Table 1 Table of evidence

Bansil et al., 2015

To evaluate and compare the performance of 3 cervical cancer screening options amongst HIV-infected women in Uganda

2337 HIV-positive and HIV-negative women; 25 and 60 years

Uganda

-Visual inspection with acetic acid (VIA)

-Vaginal careHPV

-Cervical careHPV

Observational study without a control group

Clinical performance of cervical careHPV, vaginal careHPV and VIA

Amongst HIV-positive women, cervical careHPV had sensitivity and specificity of 94.3% and 62.4%, respectively; vaginal careHPV had 80.0% sensitivity and 59.9% specificity; VIA had 77.1% sensitivity and 47.3% specificity.

Amongst HIV-negative women, VIA had 93.8% sensitivity and 60.5% specificity; cervical careHPV had 81.3% sensitivity and 80.9% specificity; vaginal careHPV had 75.0% sensitivity and 81.9% specificity.

CareHPV™ performs better for screening cervical cancer amongst HIV-positive women when compared to VIA.

VIA can be used as a triage method and reduce the number of treatment by half

Moderate

Cholli et al., 2018

Assess the feasibility and clinical outcomes of screening HIV-positive and HIV-negative Cameroonian women by pairing VIA/VILI-DC with careHPV (high-risk HPV nucleic acid test)

913 previously unscreened HIV-positive and HIV-negative women; ≥ 30 years

Cameroon

(VIA/VILI-DC) careHPV

Observational study without a control group

VIA/VILI-DC and careHPV co-testing strength amongst HIV-positive and HIV-negative women

For HIV-positive women, 8% (29/384) tested VIA/VILI-DC positive whilst 41% (157/384) tested positive on careHPV test (p < 0.0001).

HIV-positive women had more than twice VIA/VILI-DC positive results (n = 29/384, 7.6%) than HIV-negative women (n = 15/530, 2.8%) (p < 0.0001).

HIV-positive women were almost twice (1.9 times) more likely to test careHPV positive than HIV-negative for each VIA/VILI-DC result category.

Exclusively relying only on VIA/VILI-DC will likely result in about 50% overtreatment rate. Due to their weaknesses, pairing VIA/VILI-DC and HPV DNA testing is effective.

Moderate

Chung et al., 2013

To compare Papanicolau smear, visual inspection with acetic acid and human papillomavirus cervical cancer screening methods amongst HIV-positive women by immune status and antiretroviral therapy

500 HIV-positive women; 18 and 55 years

Kenya

Pap smear VIA HPV test colposcopy-directed biopsy

Observational study without a control group

Sensitivity, specificity of Pap smear, VIA and HPV testing

Individually, the most sensitive test was Pap (ASCUS+) (92.7%), which was significantly more sensitive than VIA (62.7%; p < 0.001), Pap (HSIL+) (71.8%; p < 0.001) and HPV (83.6%; p = 0.04) (Table 3). HPV was significantly more sensitive than VIA (p < 0.001) and Pap (HSIL+) (p = 0.04). Pap (HSIL+) (97.1%) was significantly more specific than VIA (65.9%; p < 0.001) and HPV (55.7%; p < 0.001), and VIA were more specific than HPV (p = 0.006). The cervical screening method with the highest AUC was Pap (HSIL+) (0.85), which was significantly greater than VIA (0.64; p < 0.001), HPV (0.70; p < 0.001), Pap (ASCUS+) (0.71; p < 0.001) and Pap (LSIL+) (0.76; p < 0.001). Combining cervical screening methods did not significantly improve test sensitivity over using Pap (ASCUS+) alone. However, combining VIA and Pap (HSIL+) to confirm positive test results had greater specificity than Pap (HSIL+) alone (99.1 vs. 97.1%; p < 0.001).

Pap smear is a robust test amongst HIV-positive women regardless of immune status or ART duration. Pap (ASCUS+) had the highest sensitivity, the combination of both Pap (HSIL+) and VIA positive had the highest specificity and Pap (HSIL+) had the highest AUC.

Moderate

Dartell et al., 2014

To examine the ability of VIA and HPV-testing to detect cytologically diagnosed high-grade lesions or cancer (HSIL+)

3603 HIV-positive and HIV-negative women; 24.4% were 29 years or younger, 35.1% were 30–39 years, 25.2% were 40–49 years and 15.3% were 50 years or older

Tanzania

Conventional cytology VIA HPV-DNA detection

Observational study without a control group

Sensitivity, specificity, positive predictive value and negative predictive value of VIA, HR HPV-testing compared to cytology

Amongst all women, VIA had a sensitivity of 28.5% (95% CI 20.9–36.0) and a specificity of 96.5% (95% CI 95.9–97.1). The sensitivity for VIA was higher in women from urban areas (39.0%) and amongst HIV-positive women (50.0%). HPV-testing had a high sensitivity (94.2%; 95% CI 90.2–98.1) and a somewhat lower specificity (82.8%; 95% CI 81.6–84.1). The specificity was lowest amongst HIV-positive women (58.2%) and amongst women 29 years or younger (74.7%). The VIA and HPV testing had a PPV ranging from 16.7 to 32.6% and from 7.2 to 22.9%, respectively. For both VIA and HPV-testing, the lowest PPV was seen amongst women below 29 years old. NPV was high for both VIA and HPV-testing (> 99.6%) and reached up to 100% for HPV testing amongst women who were below 29 years and amongst women who were HIV positive.

HPV testing would be a better primary screening tool for cervical cancer in sub-Saharan Africa, possibly with VIA as a secondary tool to increase specificity.

Moderate

Firnhaber et al., 2016

To compare VIA, cytology and HPV-DNA testing amongst HIV-positive women

688 HIV-positive women; age IQR (33–44)

South Africa

Pap smear VIA HPV testing

Observational study without a control group

Progression to CIN2+ was higher in women with positive VIA results (12.6%; 24/191) than those VIA-negative (4.4%; 19/432). HPV-positive women at baseline were more likely to progress to CIN2+ (12.3%; 36/293) than those HPV-negative (2.1%; 7/329). Cytology-positive women at baseline were more likely to progress to CIN2+ (9.6%; 37/384) than cytology-negative women (2.5%; 6/237). Approximately 10% (10.4%; 39/376) of women with CIN1 at baseline progressed to CIN2+. Women who were VIA or HPV positive at baseline were more likely to progress aIRR 1.85, (95% CI 1.46 to 2.36), aIRR 1.41 (95% CI 1.14 to 1.75), respectively.

Progression to CIN2+ in HIV-infected women is significant when measured by baseline positive VIA, HPV or Pap, and yearly screening by any method should be considered in this population if possible.

Moderate

Huchko et al., 2014

To determine the optimal strategy for cervical cancer screening in women with HIV infection by comparing two strategies: VIA and VIA followed by VILI in women with a positive VIA result

3462 HIV-positive women; 23–60 years

Kenya

VIA and VIA/VILI

Observational study with a control group

Positivity rate and PPV for VIA and VIA/VILI for CIN2+ amongst HIV-infected women

Screening positivity rate was 26.4% for VIA and 21.7% VIA/VILI (p = 0.003) with a follow-up colposcopy rate of 96.6% and 96.3%, respectively. The PPV of VIA for biopsy-confirmed CIN2+ in a single round of screening was 35.2% VIA, compared with 38.2% for VIA/VILI (p = 0.41).

The absence of much differences between VIA and VIA/VILI in detection rates or PPV for CIN2+ suggests that VIA can be used alone as a cervical cancer screening strategy in low-income settings.

High

Huchko et al., 2015

To compare the diagnostic accuracy of VIA to VILI for cervical cancer screening in HIV-infected women

654 HIV-positive women; 23–65 years

Kenya

VIA and VILI

Randomised clinical trial

Test performance of VIA or VILI

The test positivity rates were 26.2% for VIA and 30.6% for VILI (p = 0.22). The rate of detection of CIN2+ was 7.7% in the VIA arm and 11.5% in the VILI arm (p = 0.10). Sensitivity and specificity were 84.0% and 78.6%, respectively, for VIA and 84.2% and 76.4% for VILI. The positive and negative predictive values were 24.7% and 98.3% for VIA, and 31.7% and 97.4% for VILI. Amongst women with CD4+ count < 350, VILI had a significantly decreased specificity (66.2%) compared to VIA in the same group (83.9%, p = 0.02) and compared to VILI performed amongst women with CD4+ count ≥ 350 (79.7%, p = 0.02).

VIA and VILI had similar diagnostic accuracy and rates of CIN2+ detection amongst HIV-infected women.

High

Joshi et al., 2013

To evaluate an accurate, affordable, and feasible method to screen and treat HIV-infected women so that cervical cancer can be prevented amongst them

1128 HIV-positive women; 21–60 years

India

VIA, VILI, cytology, HPV testing, colposcopy

Observational study without a control group

Concurrent performance of cytology, HPV testing, VIA and VILI in detecting CIN2 and 3

The sensitivity, specificity and positive predictive values for VIA to detect CIN2 and 3 lesions were 83.6, 88.8 and 27.7%, respectively; the corresponding values for VILI were 89.1, 89.3 and 30.1%; for cytology at ASCUS threshold were 63.3, 94.5 and 35.2%; and for HPV testing were 94.6, 77.4 and 17.8%, respectively. Although VIA had a higher sensitivity than cytology, it did not reach statistical significance. HPV testing was 100% sensitive in detecting CIN3 lesions; however, it had significantly lower specificity than VIA, VILI and cytology (p < 0.001).

HPV testing, VILI and VIA have a higher sensitivity in detecting high-grade CIN than that of conventional cytology. Sequential testing with VIA and VILI is the most feasible screening approach for cervical cancer screening in HIV-infected women in low-resource countries. When HPV testing becomes feasible and affordable, HPV testing followed by VIA/VILI may be considered.

Moderate

Kuhn et al., 2010

To evaluate the efficacy amongst HIV-infected women of a simpler, screen-and-treat strategy in which all women with a positive screening test are treated with cryotherapy

6555 women, whom 956 were HIV-positive; 35–65 years

South Africa

HPV DNA based screen-and-treat and VIA-based screen-and-treat

Randomised clinical trial

Safety and efficacy of screen-and-treat amongst HIV-positive women

HPV DNA testing was highly effective in reducing the risk of CIN2+ by 36 months amongst both HIV-positive [relative risk (RR) = 0.20, 95% confidence interval (CI) 0.06–0.69] and HIV-negative (RR = 0.31, 95% CI 0.20–0.50) women. The benefit of VIA-and-treat was less marked and only reached statistical significance in HIV-positive women (RR = 0.51, 95% CI 0.29–0.89) and not in HIV-negative women (RR = 0.76, 95% CI 0.52–1.1). The sensitivity of HPV DNA testing at enrolment to detect CIN2+ through 36 months was 94.4% in HIV-positive women, whereas the sensitivity of the VIA test was 63.9% in HIV-positive women. In the HPV and-treat group, there was a slightly lower rate of CIN2+ after cryotherapy amongst HIV-positive (2.8%) vs. HIV-negative (7.1%) women, but this difference was of borderline significance (p = 0.05). In the VIA-and-treat group, CIN2+ failure rates after cryotherapy were similar in HIV-positive (4.8%) and HIV-negative (2.8%) women

HPV-based screen-and-treat is safe and effective in HIV-positive women. A single round of screening with an HPV test followed by cryotherapy of all screen-positive women reduced high-grade cervical cancer precursors (CIN2+) by 80%, and this was sustained through 36 months. VIA-based screen-and-treat was significantly less effective, although better than no intervention in HIV-positive women.

High

Lim et al., 2011

To compare Pap smear readings to VIA findings amongst HIV-infected women in Phnom Penh, Cambodia

293 HIV-infected women

Cambodia

Pap smear and VIA

Observational study without a control group

Degree of correlation between Pap smear and VIA findings

55 (19%) women screened positive on VIA; 25 (8.5%) women screened positive by Pap. Visual inspection with acetic acid detected 18 of the 25 patients with abnormal cytology and was normal in 7 women with abnormal cytology. 37 (67%) women with positive VIA were negative by cytology.

Our study shows a reasonable correlation between VIA and Pap smear, with VIA detecting more abnormalities than cytology. In the absence of Pap smear availability, VIA may be a reasonable cervical cancer screening method for HIV-infected women in Cambodia.

Low

Mabeya et al., 2012

To determine the accuracy of VIA vs. Pap smear amongst HIV-infected women

150 HIV-infected women; 20–45 years

Kenya

Pap smear and VIA

Observational study without a control group

Accuracy of VIA vs. conventional Pap smear as a screening tool for CIN/cancer amongst HIV-infected women with biopsy as the reference criterion standard

Using AUC as an overall measure of screening accuracies and using CIN1 or higher as the gold standard threshold, the performance of Pap smear is slightly better than VIA, but the difference is not significant (Pap smear: AUC = 0.596, VIA: AUC = 0.571, p = 0.64). When using CIN2 or higher as the gold standard threshold, the performance of Pap smear and VIA are more comparable (Pap smear: AUC = 0.606, VIA: AUC = 0.603, p = 0.93). Using CIN2 or higher disease on biopsy as an end point, VIA has a sensitivity of 69.6% (95% CI = 55.1–81.0%), specificity of 51.0% (95% CI = 41.5–60.4%), PPV of 38.6% (95% CI = 28.8–49.3%), and NPV of 79.1% (95% CI = 67.8–87.2%). For conventional Pap smear, sensitivity was 52.5% (95% CI = 42.1–71.5%), specificity was 66.3% (95% CI = 52.0–71.2%), PPV was 39.7% (95% CI = 27.6–51.8%) and NPV was 76.8% (95% CI = 67.0–85.6%).

Visual inspection with acetic acid is comparable to Pap smear and acceptable for screening HIV-infected women in resource-limited settings such as Western Kenya.

Moderate

Michelow et al., 2016

To evaluate the performance of the Cellslide® automated liquid-based cytology (LBC) system as a possible alternative to conventional cytology amongst HIV-positive women

348 HIV-positive women;18–65 years

South Africa

Cellslide® automated LBC

Observational study without a control group

Number of positive and negative samples tested using Cellslide®

For HSIL, Cellslide® showed sensitivity of 76.0% (95% CI 64.8–85.1) and specificity of 91.0% (95% CI 87.0–94.2), with a false-omission rate < 7%, compared with conventional cytology. When compared with conventional cytology, Cellslide® showed sensitivity of 89.6% (95% CI 82.9–94.4) and specificity of 92.2% (95% CI 87.8–95.4) for NILM, sensitivity of 70.2% (95% CI 61.3–78.0) and specificity of 87.7% (95% CI 82.6–91.7) for LSIL and sensitivity of 100% (95% CI 2.5–100) and specificity of 98.8% (95% CI 97.1–99.7) for ASCH.

The performance of the Cellslide((R)) LBC system was similar to that of conventional cytology in this population of high-risk HIV-positive women, indicating that it may be introduced successfully as part of a cervical cancer screening programme.

Low

Mutyaba et al., 2010

To evaluate the ‘see-see and treat’ strategy and the role of HIV on cervical cancer prevention in Uganda

5105 HIV-negative and HIV-positive women; 20–60 years

Uganda

VIA/VILI and cryotherapy

Observational study without a control group

Detection rates by age-group and cervical lesion treatment

Detection rates per 1000 women screened were higher amongst the older women (41–60 years) compared to women aged 20–40 years. They were accordingly 55% and 20% for inflammation, 10% and 2% for LGSIL, 5% and 2% for HGSIL and 6% and 1% for invasive cervical cancer. Of the 608 women, 103 (16%) were HIV positive. HIV positivity was associated with a higher likelihood of inflammation (RR = 1.7; 95% CI 1.2–2.4). The 32 women with SIL (19 LGSIL and 13 HGSIL) underwent treatment by cryotherapy (31 women) or LEEP (1 woman). 1 woman had persistent LSIL and 1 had inflammation; both were HIV positive. Other 27 women had normal findings.

VIA/VILI used as a sole method for cervical cancer screening would entail significant false-positive results. HIV seropositivity was associated with a higher prevalence of inflammatory cervical lesions. Cryotherapy treatment outcome was not conclusive due to the limited follow-up time

Low

Ngou et al., 2015

To compare the Hybrid Capture 2 HPV DNA assay (HC2) and the INNO-LiPA HPV Genotyping Extra assay (INNO-LiPA) for cervical cancer screening in HIV-1-infected African women

1224 HIV-positive women in Burkina Faso (N = 604) and South Africa (N = 620); 25–50 years

Burkina Faso and South Africa

HC2 and INNO-LiPA

Observational study without a control group

Agreement between HC2 and INNO-LiPA for detection of HR-HPV infection and compare their performances in diagnosing cervical lesions detected by cytology and histology

When considering the 13 h-HPV types detected by HC2, 634 (51.8%) and 849 (69.4%) samples were positive by HC2 and INNO-LiPA, respectively. Agreement between assays was 73.9% [adjusted kappa coefficient value, 0.44 (95% confidence interval 0.43 to 0.53)]. Agreement improved with analysis restricted to women with high-grade cervical lesions [adjusted kappa coefficient value, 0.83 (95% confidence interval 0.74 to 0.91)]. The prevalence of hr-HPV, as determined by HC2 and INNO-LiPA, was 34.5% and 54.5%, respectively, in samples with normal cytology, 48.0% and 68.0%, respectively, in samples with atypical squamous cells of undetermined significance, 51.8% and 75.2%, respectively, in samples with low-grade SIL and 86.3% and 89.8%, respectively, in samples with high-grade SIL/atypical squamous cells that cannot exclude HSIL. Sensitivity, specificity and positive and negative predictive values for the diagnosis of histological high-grade lesions (CIN2+) were 88.8%, 55.2%, 24.7% and 96.7%, and 92.5%, 35.1%, 19.1% and 96.6% for HC2 and INNO-LiPA, respectively

HC2 has lower analytical sensitivity but higher specificity than INNO-LiPA for diagnosing high-grade lesions; the 2 tests presented a comparable clinical sensitivity. HC2 might be suitable for cervical cancer screening in HIV-1-infected African women, but its use in resource-limited settings merits to be further evaluated in comparison with other prevention strategies.

Moderate

Ngou et al., 2013

To compare careHPV and hybrid capture 2 assays for detection of high-risk human papillomavirus DNA in cervical samples from HIV-1-infected African women

149 HIV-1-infected African women (75 in Johannesburg, South Africa and 74 in Ouagadougou, Burkina Faso); 25–50 years

Burkina Faso and South Africa

careHPV and HC2

Observational study without a control group

Agreement in detecting HR-HPV between careHPV and HC2

The HR-HPV DNA detection rates were 37.6% and 34.9% for careHPV and HC2, respectively. Agreement between the two tests was 94.6% (95% confidence interval [CI], 89.7 to 97.7%) with a kappa value of 0.88 (95% CI, 0.81 to 0.96), indicating an excellent agreement.

careHPV may be considered as suitable as HC2 for cervical cancer screening amongst HIV-infected African women.

Moderate

Obiri-Yeboah et al., 2017

To compare the performance of careHPV with HPV genotyping for the detection of cytological cervical squamous intraepithelial lesions (SIL)

175 women (94 HIV-1-seropositive and 81 HIV-seronegative women); ≥ 18 years

Ghana

HPV Genotyping vs. careHPV

Observational study with a control group

Agreement in detecting HR-HPV between careHPV and HPV genotyping

The inter-assay concordance was 94.3% (95% CI 89.7–97.2%, kappa = 0.88), similar by HIV serostatus. The careHPV assay was equally sensitive amongst HIV-1-seropositive and HIV-1-seronegative women (97.3% vs. 95.7%, p = 0.50) and slightly more specific amongst HIV-seronegative women (85.0% vs. 93.1%, p = 0.10). careHPV had good sensitivity (87.5%) but low specificity (52.1%) for the detection of low SIL or greater lesions, but its performance was superior to genotyping (87.5 and 38.8%, respectively). Reproducibility of careHPV, tested on 97 samples by the same individual, was 82.5% (95% CI 73.4–89.4%).

The performance characteristics of careHPV compared to genotyping suggest that this simpler and cheaper HPV detection assay could offer a suitable alternative for HPV screening in Ghana.

High

Obiri-Yeboah et al., 2017

To determine the acceptability, feasibility and performance of alternative self-collected vaginal samples for HPV detection using careHPV amongst Ghanaian women

194 women (97 HIV-positive); ≥ 18 years

Ghana

Self-collected vaginal samples with care HPV

Observational study with a control group

Performance of self-collected cervico-vaginal samples compared to clinician-collected samples

Overall HPV detection concordance was 94.2% (95% CI 89.9–97.1), kappa value of 0.88 (p < 0.0001), showing excellent agreement. This agreement was similar between HIV-positive (93.8%) and HIV-negative (94.7%) women. Sensitivity and specificity of SC compared to CC were 92.6% (95% CI 85.3–97.0) and 95.9% (95% CI 89.8–98.8), respectively. The highest sensitivity was amongst HIV-positive women (95.7%, 95% CI 88.0–99.1) and highest specificity amongst HIV-negative women (98.6%, 95% CI 92.4–100). Overall, 76.3% of women found SC very easy/easy to obtain, 57.7% preferred SC to CC and 61.9% felt SC would increase their likelihood to access cervical cancer screening.

The feasibility, acceptability and performance of SC using careHPV support the use of this alternative form of HPV screening amongst Ghanaian women. This could be a potential new affordable strategy to improve uptake of the national cervical cancer screening programme.

Moderate

Sahasrabuddhe et al., 2012

To rigorously evaluate the clinical accuracy of VIA and cytology amongst HIV-infected women in Pune, India

303 non-pregnant HIV-infected women; 25–40 years

India

VIA Pap smear

Observational study without a control group

Sensitivity, specificity, PPV and NPV for VIA and cytology

At CIN2+ disease threshold, the sensitivity, specificity and positive and negative predictive value estimates of VIA were 80, 82.6, 47.6 and 95.4%, respectively, compared to 60.5, 59.6, 22.4 and 88.7% for the atypical squamous cells of undetermined significance or severe (ASCUS+) cutoff on cytology; 60.5, 64.6, 24.8 and 89.4% for the low-grade squamous intraepithelial cells or severe (LSIL+) cutoff on cytology; and 20.9, 96.0, 50.0 and 86.3% for high-grade squamous intraepithelial lesion or severe (HSIL+) cutoff on cytology. A similar pattern of results was found for women with the presence of carcinogenic HPV-positive CIN2+ disease, as well as for women with CD4+ cell counts < 200 and < 350 muL(−1).

Overall, VIA performed better than cytology in this study with biologically rigorous endpoints and without verification bias, suggesting that VIA is a practical and useful alternative or adjunctive screening test for HIV-infected women.

Moderate

Wu et al., 2016

To measure the sensitivity, specificity and predictive values of p16INK4a ELISA for CIN2+

1054 HIV-infected women; ≥ 23 years

Kenya

p16(INK4a) ELISA

Observational study without a control group

Sensitivity, specificity and predictive values of p16INK4a ELISA

The p16INK4a cutoff value with the highest combined sensitivity (89.0%) and specificity (22.9%) for biopsy-proven CIN2+ was 9 U/mL. The positive predictive value was 13.6% and negative predictive value was 93.8%. Overall, the p16INK4a positivity with the selected 9 U/mL cutoff level was 828 (78.6%) women; in comparison, biopsy-proven CIN2+ was found in only 127 (12%) women.

p16(INK4a) ELISA did not perform well as a screening test for CIN2+ detection amongst HIV-infected women due to low specificity. Our study contributes to the ongoing search for a more specific alternative to HPV testing for CIN2+ detection.

Low

Akinwuntwan et al., 2008

To assess the correlation between cytology and VIA in HIV-positive women

205 consenting HIV-seropositive women; 17–60 years

Nigeria

Pap smear VIA

Observational study without a control group

Sensitivity, specificity, PPV, NPV and diagnostic accuracy of Pap smear and VIA

The sensitivity of VIA was 76.0% (95% CI 52.0–91.0), specificity 83.0% (95% CI 77.0–88.0) and positive predictive value 34.0% (95% CI 21.0–49.0). The sensitivity of Pap smear was 57.0% (95% CI 34.0–77.0), specificity of 95.0% (95% CI 90.0–97.0) and positive predictive value of 55.0% (95% CI 33.0–75.0). Diagnostic accuracy of VIA is 82.0% (95% CI 76.0–87.0) and for Pap smear is 91.0% (95% CI 86.0–98.0)

In HIV-seropositive women, the sensitivity of VIA is 76.0%, making it a useful screening test for pre-invasive lesion of the cervix in low resource settings but not a diagnostic tool.

Moderate

Firnhaber et al., 2013

To compare the sensitivity and specificity of conventional Pap smear screening to that of HPV DNA and VIA testing for detection of histologically confirmed high-grade CIN2+ in HIV-infected women

1202 HIV-infected women; 18–65 years

South Africa

Pap smear, VIA and HPV DNA test using HC2

Observational study without a control group

Sensitivity and specificity of Pap smear, HPV DNA test and VIA

VIA and HPV were positive in 45% and 61% of women, respectively. Estimated sensitivity/specificity for HPV, Pap smear and VIA for CIN2+ was 92%/51.4%, 75.8%/83.4% and 65.4/68.5% (nurse reading), respectively. Sensitivities were similar, and specificities appeared significantly lower for the HPV test, cytology and VIA amongst women with CD4 counts ≤ 200 cells/mm³ as compared to CD4 counts > 350 cells/mm³.

Although HPV was the most sensitive screening method for detecting CIN2+, it was less specific than conventional cytology and VIA with digital imaging review. Screening programmes may need to be individualised in the context of the resources and capacity in each area.

Moderate

Chibwesha et al., 20165

To determine the clinical performance of VIA, digital cervicography (DC), Xpert HPV and OncoE6 for cervical cancer screening in an HIV-infected population

200 HIV-infected women; ≥ 18 years

Zambia

VIA, DC, Xpert HPV and OncoE6

Observational study without a control group

Sensitivity and specificity of VIA, DC, Xpert HPV and OncoE6

Of the 200 women, 15% were screen positive by VIA, 20% by DC, 47% by Xpert HPV and 6% by OncoE6. Using a CIN2+ threshold, the sensitivity and specificity of VIA were 48% (95% confidence interval [CI] 30–67%) and 92% (95% CI 86–95%), respectively. Similarly, the sensitivity and specificity of DC were 59% (95% CI 41–76%) and 88% (95% CI 82–93%), respectively. The sensitivity and specificity of Xpert HPV were 88% (95% CI 71–97%) and 60% (95% CI 52–68%), respectively. Finally, the sensitivity and specificity of OncoE6 were 31% (95% CI 16–50%) and 99% (95% CI 97–100%), respectively.

VIA and DC displayed moderate sensitivity and high specificity. Xpert HPV performed equivalently to currently approved HPV DNA tests, with high sensitivity and moderate specificity. OncoE6 displayed excellent specificity but low sensitivity. These results confirm an important role for VIA, DC and Xpert HPV in screen-and-treat cervical cancer prevention in low- and middle-income countries, such as Zambia

Low

Adamson et al., 2015

To (1) compare the test positivity between the two collection methods, (2) assess the accuracy and agreement of self-collected tampons compared to clinician-collected specimens for hrHPV mRNA testing and (3) assess the acceptability of the self-collected tampon method

325 HIV-infected women; ≥ 25 years

South Africa

HrHPV messenger RNA (mRNA) test

Observational study without a control group

Sensitivity and specificity of hrHPV mRNA test

Over 90% of women reported no difficulties self-collecting specimens, and 82% were willing to perform the tampon collection at home. Based on clinician collection specimens, the prevalence of hrHPV mRNA in our study population was 36.7% (95% CI 31.4–42.0%). There was no difference in test positivity between clinician collection, 36.7%, and tampon collection, 43.5% (p = 0.08). Using clinician collection as the reference test, the sensitivity and specificity for hrHPV mRNA of tampon collection were 77.4% (95% CI 69.8–85.0%) and 77.8% (95% CI 71.9–83.6%), respectively.

Tampon-based self-collection is acceptable to women and has similar hrHPV mRNA positivity rates as clinician collection but has reduced sensitivity and specificity compared to clinician collection.

Moderate

Segondy et al., 2016

To evaluate the performance of careHPV for detecting CIN2+ amongst women living with HIV-1 in Burkina Faso and South Africa

1052 HIV-1-seropositive women; 25–50 years

South Africa and Burkina Faso

careHPV assay INNO-LiPA

Observational study without a control group

Sensitivity, specificity, positive and negative predictive values of careHPV assay

Overall, 45.1% of women had a positive careHPV test (46.5% in BF, 43.8% in SA). The careHPV positivity rate increased with the grade of cytological lesions. Sensitivity and specificity of careHPV for the diagnosis of CIN2+ (n = 60, both countries combined) were 93.3% (95% confidence interval (CI) 83.8–98.2) and 57.9% (95% CI 54.5–61.2), respectively. Specificity increased with CD4 count. careHPV had a similar clinical sensitivity but higher specificity than the INNO-LiPA assay for detection of CIN2+.

Results suggest that careHPV testing is a reliable tool for cervical cancer screening in HIV-1-infected women in sub-Saharan Africa.

Moderate

Bateman et al., 2014

To assess the clinical performance of DC, as well as cytology in HIV-infected women

303 women; 20–45 years

Zambia

DC Pap smear

Observational study without a control group

Clinical performance of each screening test to detect cervical lesions on histopathology

The sensitivity of DC for identifying CIN2+ was 84% (95% CI 72–91%), and the specificity was 58% (95% CI 52–64%) (Table 2). The sensitivity estimates of cytology for identifying CIN2+ were as follows: HSIL+, 61% (95% CI 48–72%); LSIL+, 90% (95% CI 80–95%); and ASC-US+, 100% (95% CI 94–100%). The specificity estimates of cytology for identifying CIN2+ were as follows: HSIL+, 58% (95% CI 52–64%); LSIL+, 35% (95% CI 29–41%); and ASC-US+, 13% (95% CI 10–18%). The PPVs were low (23–33%) for both tests, whilst the NPVs were correspondingly high (86–100%). A similar pattern of results was observed at the CIN3+ diagnostic threshold on histopathology (Table 2).

Digital cervicography appears to be as good as cytology in HIV-infected women.

Moderate