Self-collected versus clinician-collected samples for HPV testing: A meta-analysis revealing lower sensitivity for detecting high-grade cervical lesions

Lima, Gabriel Gonçalves Santiago; Costa, Annielson de Souza; Alves, Mayara Souza; Turri, Jose Antonio Orellana; Zuchelo, Lea Tami Suzuki; Lorenzi, Noely Paula Cristina; Veiga, Eduardo Carvalho de Arruda; Sorpreso, Isabel C.E.; Baracat, Edmund C.; Soares Junior, José Maria

doi:10.1016/j.clinsp.2026.100898

Abstract

Objective To compare the specificity and sensitivity between self-collection and collection performed by healthcare professionals in testing High-risk oncogenic Human Papillomavirus (HrHPV) to detect high-grade uterine cervical lesions.

Methods A systematic review was conducted following PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). Meta-analysis was conducted on STATA-14SE using random effects models for the aggregated values of sensitivity and specificity of studies that evaluated self-collection and collection by healthcare professionals. This review is registered on PROSPERO under CRD42022347731.

Results This study included 17 articles. Regarding the sensitivity of the collection method, self-collection was superior to professional collection in 2 studies, inferior in 8, equivalent in 1, and variable depending on the sample storage medium, HPV DNA analysis method, or CIN grade assessed in 3. The specificity between the methods was higher with self-collection in 5 studies, lower in 5 others, equivalent in 1, not evaluated in 1, and variable depending on the HPV DNA detection method in 2 articles. The results of the meta-analysis of studies that evaluated the sensitivity and specificity of devices used for HPV detection between self-collection and health-professionals collection showed significant differences of sensitivity favorable to the health-professionals (SMD = -0.540; 95% CI -0.595 ‒ -0.486; p < 0.001). The evaluation of the sensitivity and specificity of the type of lesion between self-collection and health professionals’ collection showed significant differences in sensitivity, favorable to the health professionals (SMD = -0.872; 95% CI -1.392 ‒ -0.351; p = 0.001).

Conclusion In conclusion, while specificity is comparable, this meta-analysis demonstrates that self-collected samples have significantly lower sensitivity for detecting HR-HPV and high-grade cervical lesions compared to clinician-collected samples. This suggests that self-sampling, in its current form, may be best utilized with complementary triage strategies or in settings where its logistical advantages outweigh the modest reduction in sensitivity for increasing screening coverage.

Keywords
Self-Collection; Human papillomavirus (HPV); Sensitivity and specificity; Diagnostic test accuracy; Meta-analysis; Cervical cancer screening; Cervical intraepithelial neoplasia

Introduction

Cervical cancer is still a significant public health issue and one of the leading causes of cancer-related mortality in adult women. According to current data, it is ranked 14th among all cancers and is the 4th most common cancer among women worldwide. Cervical cancer intervention focuses on prevention, such as HPV vaccination and early detection screening, which are both the most appropriate methods to decrease the burden of cervical cancer and mortality. However, its prevalence remains high despite the development of new screening techniques.¹,²

In some developing countries, as in Brazil, where barriers to prevention and difficulties in accessing screening methods are substantial, cervical cancer is a major concern among women due to limited healthcare access and resource scarcity for diagnosis and treatment.²,³ Oncological cervical cytology using the Papanicolaou method has reduced incidence and mortality from cervical cancer, and it is directly related to the intensity of screening.³ The limited and highly variable sensitivity of cytology strongly supports the transition to the more robust and reproducible HPV-based cervical screening to ensure progress towards global cervical cancer elimination targets in Latin America.⁴

Molecular testing for HPV genotypes, compared to oncological cervical cytology, offers high feasibility, greater objectivity in result interpretation, high sensitivity, a high negative predictive value for virus detection, long-term risk stratification capability, less subjective results among testing centers,⁴,⁵ and attractive specificity for detecting HrHPV. Thus, molecular tests are a viable option for detecting women at high-grade lesion risk in population-based screening programs.⁶,⁷

Self-sampling improves screening uptake, particularly in low-income countries. In fact, self-sampling is more cost-effective than supplier-required visual inspection, although testing and running costs are higher for HPV self-sampling. However, comparisons between specificity and sensitivity between self-collection and healthcare professional collection principaly in detecting high-grade endocervical lesions, have not been made.⁵-⁷ The objective of this review is to compare the specificity and sensitivity between self-collection and collection performed by healthcare professionals in testing High-risk oncogenic Human Papillomavirus (HrHPV) to detect high-grade uterine cervical lesions.

Method

This is a systematic review conducted in accordance with the guidelines of PRISMA⁸ (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and MOOSE (Meta-analyses Of Observational Studies in Epidemiology). The data were evaluated for comparison by the Discipline of Gynecology at the Faculty of Medicine of the University of São Paulo between 2021 and 2024. This review is registered on PROSPERO under CRD42022347731.

Databases of PubMed, Embase, and Web of Science were used to select articles. The P.I.C.O (Population, Intervention, Comparison, and Outcome) statement was designated as it follows: “In women eligible for cervical cancer screening (P), are the self-collected samples (I), compared with cervical samples taken by healthcare specialists (C), accurate in screening precancerous cervical lesions (O)?” The keywords used in the search were developed based on the P.I.C.O strategy, which included: “HPV” “self-collection”, “self-sampling”, “vagina smear”, “papanicolaou test”, “pap smear”, “pap stain”, “pap test”, “papanicolaou cytology”, “papanicolaou smear”, “papanicolaou method”, “papanicolaou stain”, “efficacy”, “sensitivity” and specificity”.

Selected articles involved sexually active women, with no age limits established (Population), who performed self-collection of vaginal samples for the detection of HrHPV (Intervention), as well as cervical samples collected by a healthcare professional (Comparison) to compare with their self-collected samples. The outcome studied was the accuracy of the diagnostic test, assessed through the analysis of specificity and sensitivity, comparing tests for HrHPV detection performed on both samples.

The selection process was conducted in three stages: title reading, abstract reading, and full study reading. To minimize potential biases and enhance the reliability of the process, another author was consulted when disagreements arose regarding the selection of articles. The flowchart illustrating the study screening process is presented in Fig. 1.

Fig. 1
Flowchart demonstrating the screening process.

The inclusion criteria for the studies were as follows: articles in English that compared the effectiveness of genitourinary samples collected by the patient and cervical samples collected by a healthcare professional in detecting cervical intraepithelial neoplasias Grade 2 or higher, i.e., CIN2 or CIN3, using histopathological results from cervical biopsies guided by colposcopy as the gold standard.

Studies were excluded if they were in a language other than English, published as abstracts, editorials, reviews, or meta-analyses, if they lacked methodological details, failed to report sensitivity or specificity, or were not freely available in the literature. There were no limitations regarding the date of publication or the size of the sample.

The assessment of the Risk of Bias, conducted using the QUADAS-2 tool and detailed in Table 1, indicates that none of the studies were at high risk of bias. Given their nature as observational studies, they were all deemed to provide low levels of evidence. There were no identified factors that would necessitate a downgrade or upgrade in the quality of evidence. Consequently, the inclusion of women referred due to altered Pap smear results did not pose any obstacles to the overall level of evidence in the study.

Thumbnail

Table 1
QUADAS-2 evaluation of bias of the eligible studies.

Data were extracted from the studies regarding the HrHPV genetic material analysis method, the devices employed for self-collection, and the classification of cervical lesions. These data were used to compare self-collection with collection performed by healthcare professionals, utilizing these specific parameters (Table 2).

Thumbnail

Table 2
Technical features of the eligible studies.

Meta-analysis was conducted using the aggregated values of sensitivity and specificity for the results of studies that evaluated self-collection and collection by healthcare professionals in outpatient settings. The aggregated results of the meta-analysis assessed the weighted mean difference between the mean values and standard deviations for each type of examination that evaluated HPV, divided by the type and severity of the assessed cervical lesion, the devices used for collection, and the methodology of each type of test. Random-effects models were used for the meta-analysis, and p-values <0.05 were considered significant, using STATA 14-SE software (Stata corp, Texas, USA) for all analyses (Fig. 2).

Fig. 2
Meta analysis. (A) Forest-Plot of sensitivity and specificity of detection methods used for HPV detection between self-collection and healthcare professionals. (B) Forest-Plot of sensitivity and specificity of detection methods used for HPV detection by type of lesions between self-collection and health professionals. (C) Forest-Plot of sensitivity and specificity of detection methods used for HPV detection by type of device between self-collection and healthcare professionals.

Results

A total of 474 articles were selected from database searches; 335 remained after removing duplicates. Following a review of titles and abstracts, 225 studies were excluded based on inclusion and exclusion criteria. After full-text reading of 110 articles, 17 were chosen for the systematic review. The technical details of these studies are shown in Table 3.

Thumbnail

Table 3
Technical features of the eligible studies.

The studies were published between 2000 and 2023 and were conducted in the following countries: Brazil, Canada, China, Denmark, Germany, Haiti, India, Italy, South Korea, Spain, Sweden, Switzerland, Tanzania, and the Netherlands. The number of participants ranged from 91 to 4658, totaling 11.014 women, aged between 16- and 76-years. The devices used for self-collection included swabs in 4 studies; brushes in 10 studies; vaginal lavage devices in 2 studies; and urine in 3 studies. Regarding the storage medium for self-collected samples, most studies used a liquid cell preservation method, except for the studies by Catarino et al. Geraets et al. and Shao-Ming Wang et al. with the latter conducting a comparison between liquid and dry mediums.

The most commonly used test for HPV DNA detection was the Hybrid Capture II (HC2) HPV assay (Qiagen Gaithersburg, Inc., USA), followed, in decreasing order of usage, by the following tests: careHPV test (Qiagen Inc., Gaithersburg, MD, USA), GP5+/6+-EIA kit (Diassay BV, Rijswijk, The Netherlands), SPF10-DEIA-LiPA25 version 1 (Diassay BV, Rijswijk, The Netherlands), COBAS 4800 (Roche Inc, Basel, Switzerland), Anyplex™ II HPV Kit (Seegene Inc, South Korea), RealTime High Risk (HR) HPV assay (Abbott Inc. Illinois, USA), HPVDNAchip (Biomedlab Co., South Korea), Xpert HPV assay (Sunnyvale, CA, USA),⁹ Polymerase Chain Reaction (PCR) reverse line blot assay, and the Aptima HPV mRNA test (Hologic Inc, MA, USA).

The information extracted regarding the effectiveness of the methods in detecting high-grade lesions, i.e., sensitivity and specificity of each method of HrHPV detection, is presented in Table 4.

Thumbnail

Table 4
Analyses of mean, Standard Deviation (SD), Confidence Intervals (95 % CI), and p (Mann-Whitney test for non-parametric samples) are conducted for the 17 analyzed studies, comparing the Collection Devices (A), hrHPV Detection Method (B) and Lesion Types (C) regarding Sensitivity (SE) and Specificity (SP).

Four studies showed lower effectiveness of self-collection compared to healthcare professional collection, both in terms of sensitivity and specificity of the analysis method or sampling device.¹⁰,²²,²⁴,²⁵ The study by Labani et al.¹⁷ also showed decreased effectiveness in terms of sensitivity, although both collection methods were highly specific. Leeman et al.¹⁶ showed the same sensitivity in both collection methods, with only self-collection having higher specificity.

In the study of Sang-Soo Seo et al., sensitivity was higher in self-collection than in healthcare professional collection, but with lower specificity. In contrast, 3 studies showed higher specificity of self-collection with lower sensitivity.¹¹,¹⁶,²⁴

Bhatla et al.¹⁰ separated their data according to the HPV DNA detection method used; with the use of the HC2 HPV assay, both sensitivity and specificity of self-collection were lower. However, when using the PCR reverse line blot assay method, self-collection showed superior specificity.

The study by Shao-Ming Wang et al.¹² separated sensitivity and specificity values based on the storage methods used; with the use of a liquid medium, both sensitivity and specificity of self-collection were higher than those of healthcare professional collection. However, with the use of a dry medium, specificity was higher with self-collection, but sensitivity was reduced.

Asciutto et al.¹⁸ evaluated the performance of the Aptima mRNA test, being the only study in this review to use mRNA for HPV testing. Their findings revealed overall lower sensitivity and specificity for self-collection.

Hyun-Woong Cho et al. (2020)²⁴ evaluated the performance of urine and vaginal samples in two different HPV detection devices. They found that self-collected samples generally had lower sensitivity and specificity compared to clinician-collected samples. However, the urine sample analyzed with the AnyPlex II PCR Kit showed higher specificity.

Finally, 4 studies separated the efficacy data based on the severity of the lesion found in the biopsy.¹¹,¹⁷,²⁰,²³ Boggan et al.²⁰ conducted their analysis considering only sensitivity, which was lower for self-collection for CIN2+ lesions and higher for this method when evaluating CIN3+ lesions.

Results of meta-analysis

Sensitivity and specificity by detection method

The results of the meta-analysis of studies that evaluated the sensitivity and specificity of detection methods of HrHPV between self-collection and health-professionals collection showed significant differences of sensitivity favorable to the health-professionals (SMD = −0.711; 95 % CI −0.889 ‒ −0.533; p < 0.001), and no difference in the specificity. (Table 5A and Fig. 2A).

Thumbnail

Table 5A
Meta-analysis of sensitivity and specificity of detection methods used for HPV detection between self-collection and healthcare professionals.

Sensitivity and specificity by type of lesion

The results of the meta-analysis of studies that evaluated the sensitivity and specificity of the type of lesion between self-collection and health professionals’ collection showed significant differences in sensitivity, favorable to the health professionals (SMD = −0.872; 95 % CI −1.392 ‒ −0.351; p = 0.001), and no difference in specificity (Table 5B and Fig. 2B).

Thumbnail

Table 5B
Meta-analysis of sensitivity and specificity of detection methods used for HPV detection by type of lesions between self-collection and healthcare professionals.

Sensitivity and specificity by type of device

The results of the meta-analysis of studies that evaluated the sensitivity and specificity of devices used for HPV detection between self-collection and health-professionals collection showed significant differences of sensitivity favorable to the health-professionals (SMD = −0.540; 95 % CI −0.595 ‒ −0.486; p < 0.001), and no difference in the specificity (Table 5C and Fig. 2C).

Thumbnail

Table 5C
Meta-analysis of sensitivity and specificity of detection methods used for HPV detection by type of device between self-collection and healthcare professionals.

Discussion

Cervical cancer remains a significant public health concern, especially in developing countries with limited access to screening. This review aimed to compare the accuracy of self-collected samples versus healthcare professional-collected samples in detecting High-risk HPV (HrHPV) for cervical precancerous lesions, offering a perspective that contributes to the decision to include self-collection in screening programs.

This review has limitations due to the inclusion of only freely available literature. Although this criterion excluded numerous studies, the meta-analysis benefits from a large sample size (n = 11,014), strengthening the evaluation of self-collection performance.

The seventeen eligible studies displayed a wide array of self-collection devices and high-risk HPV detection methods. Nonetheless, significant differences in the sensitivity of the analyzed parameters were discernible.

Brush and swab methods demonstrated minimal differences in sensitivity and specificity, with swabs exhibiting higher mean sensitivity in self-collected samples versus those collected by healthcare professionals. Studies conducted by Catarino et al.,⁹ Sang-Soo Seo et al.,¹⁴ and Bogan et al.¹⁹ indicated greater sensitivity in self-collected samples compared to those collected clinically, whereas Leeman et al.¹⁶ found equivalent sensitivity in both methods of collection. This could be attributed to the selection of symptomatic patients or those previously diagnosed with HPV-related precursor lesions, which may affect the concordance of the laboratory method employed.

The p-values for Heterogeneity obtained from the Meta-Analysis indicate a significant difference in sensitivity concerning the HrHPV detection method, lesion type, and device used for self-collection, favoring the cervical collection by health-care professionals, suggesting that the Papanicolaou smear is more sensitive in detecting precancerous lesions. However, there is no significant difference in specificity between the two sampling methods.

Recent studies, such as the Dutch study by Polman et al. in 2019,²⁶ with a large number of female participants, reported recognized and substantial variation in clinical accuracy when using different brushes or devices for self-collection, as well as highlighting the bias in the accuracy of different HPV DNA detection tests. In a recent study, Lorenzi et al.²¹ used a clinically validated and FDA-approved HrHPV detection test and a widely used self-collection device in numerous studies, internationally validated, associated with double staining for p16/Ki-67, cytological, in self-collected samples. In both aforementioned studies, HrHPV detection in samples from both groups, self-collection and healthcare professional collection, was similar in terms of sensitivity and specificity for detecting CIN2+ lesions, with no inferior performance of the self-collection group compared to the healthcare professional group.

The performance of cytology to detect cervical precancer and cancer has been studied largely among women from Europe and North America. There is ample evidence and agreement that HPV testing is a more sensitive and reproducible test for primary screening than cytology, which allows extension of screening intervals and can be self-collected. WHO guidelines recommend HPV testing with or without triage as the preferred alternative for cervical cancer screening. The limited accuracy of cytology, which remains the main screening method in Latin American countries, together with limited coverage and inadequate follow-up of abnormalities, likely explains the persistently high burden of disease in the region.

These findings provide regional evidence that can guide evidence-based decisions by policy makers to accelerate the transition from cytology to HPV testing for primary cervical cancer screening instead of improving current cytology capacity.

The limited and highly variable sensitivity of cytology strongly supports the transition to the more robust and reproducible HPV-based cervical screening to ensure progress towards global cervical cancer elimination targets in Latin America. Therefore, although self-collection alone appears to be as sensitive as the Papanicolaou smear in several studies, its overall lower sensitivity suggests that it cannot be solely relied upon in screening programs.

The specificity, although comparable between both types of specimen collection, remains low. This highlights the importance of incorporating additional detection methods, such as immunohistochemical markers, to enhance the overall specificity in both self-collected and healthcare professional-collected samples.

Conclusions

In conclusion, these findings challenge the suitability of a self-collection-based approach for cervical cancer screening, emphasizing the potential limitations in identifying women at risk for cervical cancer. The results underscore the necessity to employ alternative methods complementary to traditional screening processes in populations where traditional cervical cancer screening may face challenges.

Ethical aspects

This study was not funded, nor is there any conflict of interest on the part of any of the authors.

Data availability

Indication that the data must be requested from the corresponding author. - All data must be requested from the corresponding author.

References

¹ Fowler J.R., Maani E.V., Dunton C.J., Gasalberti D.P., Jack B.W. Cervical cancer. In: StatPearls. Treasure Island (FL): StatPearls Publishing; November 12, 2023.
² de Melo A.C., da Silva J.L., Dos Santos A.L.S., Thuler L.C.S. Population-based trends in cervical cancer incidence and mortality in Brazil: focusing on black and indigenous Population disparities. J Racial Ethn Health Dispar. 2024;11(1):255-263.
³ Ramírez A.T., Valls J., Baena A., Rojas F.D., Ramírez K., Álvarez R., et al.; ESTAMPA Study Group. Performance of cervical cytology and HPV testing for primary cervical cancer screening in Latin America: an analysis within the ESTAMPA study. Lancet Reg Health Am. 2023;26:100593.
⁴ Levi J.E., Martins T.R., Longatto-Filho A., Cohen D.D., Cury L., Fuza L.M., et al. High-risk HPV testing in primary screening for cervical cancer in the public health system, São Paulo, Brazil. Cancer Prev Res. 2019;12(8):539-46.
⁵ Mekuria S.F., Timmermans S., Borgfeldt C., Jerkeman M., Johansson P., Linde D.S. HPV self-sampling versus healthcare provider collection on the effect of cervical cancer screening uptake and costs in LMIC: a systematic review and meta-analysis. Syst Rev. 2023;12(1):103.
⁶ Bruni L., Serrano B., Roura E., Alemany L., Cowan M., Herrero R., et al. Cervical cancer screening programmes and age-specific coverage estimates for 202 countries and territories worldwide: a review and synthetic analysis. Lancet Glob Health. 2022;10(8):e1115-e1127. Erratum: Lancet Glob Health. 2023;11(7):e1011.
⁷ Teixeira J.C., Vale D.B., Campos C.S., Bragança J.F., Discacciati M.G., Zeferino L.C. Organization of cervical cancer screening with DNA-HPV testing impact on early-stage cancer detection: a population-based demonstration study in a Brazilian city. Lancet Reg Health Am. 2022;5:100084.
⁸ Galvao T., Andrade Pansani T., Harrad D. Principais itens para relatar revisões sistemáticas e Meta-análises: a recomendação PRISMA. Epidemiol Serv Saude. 2015;24(2):335-42.
⁹ Catarino R., Vassilakos P., Bilancioni A., Bougel S., Boukrid M., Meyer-Hamme U., et al. Accuracy of self-collected vaginal dry swabs using the Xpert human papillomavirus assay. PLoS One. 2017;12(7):e0181905.
¹⁰ Bhatla N., Dar L., Patro A.R., Kumar P., Kriplani A., Gulati A., et al. Can human papillomavirus DNA testing of self-collected vaginal samples compare with physician-collected cervical samples and cytology for cervical cancer screening in developing countries? Cancer Epidemiol. 2009;33(6):446-50.
¹¹ Geraets D.T., van Baars R., Alonso I., Ordi J., Torné A., Melchers W.J., et al. Clinical evaluation of high-risk HPV detection on self-samples using the indicating FTA-elute solid-carrier cartridge. J Clin Virol. 2013;57(2):125-9.
¹² Wang S.M., Hu S.Y., Chen F., Chen W., Zhao F.H., Zhang Y.Q., et al. Clinical evaluation of human papillomavirus detection by careHPV™ test on physician-samples and self-samples using the indicating FTA Elute® card. Asian Pac J Cancer Prev. 2014;15(17):7085-9.
¹³ Sellors J.W., Lorincz A.T., Mahony J.B., Mielzynska I., Lytwyn A., Roth P., et al. Comparison of self-collected vaginal, vulvar and urine samples with physician-collected cervical samples for human papillomavirus testing to detect high-grade squamous intraepithelial lesions. CMAJ. 2000;163(5):513-8.
¹⁴ Seo S.S., Song Y.S., Kim J.W., Park N.H., Kang S.B., Lee H.P. Good correlation of HPV DNA test between self-collected vaginal and clinician-collected cervical samples by the oligonucleotide microarray. Gynecol Oncol. 2006;102(1):67-73.
¹⁵ Brink A.A., Meijer C.J., Wiegerinck M.A., Nieboer T.E., Kruitwagen R.F., van Kemenade F., et al. High concordance of results of testing for human papillomavirus in cervicovaginal samples collected by two methods, with comparison of a novel self-sampling device to a conventional endocervical brush. J Clin Microbiol. 2006;44(7):2518-23.
¹⁶ Leeman A., Del Pino M., Molijn A., Rodriguez A., Torné A., de Koning M., et al. HPV testing in first-void urine provides sensitivity for CIN2+ detection comparable with a smear taken by a clinician or a brush-based self-sample: cross-sectional data from a triage population. BJOG. 2017;124(9):1356-63.
¹⁷ Labani S., Asthana S. Human papillomavirus viral load on careHPV testing of self-collected vaginal samples vs. clinician-collected cervical samples. Eur J Obstet Gynecol Reprod Biol. 2014;181:233-9.
¹⁸ Asciutto K.C., Ernstson A., Forslund O., Borgfeldt C. Self-sampling with HPV mRNA analyses from vagina and urine compared with cervical samples. J Clin Virol. 2018;101:69-73.
¹⁹ de Melo Kuil L., Lorenzi A.T., Stein M.D., Resende J.C.P., Antoniazzi M., Longatto-Filho A., et al. The role of self-collection by vaginal lavage for the detection of HPV and high-grade intraepithelial neoplasia. Acta Cytol. 2017;61(6):425-33.
²⁰ Boggan J.C., Walmer D.K., Henderson G., Chakhtoura N., McCarthy S.H., Beauvais H.J., et al. Vaginal self-sampling for Human papillomavirus infection as a primary cervical cancer screening tool in a Haitian population. Sex Transm Dis. 2015;42(11):655-9.
²¹ Lorenzi N.P.C., Termini L., Ferreira-Filho E.S., Nunes R.A.L., Silva G.A.F., Lepique A.P., et al. A positive HPV test with positive p16/ki-67 double staining in self-sampled vaginal material is an accurate tool to detect women at risk for cervical cancer. Cancer Cytopathol. 2022;130(1):41-54.
²² Martinelli M., Giubbi C., Di Meo M.L., Perdoni F., Musumeci R., Leone B.E., et al. Accuracy of Human papillomavirus (HPV) testing on urine and vaginal self-samples compared to clinician-collected cervical sample in women referred to colposcopy. Viruses. 2023;15(9):1889.
²³ Ørnskov D., Jochumsen K., Steiner P.H., et al. Clinical performance and acceptability of self-collected vaginal and urine samples compared with clinician-taken cervical samples for HPV testing among women referred for colposcopy. A crosssectional study. BMJ Open. 2021;11:e041512.
²⁴ Cho H.W., Hong J.H., Min K.J., et al. Performance and diagnostic accuracy of Human papillomavirus testing on self-collected urine and vaginal samples in a referral population. Cancer Res Treat. 2021;53(3):829-36.
²⁵ Mremi A., Mchome B., Mlay J., et al. Performance of HPV testing, pap smear and VIA in women attending cervical cancer screening in Kilimanjaro region, Northern Tanzania: a cross-sectional study nested in a cohort. BMJ Open. 2022;12:e064321.
²⁶ Polman N.J., de Haan Y., Veldhuijzen N.J., Heideman D.A.M., de Vet H.C.W., Meijer C.J.L.M., et al. Experience with HPV self-sampling and clinician-based sampling in women attending routine cervical screening in the Netherlands. Prev Med. 2019;125:5-11.

Edited by

Edited by Editor:
José Maria Soares Junior

Publication Dates

Publication in this collection
18 May 2026
Date of issue
2026

History

Received
15 Aug 2025
Reviewed
5 Nov 2025
Accepted
27 Nov 2025
Published
20 Mar 2026

This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

[1] ¹ Fowler J.R., Maani E.V., Dunton C.J., Gasalberti D.P., Jack B.W. Cervical cancer. In: StatPearls. Treasure Island (FL): StatPearls Publishing; November 12, 2023.

[2] ² de Melo A.C., da Silva J.L., Dos Santos A.L.S., Thuler L.C.S. Population-based trends in cervical cancer incidence and mortality in Brazil: focusing on black and indigenous Population disparities. J Racial Ethn Health Dispar. 2024;11(1):255-263.

[3] ³ Ramírez A.T., Valls J., Baena A., Rojas F.D., Ramírez K., Álvarez R., et al.; ESTAMPA Study Group. Performance of cervical cytology and HPV testing for primary cervical cancer screening in Latin America: an analysis within the ESTAMPA study. Lancet Reg Health Am. 2023;26:100593.

[4] ⁴ Levi J.E., Martins T.R., Longatto-Filho A., Cohen D.D., Cury L., Fuza L.M., et al. High-risk HPV testing in primary screening for cervical cancer in the public health system, São Paulo, Brazil. Cancer Prev Res. 2019;12(8):539-46.

[5] ⁵ Mekuria S.F., Timmermans S., Borgfeldt C., Jerkeman M., Johansson P., Linde D.S. HPV self-sampling versus healthcare provider collection on the effect of cervical cancer screening uptake and costs in LMIC: a systematic review and meta-analysis. Syst Rev. 2023;12(1):103.

[6] ⁶ Bruni L., Serrano B., Roura E., Alemany L., Cowan M., Herrero R., et al. Cervical cancer screening programmes and age-specific coverage estimates for 202 countries and territories worldwide: a review and synthetic analysis. Lancet Glob Health. 2022;10(8):e1115-e1127. Erratum: Lancet Glob Health. 2023;11(7):e1011.

[7] ⁷ Teixeira J.C., Vale D.B., Campos C.S., Bragança J.F., Discacciati M.G., Zeferino L.C. Organization of cervical cancer screening with DNA-HPV testing impact on early-stage cancer detection: a population-based demonstration study in a Brazilian city. Lancet Reg Health Am. 2022;5:100084.

[8] ⁸ Galvao T., Andrade Pansani T., Harrad D. Principais itens para relatar revisões sistemáticas e Meta-análises: a recomendação PRISMA. Epidemiol Serv Saude. 2015;24(2):335-42.

[9] ⁹ Catarino R., Vassilakos P., Bilancioni A., Bougel S., Boukrid M., Meyer-Hamme U., et al. Accuracy of self-collected vaginal dry swabs using the Xpert human papillomavirus assay. PLoS One. 2017;12(7):e0181905.

[10] ¹⁰ Bhatla N., Dar L., Patro A.R., Kumar P., Kriplani A., Gulati A., et al. Can human papillomavirus DNA testing of self-collected vaginal samples compare with physician-collected cervical samples and cytology for cervical cancer screening in developing countries? Cancer Epidemiol. 2009;33(6):446-50.

[11] ¹¹ Geraets D.T., van Baars R., Alonso I., Ordi J., Torné A., Melchers W.J., et al. Clinical evaluation of high-risk HPV detection on self-samples using the indicating FTA-elute solid-carrier cartridge. J Clin Virol. 2013;57(2):125-9.

[12] ¹² Wang S.M., Hu S.Y., Chen F., Chen W., Zhao F.H., Zhang Y.Q., et al. Clinical evaluation of human papillomavirus detection by careHPV™ test on physician-samples and self-samples using the indicating FTA Elute® card. Asian Pac J Cancer Prev. 2014;15(17):7085-9.

[13] ¹³ Sellors J.W., Lorincz A.T., Mahony J.B., Mielzynska I., Lytwyn A., Roth P., et al. Comparison of self-collected vaginal, vulvar and urine samples with physician-collected cervical samples for human papillomavirus testing to detect high-grade squamous intraepithelial lesions. CMAJ. 2000;163(5):513-8.

[14] ¹⁴ Seo S.S., Song Y.S., Kim J.W., Park N.H., Kang S.B., Lee H.P. Good correlation of HPV DNA test between self-collected vaginal and clinician-collected cervical samples by the oligonucleotide microarray. Gynecol Oncol. 2006;102(1):67-73.

[15] ¹⁵ Brink A.A., Meijer C.J., Wiegerinck M.A., Nieboer T.E., Kruitwagen R.F., van Kemenade F., et al. High concordance of results of testing for human papillomavirus in cervicovaginal samples collected by two methods, with comparison of a novel self-sampling device to a conventional endocervical brush. J Clin Microbiol. 2006;44(7):2518-23.

[16] ¹⁶ Leeman A., Del Pino M., Molijn A., Rodriguez A., Torné A., de Koning M., et al. HPV testing in first-void urine provides sensitivity for CIN2+ detection comparable with a smear taken by a clinician or a brush-based self-sample: cross-sectional data from a triage population. BJOG. 2017;124(9):1356-63.

[17] ¹⁷ Labani S., Asthana S. Human papillomavirus viral load on careHPV testing of self-collected vaginal samples vs. clinician-collected cervical samples. Eur J Obstet Gynecol Reprod Biol. 2014;181:233-9.

[18] ¹⁸ Asciutto K.C., Ernstson A., Forslund O., Borgfeldt C. Self-sampling with HPV mRNA analyses from vagina and urine compared with cervical samples. J Clin Virol. 2018;101:69-73.

[19] ¹⁹ de Melo Kuil L., Lorenzi A.T., Stein M.D., Resende J.C.P., Antoniazzi M., Longatto-Filho A., et al. The role of self-collection by vaginal lavage for the detection of HPV and high-grade intraepithelial neoplasia. Acta Cytol. 2017;61(6):425-33.

[20] ²⁰ Boggan J.C., Walmer D.K., Henderson G., Chakhtoura N., McCarthy S.H., Beauvais H.J., et al. Vaginal self-sampling for Human papillomavirus infection as a primary cervical cancer screening tool in a Haitian population. Sex Transm Dis. 2015;42(11):655-9.

[21] ²¹ Lorenzi N.P.C., Termini L., Ferreira-Filho E.S., Nunes R.A.L., Silva G.A.F., Lepique A.P., et al. A positive HPV test with positive p16/ki-67 double staining in self-sampled vaginal material is an accurate tool to detect women at risk for cervical cancer. Cancer Cytopathol. 2022;130(1):41-54.

[22] ²² Martinelli M., Giubbi C., Di Meo M.L., Perdoni F., Musumeci R., Leone B.E., et al. Accuracy of Human papillomavirus (HPV) testing on urine and vaginal self-samples compared to clinician-collected cervical sample in women referred to colposcopy. Viruses. 2023;15(9):1889.

[23] ²³ Ørnskov D., Jochumsen K., Steiner P.H., et al. Clinical performance and acceptability of self-collected vaginal and urine samples compared with clinician-taken cervical samples for HPV testing among women referred for colposcopy. A crosssectional study. BMJ Open. 2021;11:e041512.

[24] ²⁴ Cho H.W., Hong J.H., Min K.J., et al. Performance and diagnostic accuracy of Human papillomavirus testing on self-collected urine and vaginal samples in a referral population. Cancer Res Treat. 2021;53(3):829-36.

[25] ²⁵ Mremi A., Mchome B., Mlay J., et al. Performance of HPV testing, pap smear and VIA in women attending cervical cancer screening in Kilimanjaro region, Northern Tanzania: a cross-sectional study nested in a cohort. BMJ Open. 2022;12:e064321.

[26] ²⁶ Polman N.J., de Haan Y., Veldhuijzen N.J., Heideman D.A.M., de Vet H.C.W., Meijer C.J.L.M., et al. Experience with HPV self-sampling and clinician-based sampling in women attending routine cervical screening in the Netherlands. Prev Med. 2019;125:5-11.

Study	Risk of bias				Applicability concerns
Study	Patient selection	Index test	Reference standard	Flow and timing	Patient selection	Index test	Reference standard
Catarino et al.⁹ (2017)	☺	☺	☺	?	☺	☺	☺
Bhatla et al.¹⁰ (2009)	☺	☺	☺	☺	☺	☺	☺
Geraets et al.¹¹ (2013)	☺	☺	☺	☺	☺	☺	☺
Shao-Ming Wang et al.¹² (2014)	☺	☺	?	☺	☺	☺	☺
Sellors et al.¹³ (2000)	☺	☺	☺	☺	☺	☺	☺
Sang-Soo Seo et al.⁷ (2006)	☺	☺	☺	☺	☺	☺	☺
Brink et al.¹⁵ (2006)	☺	☺	☺	☺	☺	☺	☺
Leeman et al.¹⁶ (2017)	☺	☺	☺	☺	☺	☺	☺
Satyanarayanai et al.¹⁷ (2014)	☺	☺	☺	☺	☺	☺	☺
Asciutto et al.¹⁹ (2008)	☺	☺	☺	☺	☺	☺	☺
de Melo Kuil et al.²⁰ (2017)	☺	☺	☺	☺	☺	☺	☺
Boggan et al.²¹ (2015)	☺	☺	☺	?	☺	☺	☺
Lorenzi et al. (2022)	☺	☺	☺	☺	☺	☺	☺
Martinelli et al. (2023)	☺	☺	☺	☺	☺	☺	☺
Ørnskov D, et al. (2020)	☺	☺	☺	☺	☺	☺	☺
Hyun-Woong Cho et al. (2020)	☺	☺	☺	☺	☺	☺	☺
Mremi A, et al. (2022)	?	☺	☺	☺	☺	☺	☺

Author (year) Study origin	Number of participants	Self-Collection device	Diagnostic DNA Method	Sampling storage medium for self-sample	GRADE Level of evidence
Catarino et al.⁹ (2017), Switzerland	150	Swab	Xpert HPV (PCR assay)	Dry medium	Low
Bhatla et al.¹⁰ (2009), India	512	Brush	Hybrid Capture II (HC2) and PCR reverse line blot assay	Digene HPV collection tube containing Specimen Transport Medium (liquid medium)	Low
Geraets et al.¹¹ (2013), Spain	182	Brush	SPF10 system and GP5+/6+ system (PCR assays)	FTA-elute solid-carrier cartridge (dry medium)	Low
Shao-Ming Wang et al.¹² (2014), China	396	Brush	careHPV™	DCM preservative solution (liquid medium) and FTA-elute (dry medium)	Low
Sellors et al.¹³ (2000) Canada	200	Swab	Hybrid Capture II (HC2)	Digene HPV collection tube containing Specimen Transport Medium (liquid medium)	Low
Jentschke et al.¹⁴ (2012), Germany	140	Vaginal lavage	Hybrid Capture II (HC2)	ThinPrep PreservCyt solution and Cervatec medium (liquid mediums)	Low
Sang-Soo Seo et al.⁷ (2006), South Korea	118	Swab	HPV oligonucleotide microarray system (HPVDNAChip™)	Sample Transport Medium (liquid medium)	Low
Brink et al.¹⁵ (2006), The Netherlands	96	Lavage	GP5+/6+ system (PCR assay)	Surepath preservation solution (liquid medium)	Low
Leeman et al.¹⁶ (2017), Spain	91	Brush	SPF10 system and GP5+/6+ system (PCR assays)	ThinPrep (liquid medium)	Low
Labani et al.¹⁷ (2014), India	4658	Brush	careHPV™	Digene Co-collection Media (liquid medium)	Low
Longatto-filho et al.¹⁸ (2008), Brazil	1081	Tampon	Hybrid Capture II (HC2)	‒	Low
Asciutto et al.¹⁹ (2008), Sweden	209	Swab	Aptima HPV mRNA test	‒	Low
de Melo Kuil et al.²⁰ (2017) Brazil	163	Lavage	Hybrid Capture II (HC2)	SurePath (liquid medium)	Low
Boggan et al.²¹ (2015), Haiti	1836	Brush	Hybrid Capture II (HC2)	Specimen Transport Medium (liquid medium)	Low

Author (year) Study origin	Number of participants	Self-Collection device	Genetic Testing	Sampling storage medium for self-sample	HrHPV types tested in Self-sample	Level of Evidence (GRADE)
Catarino et al.⁹ (2017), Switzerland⁹	150	Swab (No commercial name mentioned)	Xpert HPV (PCR assay)	Dry medium	PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68)	Low
Bhatla et al.¹⁰ (2009), India¹⁰	512	Brush (No commercial name mentioned)	Hybrid Capture II (HC2) and PCR reverse line blot assay	Digene HPV collection tube containing Specimen Transport Medium (liquid medium) [1 mL of transport medium]	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68) / PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68)	Low
Geraets et al.¹¹ (2013), Spain¹¹	182	Brush (Rovers Viba-Brush)	SPF10 system and GP5+/6+ system (PCR assays)	FTA-elute solid-carrier cartridge (dry medium)	PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
Shao-Ming Wang et al.¹² (2014), China¹²	396	Brush (Qiagen Inc. Conical Cervical Sampler)	careHPV™ (HC2)	DCM preservative solution (liquid medium) and FTA-elute (dry medium)	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
Sellors et al.¹³ (2000), Canada¹³	200	Swab (No commercial name mentioned)	Hybrid Capture II (HC2)	Digene HPV collection tube containing Specimen Transport Medium (liquid medium)	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68).	Low
Sang-Soo Seo et al.⁷ (2006), South Korea¹⁴	118	Swab (No commercial name mentioned)	HPV oligonucleotide microarray system (HPVDNAChip™)	Sample Transport Medium (liquid medium)	HPVDNA Chip (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 66)	Low
Brink et al.¹⁵ (2006), The Netherlands¹⁵	96	Lavage (Mermaid)	GP5+/6+ system (PCR assay)	Surepath preservation solution (liquid medium)	PCR (16, 18, 26 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 6, 68, 73 and 82)	Low
Leeman et al.¹⁶ (2017), Spain¹⁶	91	Brush (Colli-Pee™)	SPF10 system and GP5+/6+ system (PCR assays)	ThinPrep (liquid medium) [4 mL of transport medium]	SPF10 PCR (6, 11, 16, 18, 31, 33, 34, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 56, 58, 59, 66, 68, 73, 70, and 74) / GP5+/6+ PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68)	Low
Labani et al.¹⁷ (2014), India¹⁷	4658	Brush (Qiagen Inc. Conical Cervical Sampler)	careHPV™ (HC2)	Digene Co-collection Media (liquid medium)	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
Asciutto et al.¹⁹ (2008), Sweden¹⁸	209	Swab (Aptima Vaginal Swab Specimen Collection Kit)	Aptima HPV mRNA test	Aptima Vaginal Swab Specimen Collection Kit (liquid medium) [2.9 mL of transport medium]	Aptma HPV mRNA (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
de Melo Kuil et al.²⁰ (2017), Brazil¹⁹	163	Lavage (No commercial name mentioned)	Hybrid Capture II (HC2)	SurePath (liquid medium) [10 mL collected]	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68).	Low
Boggan et al.²¹ (2015), Haiti²⁰	1836	Brush (Qiagen Inc. Dacron Brush)	Hybrid Capture II (HC2)	Specimen Transport Medium (liquid medium)	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68)	Low
Lorenzi et al. (2022), Brazil²¹	232	Brush (Evalyn Brush)	COBAS 4800 HPV test (PCR)	BD SurePath Cell-preservation liquid [10 mL]	PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
Martinelli et al. (2023), Italy²²	245	Urine (Colli-pee®), Swab (FLOQSwab®)	Anyplex™Ⅱ HPV28 (PCR)	ThinPrep PreservCyt solution [5 mL]	PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68)	Low
Ørnskov D, et al. (2020) Denmark²³	305	Brush (Evalyn Brush) Urine (No commercial name mentioned)	COBAS 4800 HPV test (PCR)	Thin Prep Hologic (Brush) PreservCyt (Urine)	PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68)	Low
Hyun-Woong Cho et al. (2020) South Korea²⁴	314	Brush (Flocked Swab) Urine (No commercial name mentioned)	Anyplex™Ⅱ HPV28 (PCR) RealTime HR-S (PCR)	Thin Prep PreservCyt [5 mL]	Anyplex™ and RealTime PCR (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68).	Low
Mremi A, et al.(2022), Tanzania²⁵	512	Brush (Evalyn Brush)	careHPV™ (HC2)	[Sample Storage medium not mentioned]	HC2 (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68).	Low

A) Comparison of Self Collection Devices
		Self Collection		Healthcare Professional Collection
		Mean ± SD	95 % CI	Mean ± SD	95 % CI	p
Brush	SE	82.07 ± 16.13	(74.29 ± 89.84)	90.66 ± 10.91	(85.4 ± 95.91)	0.0625
Brush	SP	51.83 ± 20.03	(41.53 ± 62.13)	51.37 ± 22.11	(40 ± 62.73)	0.9494
Lavage	SE	70.95 ± 29.63	(−195.25 ± 337.15)	88.05 ± 9.26	(4.82 ± 171.28)	0.5175
Lavage	SP	63.85 ± 11.1	(−35.89 ± 163.59)	62.65 ± 12.23	(−47.26 ± 172.56)	0.9275
Swab	SE	90.53 ± 6.15	(84.08 ± 96.98)	74.88 ± 37.25	(35.79 ± 113.98)	0.3339
Swab	SP	40.13 ± 9.14	(25.58 ± 54.67)	45.1 ± 11.44	(26.89 ± 63.31)	0.5222
Urine	SE	83.72 ± 12.98	(67.61 ± 99.83)	94.34 ± 2.77	(90.9 ± 97.79)	0.1112
Urine	SP	41.27 ± 5.81	(34.06 ± 48.49)	38.56 ± 5.34	(31.93 ± 45.2)	0.4648

B) Comparison of Detection Methods
		Self Collection		Healthcare Professional Collection
		Mean ± SD	95 % CI	Mean ± SD	95 % CI	P
GP5/6+ PCR	SE	91.1 ± 3.09	(86.18 ‒ 96.02)	96.05 ± 1.56	(93.57 ‒ 98.53)	0.0289
GP5/6+ PCR	SP	53.35 ± 7.34	(41.66 ‒ 65.04)	45.73 ± 6.7	(35.06 ‒ 56.39)	0.176
SFP10 PCR	SE	96.33 ± 3.47	(87.71 ‒ 104.95)	100 ± 0	(100 ‒ 100)	0.1412
SFP10 PCR	SP	38.4 ± 4.06	(28.31 ‒ 48.49)	32.07 ± 2.34	(26.24 ‒ 37.89)	0.0795
Qiagen HC2	SE	80.74 ± 18.65	(57.58 ‒ 103.9)	92.28 ± 6.8	(83.84 ‒ 100.72)	0.2299
Qiagen HC2	SP	69.23 ± 14.66	(32.82 ‒ 105.64)	70.3 ± 17.72	(26.28 ‒ 114.32)	0.9398
Care HPV HC2	SE	59.03 ± 15.07	(35.05 ‒ 83)	76.08 ± 15.33	(51.68 ‒ 100.47)	0.1637
Care HPV HC2	SP	75.3 ± 5.23	(28.29 ‒ 122.31)	75.45 ± 3.46	(44.32 ‒ 106.58)	0.9761
COBAS PCR	SE	95.02 ± 1.56	(93.08 ‒ 96.96)	95.95 ± 1.33	(94.3 ‒ 97.6)	0.3422
COBAS PCR	SP	40.21 ± 5.71	(33.12 ‒ 47.29)	40.46 ± 6.02	(32.98 ‒ 47.94)	0.9471
Anyplex PCR	SE	78.79 ± 14.53	(60.75 ‒ 96.83)	90.79 ± 5.23	(84.3 ‒ 97.28)	0.1204
Anyplex PCR	SP	42.2 ± 11.41	(28.03 ‒ 56.37)	46.07 ± 20.48	(20.64 ‒ 71.51)	0.7215
RealTime PCR	SE	79.01 ± 8.1	(6.26 ‒ 151.75)	93.13 ± 0	(93.13 ‒ 93.13)	0.1324
RealTime PCR	SP	29.47 ± 3.78	(−4.52 ‒ 63.45)	32.69 ± 0	(32.69 ‒ 32.69)	0.3513
Aptima mRNA Array	SE	85.5 ±	(- ‒ -)	1 ±	(- ‒ -)	(- ‒ -)
Aptima mRNA Array	SP	48.1 ±	(- ‒ -)	49 ±	(- ‒ -)	(- ‒ -)
Other PCR	SE	85.73 ± 4.21	(75.26 ‒ 96.2)	82.9 ± 8.49	(61.8 ‒ 104)	0.632
Other PCR	SP	56.57 ± 33.32	(−26.22 ‒ 139.35)	61.27 ± 30.73	(−15.07 ‒ 137.61)	0.8662

Study	SMD	95 %CI	Weight	Heterogeneity ( %)
Detection Method ‒ Sensitivity
PCR	−0.62	−0.648 ‒ −0.592	49.99	98.8
Hybrid Capture	−0.802	−0.83 ‒ −0.774	50.01	98.8
D + L pooled SMD	−0.711	−0.889 ‒ −0.533	100	p < 0.001
Detection Method - Specificity
PCR	0.103	0.075 ‒ 0.131	49.98	99.1
Hybrid Capture	−0.107	−0.134 ‒ −0.081	50.02	99.1
D + L pooled SMD	−0.002	−0.208 ‒ 0.204	100	p = 0.983

Study	SMD	95 %CI	Weight	Heterogeneity ( %)
Type of Lesion ‒ Sensitivity
CIN2	−0.89	−0.917 ‒ −0.863	33.43	99.8
HSIL	−1.433	−1.518 ‒ −1.347	33.16
CIN3	−0.297	−0.329 ‒ −0.265	33.42
D + L pooled SMD	−0.872	−1.392 - −0.351	100	p = 0.001
Type of Lesion ‒ Specificity
CIN2	0.356	0.33 ‒ 0.383	33.43	99.8
HSIL	0.031	−0.045 ‒ 0.108	33.15
CIN3	−0.26	−0.292 ‒ −0.228	33.41
D + L pooled SMD	0.043	−0.407 ‒ 0.492	100	p = 0.853

Study	SMD	95 %CI	Weight	Heterogeneity ( %)
Type of Device ‒ Sensitivity
Brush	−0.551	−0.574 ‒ −0.527	37.25	73.9
Vaginal lavage	−0.779	−0.958 ‒ −0.6	7.55
Swab	−0.508	−0.551 ‒ −0.464	31.91
Urine	−0.49	−0.563 ‒ −0.417	23.29
D + L pooled SMD	−0.54	−0.595 - −0.486	100	p < 0.001
Type of Device ‒ Specificity
Brush	0.055	0.032 ‒ 0.078	25.69	99.2
Vaginal lavage	0.103	−0.07 ‒ 0.275	23.44
Swab	−0.388	−0.431 ‒ −0.345	25.57
Urine	0.215	0.143 ‒ 0.287	25.3
D + L pooled SMD	−0.007	−0.286 ‒ 0.273	100	p = 0.963

C) Comparison of Lesions Detected
		Self Collection		Healthcare Professional Collection
		Mean ± SD	95 % CI	Mean ± SD	95 % CI	p
CIN2+	SE	80.89 ± 16.54	(73.15 ‒ 88.63)	89.51 ± 11.14	(84.29 ‒ 94.72)	0.0607
CIN2+	SP	46.69 ± 15.58	(38.95 ‒ 54.44)	47.27 ± 17.44	(38.6 ‒ 55.94)	0.9173
CIN3+	SE	88.44 ± 15.69	(73.93 ‒ 102.96)	94 ± 5.54	(88.87 ‒ 99.13)	0.3943
CIN3+	SP	40.3 ± 9.14	(28.95 ‒ 51.65)	37.02 ± 5.89	(29.71 ‒ 44.33)	0.5192
HSIL	SE	85.22 ± 4.48	(79.66 ‒ 90.78)	74.28 ± 41.18	(23.15 ‒ 125.41)	0.5711
HSIL	SP	66.74 ± 20.05	(41.85 ‒ 91.63)	67.16 ± 21.34	(40.66 ‒ 93.66)	0.9752