PERFORMANCE OF DIFFERENT CERVICAL CANCER PRIMARY SCREENING METHODS: AN UMBRELLA REVIEW AND UPDATED META-ANALYSIS

Abstract

Previous systematic reviews and meta-analyses (SRMA) have assessed individual cervical cancer screening tests, but none has comprehensively assessed all major modalities using harmonized accuracy metrics. The objective of this research was to pool the diagnostic performance of visual inspection with acetic acid (VIA), visual inspection with Lugol’s iodine (VILI), cytology, human papillomavirus (HPV) DNA (clinician-collected and self-sampling), and HPV mRNA testing for detecting cervical intraepithelial neoplasia grade 2 or worse (CIN2+) in primary screening. We combined an umbrella review of 27 SRMAs (570 primary studies, >1.15 million women) with an updated meta-analysis of 34 individual studies using colposcopy-directed biopsy as the reference standard. Pooled sensitivity (Sn), specificity (Sp), positive and negative likelihood ratios (LR+ and LR-, respectively), and diagnostic odds ratios (DOR) were estimated using bivariate hierarchical summary receiver operating characteristic models. The results show that clinician-collected HPV DNA and mRNA testing had strong performance (HPV DNA: Sn 0.88, Sp 0.80, LR- 0.15, DOR 30; HPV mRNA: Sn 0.86, Sp 0.72, LR- 0.19, DOR 16), but the latter was based on only four studies. VILI appeared comparable (Sn 0.86, Sp 0.89, LR- 0.16, DOR 47) but VIA did not (Sn 0.71, Sp 0.87, LR- 0.33, DOR 17). Weaker accuracy was observed with self-sampled HPV DNA testing. Performance of cytology varied by threshold. In conclusion, clinician-collected HPV DNA testing showed the strongest overall accuracy for CIN2+ detection, supporting WHO recommendations. VILI is a viable alternative in resource-limited settings. HPV mRNA and self-sampling approaches need further evaluation.

Keywords: Cervical Cancer Screening, Diagnostic Accuracy, Umbrella Review, Meta-Analysis, HPV Testing, VIA, VILI, Cytology, CIN2+