Cited by 852Cervical Cancer and HPV ResearchWomen's cancer prevention and managementReproductive tract infections research

Detecting cervical precancer and reaching underscreened women by using HPV testing on self samples: updated meta-analyses

Open access
Arbyn, Marc, Smith, Sara B, Temin, Sarah, Sultana, Farhana, Castle, Philip·2018·BMJ

Abstract

Objective: To evaluate the diagnostic accuracy of high-risk human papillomavirus (hrHPV) assays on self samples and the efficacy of self sampling strategies to reach underscreened women. Design: Updated meta-analysis. Data sources: Medline (PubMed), Embase, and CENTRAL from 1 January 2013 to 15 April 2018 (accuracy review), and 1 January 2014 to 15 April 2018 (participation review). Review methods: Accuracy review: hrHPV assay on a vaginal self sample and a clinician sample; and verification of the presence of cervical intraepithelial neoplasia grade 2 or worse (CIN2+) by colposcopy and biopsy in all enrolled women or in women with positive tests. Participation review: study population included women who were irregularly or never screened; women in the self sampling arm (intervention arm) were invited to collect a self sample for hrHPV testing; women in the control arm were invited or reminded to undergo a screening test on a clinician sample; participation in both arms was documented; and a population minimum of 400 women. Results: 56 accuracy studies and 25 participation trials were included. hrHPV assays based on polymerase chain reaction were as sensitive on self samples as on clinician samples to detect CIN2+ or CIN3+ (pooled ratio 0.99, 95% confidence interval 0.97 to 1.02). However, hrHPV assays based on signal amplification were less sensitive on self samples (pooled ratio 0.85, 95% confidence interval 0.80 to 0.89). The specificity to exclude CIN2+ was 2% or 4% lower on self samples than on clinician samples, for hrHPV assays based on polymerase chain reaction or signal amplification, respectively. Mailing self sample kits to the woman’s home address generated higher response rates to have a sample taken by a clinician than invitation or reminder letters (pooled relative participation in intention-to-treat-analysis of 2.33, 95% confidence interval 1.86 to 2.91). Opt-in strategies where women had to request a self sampling kit were generally not more effective than invitation letters (relative participation of 1.22, 95% confidence interval 0.93 to 1.61). Direct offer of self sampling devices to women in communities that were underscreened generated high participation rates (>75%). Substantial interstudy heterogeneity was noted (I 2 >95%). Conclusions: When used with hrHPV assays based on polymerase chain reaction, testing on self samples was similarly accurate as on clinician samples. Offering self sampling kits generally is more effective in reaching underscreened women than sending invitations. However, since response rates are highly variable among settings, pilots should be set up before regional or national roll out of self sampling strategies.

AI snapshot

Written from the paper’s full text.

TL;DR

In updated meta-analyses of 56 diagnostic accuracy studies and 25 randomized participation trials, hrHPV testing on self samples was as sensitive as clinician samples when using PCR-based assays, but less sensitive with signal-amplification assays. Mailing self-sampling kits increased participation among underscreened women, though effects varied substantially by invitation strategy and setting.

Key findings

  • PCR-based hrHPV assays on self samples were as sensitive as clinician samples for CIN2+ (pooled ratio 0.99, 95% CI 0.97 to 1.02) and CIN3+ (0.99, 0.96 to 1.02), with slightly lower specificity to exclude CIN2+ (0.98, 0.97 to 0.99).
  • Signal-amplification hrHPV assays were less sensitive on self samples for CIN2+ (0.85, 0.80 to 0.89) and CIN3+ (0.86, 0.76 to 0.98), and less specific to exclude CIN2+ (0.96, 0.93 to 0.98).
  • In intention-to-treat analysis, mailing self-sampling kits to the home increased participation versus invitation/reminder letters (relative participation 2.33, 95% CI 1.86 to 2.91), whereas opt-in strategies were not significantly more effective (1.22, 0.93 to 1.61).
  • Door-to-door delivery generated high participation (>75%; pooled self-sampling participation 94.2% per protocol and 94.6% intention-to-treat), and CIN2+ detection per 1000 invited women was higher in self-sampling arms (relative proportion 2.28, 95% CI 1.44 to 3.61).

Methods

  • Updated meta-analyses: accuracy review searched Medline, Embase, and CENTRAL from 1 January 2013 to 15 April 2018; participation review searched from 1 January 2014 to 15 April 2018.
  • Included 56 diagnostic test accuracy studies comparing hrHPV testing on vaginal self samples versus clinician samples with CIN2+/CIN3+ verification, and 25 randomized trials of self-sampling offers among irregularly/never-screened women with at least 400 women.
  • Pooled accuracy using a bivariate normal model for logit-transformed sensitivity/specificity; pooled participation using random-effects models for proportions and ratios; assessed quality with QUADAS-2 and Cochrane risk-of-bias tools.

Limitations

  • Stated: meta-analyses based on aggregated published results lacked detailed data equally stratified by potentially influencing factors; meta-regression including only eight studies did not reveal age effects, and subgroup analyses/meta-regressions were not possible because covariates were categorized differently across studies.
  • Stated: the authors propose pooling individual patient data from the best trials to evaluate influential factors more efficiently.
  • Evident: substantial interstudy heterogeneity was noted (I²>95%), including high heterogeneity for participation outcomes and opt-in results (P<0.001), and quality reporting was incomplete for some domains, such as uninterpretable results poorly reported for evaluated tests (39%) and reference standard (45%).

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1136/bmj.k4823

Cites

130 of 130 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

14 of 33 cited sources checked · 6 supported · 0 flagged · accuracy not yet measured

  • Findingcitation

    Participation was not significantly different when hrHPV testing was performed versus when cytology was performed in two Italian studies (P=0.60 v P=0.76).

    “The participation was not significantly different when hrHPV testing was performed versus when cytology was performed (P=0.60 v P=0.76).104 117”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [104] source

      The Results (‘Effect on compliance’; page unavailable) report similar participation after invitations for clinic-based HPV testing (14.9%) and standard Pap-test recall (13.9%). The Abstract (‘Results’; page unavailable) explicitly calls the difference nonsignificant. Our copy does not report the cited P=0.60 or P=0.76 values; this paper also accounts for only one of the two studies cited. (LLM, conf=0.83)

      Experimental — accuracy not yet measured
    • cites [117] source

      The Results (page unavailable) report no difference in participation between the cytology and HPV clinic-control groups at one centre. However, the reported P value is 0.9, not 0.60 or 0.76; this source does not support the quoted figures or the claim about two studies. (LLM, conf=0.88)

      Experimental — accuracy not yet measured
  • Premisecitation

    A proportion of screened women with cervical precancer and cancer show false negative Pap tests.

    “Third, a proportion of screened women with cervical precancer and cancer show false negative Pap tests.18 19”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [18] source
    • cites [19] Koliopoulos, 2017, Cytology versus HPV testing for cervical cancer screening in the general population, Cochrane Database Syst Rev source
  • Findingcitation

    In the Belgian randomized clinical trial, the response rate was 10.5% (95% CI 9.9% to 11.2%) in the control arm with a reminder letter.

    “In the first control arm (reminder letter) the response rate was 10.5% (95% confidence interval 9.9% to 11.2%)”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [69] source

      The Participation section reports that 931 of 8,830 women (10.5%) in the recall-letter control group had a Pap smear within 12 months. Our copy does not report the claimed 95% confidence interval of 9.9% to 11.2%, although it is consistent with the reported counts. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Study design and conductcitation

    Three (12%) of the 25 trials met the criteria for low risk of bias in all categories.

    “Three (12%) of the 25 trials met the criteria for low risk of bias in all categories.114 115 117”
    Quote in the paperNot a substantial claim3 of 3 sources checked
    • cites [114] source

      The Design (p. 2) identifies this source as a randomized trial, and Randomization and blinding (p. 4) describes its allocation and blinding procedures. The paper does not assess 25 trials or establish that three met low-risk-of-bias criteria in every category; it supports identifying one trial, not the aggregate claim. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites [115] source

      The Materials and Methods > Design section describes this paper as a single three-group randomised trial, not an assessment of 25 trials. Searches for “25 trials” and “low risk of bias” found no passage in our copy reporting that three trials (12%) met low-risk criteria in all categories. Page numbers were not available in the indexed passages. (LLM, conf=0.77)

      Experimental — accuracy not yet measured
    • cites [117] source

      The Methods (Analysis and Study design; page unavailable) describe this paper’s own randomized trial, not an assessment of 25 trials. Searches for “25 trials,” “low risk of bias,” and “three trials 12%” found no passage reporting the claimed count or bias rating. (LLM, conf=0.75)

      Experimental — accuracy not yet measured
  • Findingcitation

    In the Belgian randomized clinical trial, the response rate was 8.0% (95% CI 7.5% to 8.6%) in the control arm with no letter.

    “whereas in the second control arm (no letter), the response rate was slightly lower (8.0%, 95% confidence interval 7.5% to 8.6%)”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [69] source

      The Participation section reports that 709 women (8%) in the no-intervention arm had a PAP smear within 12 months, and the study description identifies this as the second control group. Our copy does not report the claimed 95% confidence interval of 7.5%–8.6% for that rate; the Statistical analysis section discusses confidence intervals for odds ratios instead. Page numbers are not available for these passages. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Classified as probably not a substantial claim. 9
  • May combine two claims. 19
  • The quote only partly states this claim. 2
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    The authors assessed the quality of randomized clinical trials for the participation review by using the Cochrane Collaboration's tool for risk of bias in randomized trials.

    “We assessed the quality of the randomized clinical trials for the participation review by using the Cochrane Collaboration’s tool for risk of bias in randomized trials.30”
    Quote in the paperNot a substantial claim1 of 1 sources checked
  • May combine two claims.

    Findingown data

    Each hrHPV assay based on polymerase chain reaction was equally sensitive for CIN2+ and for CIN3+ on self samples versus clinician samples.

    “Each hrHPV assay based on polymerase chain reaction was equally sensitive for CIN2+ and for CIN3+ on self samples versus clinician samples.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The updated meta-analyses comprised 56 diagnostic test accuracy studies and 25 randomized trials.

    “56 accuracy studies and 25 participation trials were included.”
    • Also stated in: § Results · plain
      “The updated meta-analyses finally comprised 56 diagnostic test accuracy studies and 25 randomized trials.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingown data

    In general, hrHPV assays were less specific on self samples except for the HC2 assay on a tampon self sample (n=1) or an assay based on polymerase chain reaction on a vaginal lavage self sample (2).

    “In general, hrHPV assays were less specific on self samples except for the HC2 assay on a tampon self sample (n=1) or an assay based on polymerase chain reaction on a vaginal lavage self sample (2).”
    Quote in the paperSubstantial claim
  • The quote only partly states this claim.

    Findingown data

    For signal amplification assays, the test positivity rate was on average 14% higher on self samples than clinician samples (ratio 1.14, 95% CI 1.05 to 1.24).

    “The test positivity rate was, on average, 14% higher and the positive predictive value was significantly lower for both CIN2+ and CIN3+ for self samples (positive predictive value< 1).”
    Quote in the paperNot a substantial claim
  • The quote only partly states this claim.

    Findingown data

    For PCR assays, the test positivity rate was similar on self samples and clinician samples (ratio 1.00, 95% CI 0.94 to 1.06).

    “The test positivity rate was similar in self samples versus clinician samples.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingown data

    The participation difference was nearly always positive except for two studies where the difference was significantly lower than zero.

    “The participation difference was nearly always positive except for two studies where the difference was significantly lower than zero.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingown data

    In the intention-to-treat analysis of trials with opt-in scenarios, the pooled relative participation exceeded unity, although not significantly (1.22, 95% CI 0.93 to 1.61).

    “whereas on average in the intention-to-treat analysis, the pooled relative participation exceeded unity, although not significantly (1.22, 0.93 to 1.61)”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Study design and conductown data

    The quality of verification with a reference standard was good in 84%, moderate in 13%, and problematic in 4% of the studies.

    “The quality of verification with a reference standard (acceptable validity, blinding toward tests, and avoidance of incorporating test results in final conclusion of disease outcome) was good in 84%, moderate in 13%, and problematic in 4% of the studies.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Study design and conductown data

    Reporting and execution of tests was adequate in 73% and unclear in 27% of the 56 diagnostic studies, and risk of bias was never assessed as high.

    “Reporting and execution of tests (description of cut-off, blinding of the index test toward comparator and reference test) was adequate in 73% and unclear in 27%, but the risk of bias was never assessed as high.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Study design and conductown data

    The risk of bias for enrolment of women in the 56 diagnostic studies was considered low in 59%, moderate in 39%, and high in 2%.

    “The risk of bias for enrolment of women was considered low in 59%, moderate in 39%, and high in 2% of the 56 diagnostic studies (supplementary materials).”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Study design and conductown data

    Withdrawal of patients was explained appropriately in 70%, but not in 18%, and unclear in 12% of the studies.

    “Withdrawal of patients was explained appropriately in 70%, but not in 18%, and unclear in 12% of the studies.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The authors assessed the quality of diagnostic studies for the accuracy review by using the QUADAS-2 check list.

    “We assessed the quality of the diagnostic studies for the accuracy review by using the QUADAS-2 check list.29”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [29] source
  • May combine two claims.

    Study design and conductown data

    The delay between self sampling, clinician sampling, and verification with the reference standard was short in 61%, unreported in 38%, and long in 2% of the studies.

    “The delay between self sampling, clinician sampling, and verification with the reference standard was short in 61%, unreported in 38%, and long in 2% of the studies.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    The meta-analyses followed PRISMA guidelines for reporting of meta-analyses, and recommendations established by the Cochrane Collaboration for diagnostic test accuracy and intervention trials.

    “The meta-analyses followed PRISMA guidelines for reporting of meta-analyses, and recommendations established by the Cochrane Collaboration for diagnostic test accuracy and intervention trials.31 32”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [31] Moher, 2009, Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement, Ann Intern Med source
    • cites [32] Deeks, 2013

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Study design and conductown data

    All trials had complete data for the participation outcome and were therefore at low risk of attrition bias.

    “All trials had complete data for the participation outcome and were therefore at low risk of attrition bias.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Recent trend analyses reveal an increasing burden of cervical cancer, even in countries with well organized screening programs based on cytology and good coverage.

    “However, recent trend analyses reveal an increasing burden of cervical cancer, even in countries with well organized screening programs based on cytology and good coverage.8 9 10 11 12”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [8] source
    • cites [9] source
    • cites [10] Finnish Cancer Registry. Cancer Statistics. Cancer Society of Finland. https://cancerregistry.fi/statistics/cancer-statistics/

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    Publication bias was assessed by using Deeks' and Harbord's regression tests for the pooled absolute and relative accuracy estimates, respectively.

    “We assessed publication bias by using Deeks' and Harbord's regression tests for the pooled absolute and relative accuracy estimates, respectively.36 37”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [36] source
    • cites [37] Harbord, 2009, Updated tests for small-study effects in meta-analyses, Stata J source
  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    The previous accuracy meta-analysis concluded that hrHPV testing was less sensitive on self samples than on clinician samples, but also found that the reason for lower sensitivity was the use of assays based on signal amplification.

    “The previous accuracy meta-analysis concluded that hrHPV testing was less sensitive on self samples than on clinician samples, but also found that the reason for lower sensitivity was the use of assays based on signal amplification.22”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [22] source
  • May combine two claims.

    Study design and conductown data

    Partial verification was avoided in 70% but clearly present in 25%, whereas differential verification was absent in 89% but unclear or present in 11% of the studies.

    “Partial verification was avoided in 70% but clearly present in 25%, whereas differential verification was absent in 89% but unclear or present in 11% of the studies.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Premisecitation

    hrHPV testing can be done on vaginal samples collected by the patient (self samples), whereas cytology on self samples shows poor accuracy.

    “hrHPV testing provides another advantage: it can be done on vaginal samples collected by the patient (self samples), whereas cytology on self samples shows poor accuracy.21”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [21] source
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    MA and one other author independently performed the study selection and data extraction, and PC judged any unresolvable discordances.

    “MA and one other author (SBS, ST, FS, or a Collaboration on Self-Sampling and HPV Testing group member) independently performed the study selection and data extraction. PC judged any unresolvable discordances.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Premisecitation

    Most cervical cancer cases occur in women who have never been screened for cervical cancer, or do not participate in routine screening.

    “Most cervical cancer cases occur in women who have never been screened for cervical cancer, or do not participate in routine screening.7”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [7] IARC, 2005

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    The quality of reporting was adequate in 12 trials (48%), incomplete in one trial, and intermediate in the other trials.

    “The quality of reporting was adequate in 12 trials (48%),26 67 102 103 104 107 108 109 114 115 116 117 118 incomplete in one trial,106 and intermediate in the other trials.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Study design and conductown data

    In eight trials (32%) allocation was random and concealed, in 14 trials (56%) allocation was unclear, and in three trials (12%) allocation was problematic.

    “In eight trials (32%) allocation was random and concealed,26 63 64 65 104 114 115 117 in 14 trials (56%) allocation was unclear,66 67 102 103 105 106 107 108 109 110 111 112 113 116 and in three trials (12%) allocation was problematic.62 68 69”
    Quote in the paperNot a substantial claim
25 of 25
  • The paper relies on this without support.

    Intention-to-treat analyses of participation trials reflect the overall public health effect in a real world situation.

    “In the intention-to-treat analyses, which reflected the overall public health effect in a real world situation, we additionally included women in the experimental arm who had been offered self sampling but choose to have a clinician sample taken instead.”
  • The paper relies on this without support.

    New assays, self sampling devices, storage media, and participation strategies have entered the market since the publication of the two previous meta-analyses, and more accuracy and participation studies have been conducted.

    “New assays, self sampling devices, storage media, and participation strategies have entered the market since the publication of these two meta-analyses; and more accuracy and participation studies have been conducted.”
2 of 2

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Vaginal-lavage study count differs between text and table

    Minor

    The text gives a count of two for polymerase-chain-reaction testing on vaginal lavage self samples, but table 2 lists four studies for that category.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. IARC, 2005, Cervix Cancer Screening. IARC Handbooks of Cancer Prevention. Vol. 10
  2. Deeks, 2013, Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy Version 1.0.0.

Cited by

Sorted by citations

Authors
Arbyn, M., Smith, S. B., Temin, S., Sultana, F., & Castle, P.
Affiliations
  • Sciensano (Belgium) (BE)
  • American Society of Clinical Oncology (US)
  • The University of Melbourne (AU)
  • Albert Einstein College of Medicine (US)
Corresponding author
Arbyn, Marc
Venue
BMJ
Publisher
BMJ
Published
2018-12-05
Volume
363
Pages
k4823
ISSN
  • 1756-1833 (ISSN-L)
  • 0959-8138
OpenAlex
W2902810541
Cited by
851
References
130
Citations by year
2026: 93 · 2025: 136 · 2024: 136 · 2023: 137 · 2022: 116 · 2021: 104 · 2020: 89 · 2019: 39
FWCI
39.55
Type
journal article
Language
en
Open access
Locations
  • BMJ (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (submittedVersion, other-oa)
  • PubMed Central (submittedVersion, cc-by-nc)
  • HighWire Press Open Archive (submittedVersion)
Topics
  • Health Sciences › Medicine › Epidemiology › Cervical Cancer and HPV Research
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Women's cancer prevention and management
  • Life Sciences › Immunology and Microbiology › Microbiology › Reproductive tract infections research
Keywords
cervical cancer screening, HPV testing, self-sampling, cervical intraepithelial neoplasia, CIN2+ detection, diagnostic accuracy, underscreened women, screening uptake, reminder letters, clinician-collected samples, community-based screening
MeSH terms
Uterine Cervical Dysplasia, Uterine Cervical Neoplasms, Colposcopy, Female, Humans, Mass Screening, Sensitivity and Specificity, Specimen Handling, Vagina, Vaginal Smears, Incidence, Randomized Controlled Trials as Topic, Papillomaviridae, Papillomavirus Infections, Early Detection of Cancer, Intention to Treat Analysis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-16.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-16.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review