
large. Furthermore, our meta-analysis was based on aggregate data and thus only accounted for the dependence

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 10

Vol:.(1234567890)
www.nature.com/scientificreports/

of the two tests at the aggregate data ­level82; however, a more sophisticated approach to address these limitations
would require individual-level data.
Our GRADE assessment used a typical population-based screening context in high-income countries as
adopted in a previous r­ eview13; however, the large spread of the credible and predictive accuracy values in our
study suggests wide-ranging real-life variations, implying that specific scenarios with different risks might yield
divergent conclusions. Finally, we did not assess combinations involving newer screening modalities, such as
p16/Ki-67 dual-stain-based c­ ytology83, as this was beyond the scope of our meta-analysis.

Conclusions
Limited evidence suggests that specific test combinations might complement the weaknesses of standalone
cytological or hrHPV screening and help reduce FN and/or FP results. However, the strategies that provide
more benefits than harms at reasonable cost in a population need to be assessed at the program level. As com-
parative evidence on alternative hrHPV assays is sparse, further research is needed to acquire relevant data.
Additionally, future research should elucidate long-term outcomes of specific algorithms and acquire data from
HPV-vaccinated populations.

Data availability
The data and statistical codes that supports the findings of this study will be shared on reasonable request to the
corresponding author.

Received: 21 July 2021; Accepted: 17 December 2021

References
1. Bray, F. et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185
countries. CA Cancer J. Clin. 68, 394–424. https://​doi.​org/​10.​3322/​caac.​21492 (2018).
2. WHO. Human Papillomavirus (HPV) and Cervical Cancer (WHO, 2019).
3. Peirson, L., Fitzpatrick-Lewis, D., Ciliska, D. & Warren, R. Screening for cervical cancer: A systematic review and meta-analysis.
Syst. Rev. 2, 35. https://​doi.​org/​10.​1186/​2046-​4053-2-​35 (2013).
4. Crosbie, E. J., Einstein, M. H., Franceschi, S. & Kitchener, H. C. Human papillomavirus and cervical cancer. Lancet 382, 889–899.
https://​doi.​org/​10.​1016/​s0140-​6736(13)​60022-7 (2013).
5. Ronco, G. et al. Efficacy of HPV-based screening for prevention of invasive cervical cancer: Follow-up of four European randomised
controlled trials. Lancet 383, 524–532. https://​doi.​org/​10.​1016/​s0140-​6736(13)​62218-7 (2014).
6. Curry, S. J. et al. Screening for cervical cancer: US Preventive Services Task Force Recommendation Statement. JAMA 320, 674–686.
https://​doi.​org/​10.​1001/​jama.​2018.​10897 (2018).
7. Jeronimo, J., Castle, P. E., Temin, S. & Shastri, S. S. Secondary Prevention of Cervical Cancer: American Society of Clinical Oncol-
ogy Resource-stratified clinical practice guideline summary. J. Oncol. Pract. 13, 129–133. https://​doi.​org/​10.​1200/​jop.​2016.​017889
(2017).
8. Sawaya, G. F., Kulasingam, S., Denberg, T. D. & Qaseem, A. Cervical cancer screening in average-risk women: Best practice advice
from the Clinical Guidelines Committee of the American College of Physicians. Ann. Intern. Med. 162, 851–859. https://​doi.​org/​
10.​7326/​m14-​2426 (2015).
9. Huh, W. K. et al. Use of primary high-risk human papillomavirus testing for cervical cancer screening: Interim clinical guidance.
Obstet. Gynecol. 125, 330–337. https://​doi.​org/​10.​1097/​aog.​00000​00000​000669 (2015).
10. Fontham, E. T. H. et al. Cervical cancer screening for individuals at average risk: 2020 guideline update from the American Cancer
Society. CA Cancer J. Clin. 70, 321–346. https://​doi.​org/​10.​3322/​caac.​21628 (2020).
11. Saslow, D. et al. American Cancer Society, American Society for Colposcopy and Cervical Pathology, and American Society for
Clinical Pathology screening guidelines for the prevention and early detection of cervical cancer. CA Cancer J. Clin. 62, 147–172.
https://​doi.​org/​10.​3322/​caac.​21139 (2012).
12. Dahabreh, I. J., Trikalinos, T. A., Balk, E. M. & Wong, J. B. Methods Guide for Effectiveness and Comparative Effectiveness Reviews
(Agency for Healthcare Research and Quality (US), 2016).
13. Koliopoulos, G. et al. Cytology versus HPV testing for cervical cancer screening in the general population. Cochrane. Database.
Syst. Rev. 8, CD008587. https://​doi.​org/​10.​1002/​14651​858.​CD008​587.​pub2 (2017).
14. Mustafa, R. A. et al. Systematic reviews and meta-analyses of the accuracy of HPV tests, visual inspection with acetic acid, cytology,
and colposcopy. Int. J. Gynaecol. Obstet. 132, 259–265. https://​doi.​org/​10.​1016/j.​ijgo.​2015.​07.​024 (2016).
15. Fokom-Domgue, J. et al. Performance of alternative strategies for primary cervical cancer screening in sub-Saharan Africa: Sys-
tematic review and meta-analysis of diagnostic test accuracy studies. BMJ 351, h3084. https://​doi.​org/​10.​1136/​bmj.​h3084 (2015).
16. Li, T. et al. Diagnostic value of combination of HPV testing and cytology as compared to isolated cytology in screening cervical
cancer: A meta-analysis. J. Cancer Res. Ther. 12, 283–289. https://​doi.​org/​10.​4103/​0973-​1482.​154032 (2016).
17. Biondi-Zoccai, G. (ed.) Diagnostic Meta-analysis: A Useful Tool for Clinical Decision-Making 183–197 (Springer, 2018).
18. Hamashima, C. et al. The Japanese guideline for cervical cancer screening. Jpn. J. Clin. Oncol. 40, 485–502. https://d ​ oi.o
​ rg/1​ 0.1​ 093/​
jjco/​hyq036 (2010).
19. The Japanese Research Group for Systematic Review and Guideline Development for Cancer Screening. An Evidence Report for the
Japanese Guideline for Cervical Cancer Screening 2019, 1–258 (2020). http://​cansc​reen.​ncc.​go.​jp/​guide​line/​shiky​ukeir​eport​2019.​
pdf. Accessed 20 Dec 2021.
20. McInnes, M. D. F. et al. Preferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies:
The PRISMA-DTA statement. JAMA 319, 388–396. https://​doi.​org/​10.​1001/​jama.​2017.​19163 (2018).
21. Denton, K. J. et al. The revised BSCC terminology for abnormal cervical cytology. Cytopathology 19, 137–157. https://​doi.​org/​10.​
1111/j.​1365-​2303.​2008.​00585.x (2008).
22. Cirkel, C., Barop, C. & Beyer, D. A. Method comparison between Munich II and III nomenclature for Pap smear samples. J. Turk.
Ger. Gynecol. Assoc. 16, 203–207. https://​doi.​org/​10.​5152/​jtgga.​2015.​0147 (2015).
23. Schiffman, M. et al. Human papillomavirus testing in the prevention of cervical cancer. J. Natl. Cancer Inst. 103, 368–383. https://​
doi.​org/​10.​1093/​jnci/​djq562 (2011).
24. Macaskill, P., Walter, S. D., Irwig, L. & Franco, E. L. Assessing the gain in diagnostic performance when combining two diagnostic
tests. Stat. Med. 21, 2527–2546. https://​doi.​org/​10.​1002/​sim.​1227 (2002).
25. Dickinson, J. et al. Recommendations on screening for cervical cancer. CMAJ 185, 35–45. https://​doi.​org/​10.​1503/​cmaj.​121505
(2013).

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 11

Vol.:(0123456789)
 www.nature.com/scientificreports/

26. Yang, B. et al. Development of QUADAS-C, a risk of bias tool for comparative diagnostic accuracy studies. https://​doi.​org/​10.​
17605/​OSF.​IO/​HQ8MF (2021).
27. Whiting, P. F. et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann. Intern. Med. 155,
529–536. https://​doi.​org/​10.​7326/​0003-​4819-​155-8-​20111​0180-​00009 (2011).
28. Phillips, B., Stewart, L. A. & Sutton, A. J. “Cross hairs” plots for diagnostic meta-analysis. Res. Synth. Methods 1, 308–315. https://​
doi.​org/​10.​1002/​jrsm.​26 (2010).
29. Owen, R. K., Cooper, N. J., Quinn, T. J., Lees, R. & Sutton, A. J. Network meta-analysis of diagnostic test accuracy studies identifies
and ranks the optimal diagnostic tests and thresholds for health care policy and decision-making. J. Clin. Epidemiol. 99, 64–74.
https://​doi.​org/​10.​1016/j.​jclin​epi.​2018.​03.​005 (2018).
30. Harbord, R. M., Deeks, J. J., Egger, M., Whiting, P. & Sterne, J. A. A unification of models for meta-analysis of diagnostic accuracy
studies. Biostatistics 8, 239–251. https://​doi.​org/​10.​1093/​biost​atist​ics/​kxl004 (2007).
31. Chu, H., Nie, L., Cole, S. R. & Poole, C. Meta-analysis of diagnostic accuracy studies accounting for disease prevalence: Alternative
parameterizations and model selection. Stat. Med. 28, 2384–2399. https://​doi.​org/​10.​1002/​sim.​3627 (2009).
32. Arends, L. R. et al. Bivariate random effects meta-analysis of ROC curves. Med. Decis. Making 28, 621–638. https://​doi.​org/​10.​
1177/​02729​89x08​319957 (2008).
33. Human Development Indicators and Indices: 2018 Statistical Update Team. In United Nations Development Programme; 2018
(United Nations Development Programme, 2018).
34. Schunemann, H. J. et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ
336, 1106–1110. https://​doi.​org/​10.​1136/​bmj.​39500.​677199.​AE (2008).
35. Thompson, J. Bayesian Analysis with Stata (Stata Press, 2014).
36. Belinson, J. et al. Shanxi Province cervical cancer screening study: A cross-sectional comparative trial of multiple techniques to
detect cervical neoplasia. Gynecol. Oncol. 83, 439–444. https://​doi.​org/​10.​1006/​gyno.​2001.​6370 (2001).
37. Pan, Q. et al. A thin-layer, liquid-based pap test for mass screening in an area of China with a high incidence of cervical carcinoma.
A cross-sectional, comparative study. Acta. Cytol. 47, 45–50. https://​doi.​org/​10.​1159/​00032​6474 (2003).
38. Zhao, F. H. et al. A study of cervical cancer screening algorithms. Zhonghua Zhong Liu Za Zhi 32, 420–424 (2010).
39. Cardenas-Turanzas, M. et al. The performance of human papillomavirus high-risk DNA testing in the screening and diagnostic
settings. Cancer Epidemiol. Biomark. Prev. 17, 2865–2871. https://​doi.​org/​10.​1158/​1055-​9965.​epi-​08-​0137 (2008).
40. Hovland, S. et al. A comprehensive evaluation of the accuracy of cervical pre-cancer detection methods in a high-risk area in East
Congo. Br. J. Cancer 102, 957–965. https://​doi.​org/​10.​1038/​sj.​bjc.​66055​94 (2010).
41. Schneider, A. et al. Screening for high-grade cervical intra-epithelial neoplasia and cancer by testing for high-risk HPV, routine
cytology or colposcopy. Int. J. Cancer 89, 529–534 (2000).
42. Kulasingam, S. L. et al. Evaluation of human papillomavirus testing in primary screening for cervical abnormalities: Comparison
of sensitivity, specificity, and frequency of referral. JAMA 288, 1749–1757 (2002).
43. Balasubramanian, A. et al. Accuracy and cost-effectiveness of cervical cancer screening by high-risk human papillomavirus DNA
testing of self-collected vaginal samples. J. Low. Genit. Tract. Dis. 14, 185–195. https://​doi.​org/​10.​1097/​LGT.​0b013​e3181​cd6d36
(2010).
44. Bigras, G. & de Marval, F. The probability for a Pap test to be abnormal is directly proportional to HPV viral load: Results from a
Swiss study comparing HPV testing and liquid-based cytology to detect cervical cancer precursors in 13,842 women. Br. J. Cancer
93, 575–581. https://​doi.​org/​10.​1038/​sj.​bjc.​66027​28 (2005).
45. Mayrand, M. H. et al. Human papillomavirus DNA versus Papanicolaou screening tests for cervical cancer. N. Engl. J. Med. 357,
1579–1588. https://​doi.​org/​10.​1056/​NEJMo​a0714​30 (2007).
46. Mayrand, M. H. et al. Randomized controlled trial of human papillomavirus testing versus Pap cytology in the primary screening
for cervical cancer precursors: Design, methods and preliminary accrual results of the Canadian cervical cancer screening trial
(CCCaST). Int. J. Cancer 119, 615–623. https://​doi.​org/​10.​1002/​ijc.​21897 (2006).
47. Li, N. et al. Different cervical cancer screening approaches in a Chinese multicentre study. Br. J. Cancer 100, 532–537. https://​doi.​
org/​10.​1038/​sj.​bjc.​66048​40 (2009).
48. Castle, P. E. et al. Performance of carcinogenic human papillomavirus (HPV) testing and HPV16 or HPV18 genotyping for cervical
cancer screening of women aged 25 years and older: A subanalysis of the ATHENA study. Lancet Oncol. 12, 880–890. https://​doi.​
org/​10.​1016/​s1470-​2045(11)​70188-7 (2011).
49. Stoler, M. H. et al. High-risk human papillomavirus testing in women with ASC-US cytology: Results from the ATHENA HPV
study. Am. J. Clin. Pathol. 135, 468–475. https://​doi.​org/​10.​1309/​ajcpz​5jy6f​cvnmot (2011).
50. Mahmud, S. M. et al. Comparison of human papillomavirus testing and cytology for cervical cancer screening in a primary health
care setting in the Democratic Republic of the Congo. Gynecol. Oncol. 124, 286–291. https://​doi.​org/​10.​1016/j.​ygyno.​2011.​10.​031
(2012).
51. Sangrajrang, S. et al. Comparative accuracy of Pap smear and HPV screening in Ubon Ratchathani in Thailand. Papillomavirus.
Res. 3, 30–35. https://​doi.​org/​10.​1016/j.​pvr.​2016.​12.​004 (2017).
52. Sangrajrang, S. et al. Human papillomavirus (HPV) DNA and mRNA primary cervical cancer screening: Evaluation and triaging
options for HPV-positive women. J. Med. Screen. 26, 212–218. https://​doi.​org/​10.​1177/​09691​41319​865922 (2019).
53. Kurokawa, T. et al. The ideal strategy for cervical cancer screening in Japan: Result from the Fukui Cervical cancer screening study.
Cytopathology 29, 361–367. https://​doi.​org/​10.​1111/​cyt.​12576 (2018).
54. Blumenthal, P. D. et al. Adjunctive testing for cervical cancer in low resource settings with visual inspection, HPV, and the Pap
smear. Int. J. Gynaecol. Obstet. 72, 47–53 (2001).
55. Visual inspection with acetic acid for cervical-cancer screening: Test qualities in a primary-care setting. University of Zimbabwe/
JHPIEGO Cervical Cancer Project. Lancet 353, 869–873 (1999).
56. Coste, J. et al. Cross sectional study of conventional cervical smear, monolayer cytology, and human papillomavirus DNA testing
for cervical cancer screening. BMJ 326, 733. https://​doi.​org/​10.​1136/​bmj.​326.​7392.​733 (2003).
57. de Cremoux, P. et al. Efficiency of the hybrid capture 2 HPV DNA test in cervical cancer screening. A study by the French Society
of Clinical Cytology. Am. J. Clin. Pathol. 120, 492–499. https://​doi.​org/​10.​1309/​xfuc-​pp6m-​5xua-​94b8 (2003).
58. Sankaranarayanan, R. et al. Accuracy of human papillomavirus testing in primary screening of cervical neoplasia: Results from a
multicenter study in India. Int. J. Cancer 112, 341–347. https://​doi.​org/​10.​1002/​ijc.​20396 (2004).
59. Qiao, Y. L. 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. 9, 929–936. https://​doi.​org/​10.​1016/​s1470-​2045(08)​70210-9 (2008).
60. McAdam, M., Sakita, J., Tarivonda, L., Pang, J. & Frazer, I. H. Evaluation of a cervical cancer screening program based on HPV
testing and LLETZ excision in a low resource setting. PLoS ONE 5, e13266. https://​doi.​org/​10.​1371/​journ​al.​pone.​00132​66 (2010).
61. Quincy, B. L., Turbow, D. J., Dabinett, L. N., Dillingham, R. & Monroe, S. Diagnostic accuracy of self-collected human papilloma-
virus specimens as a primary screen for cervical cancer. J. Obstet. Gynaecol. 32, 795–799. https://​doi.​org/​10.​3109/​01443​615.​2012.​
717989 (2012).
62. Cuzick, J. et al. Management of women who test positive for high-risk types of human papillomavirus: The HART study. Lancet
362, 1871–1876. https://​doi.​org/​10.​1016/​s0140-​6736(03)​14955-0 (2003).
63. Petry, K. U. et al. Inclusion of HPV testing in routine cervical cancer screening for women above 29 years in Germany: Results for
8466 patients. Br. J. Cancer 88, 1570–1577. https://​doi.​org/​10.​1038/​sj.​bjc.​66009​18 (2003).

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 12

Vol:.(1234567890)
www.nature.com/scientificreports/

64. Gravitt, P. E. et al. Effectiveness of VIA, Pap, and HPV DNA testing in a cervical cancer screening program in a peri-urban com-
munity in Andhra Pradesh, India. PLoS ONE 5, e13711. https://​doi.​org/​10.​1371/​journ​al.​pone.​00137​11 (2010).
65. Moy, L. M. et al. Human papillomavirus testing and cervical cytology in primary screening for cervical cancer among women in
rural China: Comparison of sensitivity, specificity, and frequency of referral. Int. J. Cancer 127, 646–656. https://​doi.​org/​10.​1002/​
ijc.​25071 (2010).
66. Monsonego, J. et al. Evaluation of oncogenic human papillomavirus RNA and DNA tests with liquid-based cytology in primary
cervical cancer screening: The FASE study. Int. J. Cancer 129, 691–701. https://​doi.​org/​10.​1002/​ijc.​25726 (2011).
67. Ferreccio, C. et al. Screening trial of human papillomavirus for early detection of cervical cancer in Santiago. Chile. Int. J. Cancer
132, 916–923. https://​doi.​org/​10.​1002/​ijc.​27662 (2013).
68. Agorastos, T. et al. Primary screening for cervical cancer based on high-risk human papillomavirus (HPV) detection and HPV
16 and HPV 18 genotyping, in comparison to cytology. PLoS ONE 10, e0119755. https://​doi.​org/​10.​1371/​journ​al.​pone.​01197​55
(2015).
69. Iftner, T. et al. Head-to-head comparison of the RNA-based aptima human papillomavirus (HPV) assay and the DNA-based hybrid
capture 2 HPV test in a routine screening population of women aged 30 to 60 years in Germany. J. Clin. Microbiol. 53, 2509–2516.
https://​doi.​org/​10.​1128/​jcm.​01013-​15 (2015).
70. Wu, Q. et al. A cross-sectional study on HPV testing with type 16/18 genotyping for cervical cancer screening in 11,064 Chinese
women. Cancer Med. 6, 1091–1101. https://​doi.​org/​10.​1002/​cam4.​1060 (2017).
71. Gray, J. A., Patnick, J. & Blanks, R. G. Maximising benefit and minimising harm of screening. BMJ 336, 480–483. https://​doi.​org/​
10.​1136/​bmj.​39470.​643218.​94 (2008).
72. Dillner, J. et al. Long term predictive values of cytology and human papillomavirus testing in cervical cancer screening: Joint
European cohort study. BMJ 337, a1754. https://​doi.​org/​10.​1136/​bmj.​a1754 (2008).
73. Katki, H. A. et al. Cervical cancer risk for women undergoing concurrent testing for human papillomavirus and cervical cytology:
A population-based study in routine clinical practice. Lancet Oncol. 12, 663–672. https://​doi.​org/​10.​1016/​s1470-​2045(11)​70145-0
(2011).
74. Kitchener, H. C. et al. A comparison of HPV DNA testing and liquid based cytology over three rounds of primary cervical screen-
ing: Extended follow up in the ARTISTIC trial. Eur. J. Cancer 47, 864–871. https://​doi.​org/​10.​1016/j.​ejca.​2011.​01.​008 (2011).
75. Luyten, A. et al. Early detection of CIN3 and cervical cancer during long-term follow-up using HPV/Pap smear co-testing and
risk-adapted follow-up in a locally organised screening programme. Int. J. Cancer 135, 1408–1416. https://​doi.​org/​10.​1002/​ijc.​
28783 (2014).
76. Tainio, K. et al. Clinical course of untreated cervical intraepithelial neoplasia grade 2 under active surveillance: Systematic review
and meta-analysis. BMJ 360, k499. https://​doi.​org/​10.​1136/​bmj.​k499 (2018).
77. Perkins, R. B. et al. 2019 ASCCP risk-based management consensus guidelines for abnormal cervical cancer screening tests and
cancer precursors. J. Low Genit. Tract. Dis. 24, 102–131. https://​doi.​org/​10.​1097/​lgt.​00000​00000​000525 (2020).
78. Wentzensen, N. et al. Multiple biopsies and detection of cervical cancer precursors at colposcopy. J. Clin. Oncol. 33, 83–89. https://​
doi.​org/​10.​1200/​jco.​2014.​55.​9948 (2015).
79. Brown, B. H. & Tidy, J. A. The diagnostic accuracy of colposcopy—A review of research methodology and impact on the outcomes
of quality assurance. Eur. J. Obstet. Gynecol. Reprod. Biol. 240, 182–186. https://​doi.​org/​10.​1016/j.​ejogrb.​2019.​07.​003 (2019).
80. Naaktgeboren, C. A. et al. Anticipating missing reference standard data when planning diagnostic accuracy studies. BMJ. https://​
doi.​org/​10.​1136/​bmj.​i402 (2016).
81. Cronin, A. M. & Vickers, A. J. Statistical methods to correct for verification bias in diagnostic studies are inadequate when there
are few false negatives: A simulation study. BMC Med. Res. Methodol. 8, 75–75. https://​doi.​org/​10.​1186/​1471-​2288-8-​75 (2008).
82. Menten, J. & Lesaffre, E. A general framework for comparative Bayesian meta-analysis of diagnostic studies. BMC Med. Res.
Methodol. 15, 70. https://​doi.​org/​10.​1186/​s12874-​015-​0061-7 (2015).
83. Wentzensen, N. et al. Clinical evaluation of human papillomavirus screening with p16/Ki-67 dual stain triage in a large organized
cervical cancer screening program. JAMA Intern. Med. 179, 881–888. https://​doi.​org/​10.​1001/​jamai​ntern​med.​2019.​0306 (2019).

Acknowledgements
We thank Dr. Alejandra Castanon (on behalf of Professor Thomas Iftner and Professor Peter Sasieni), Dr. Joel
Coste, and Dr. Tetsuji Kurokawa for the provision of the additional information on their original work; and
MARUZEN-YUSHODO Co., Ltd. (https://k​ w.m ​ aruze​ n.c​ o.j​ p/k​ ousei-h
​ onyak​ u/) for the English language editing.

Author contributions
T.T.: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project
administration, resources, software, supervision, validation, visualization, writing—original draft, writing—
review & editing. S.H.: conceptualization, investigation, methodology, validation, writing—review & editing.
S.S.: conceptualization, investigation, validation, writing—review & editing. K.H.: conceptualization, investi-
gation, validation, visualization, writing—review & editing. Y.H.: conceptualization, investigation, validation,
writing—review & editing. T.K.: conceptualization, investigation, validation, writing—review & editing. C.H.:
conceptualization, funding acquisition, investigation, methodology, project administration, resources, supervi-
sion, validation, writing—review & editing.

Funding
This work was supported by the National Cancer Center Research and Development Fund from the National
Cancer Center, Tokyo, Japan (Grant Numbers 26-A-30, 29-A-16); and a Grant-in-Aid for Scientific Research
from the Ministry of Education, Culture, Sports, Science, and Technology, Japan (Grant Number 26460755 to
TT and CH).

Competing interests 
The authors declare no competing interests.

Additional information
Supplementary Information The online version contains supplementary material available at https://​doi.​org/​
10.​1038/​s41598-​021-​04201-y.
Correspondence and requests for materials should be addressed to T.T.

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 13

Vol.:(0123456789)
 www.nature.com/scientificreports/

Reprints and permissions information is available at www.nature.com/reprints.

Publisher’s note  Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access  This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© The Author(s) 2022

Scientific Reports | (2022) 12:94 | https://doi.org/10.1038/s41598-021-04201-y 14

Vol:.(1234567890)