HPV自采样与传统筛查方法的比较：系统综述与双变量荟萃分析

原文

BACKGROUND: Cervical cancer is a major public health challenge, especially in low-resource settings with suboptimal screening. HPV self-sampling (SS) is a promising strategy to overcome logistical barriers, but its diagnostic accuracy relative to clinician-collected (CC) sampling, Visual Inspection with Acetic Acid (VIA), and Papanicolaou (Pap) cytology requires rigorous validation. METHODS: Following PRISMA-DTA guidelines, we searched PubMed/Medline, Scopus, Embase, and the Cochrane Library through March 2025 for studies comparing high-risk HPV SS against CC, VIA, or Pap cytology, using biopsy-confirmed CIN2+ as the reference standard. A bivariate random-effects model pooled sensitivity and specificity. Hierarchical Summary Receiver Operating Characteristic (HSROC) curves and Deeks' Funnel Plots assessed discriminatory power and publication bias. RESULTS: Thirty-one studies (98,346 women) were included. Pooled sensitivity for SS was 0.745 (95% CI: 0.68-0.81), compared to 0.886 (95% CI: 0.83-0.93) for CC sampling. Specificity was comparable (SS: 0.850; CC: 0.870). SS was significantly superior to VIA (Sensitivity: 0.646 vs. 0.536). HPV SS was significantly superior to Pap cytology in sensitivity (0.74 vs. 0.67), though at a marginal cost to specificity (0.83 vs. 0.92), consistent with the diagnostic trade-off identified by the bivariate random-effects model. Brush-based samplers were the most effective device (Sensitivity: 0.78). Agreement between SS and CC was substantial (pooled Cohen's kappa: 0.67). No significant publication bias was detected (p = 0.921). CONCLUSION: Although clinician-collected sampling remains the most sensitive, HPV self-sampling is a robust alternative that significantly outperforms traditional VIA screening. The substantial concordance with CC sampling, particularly using brush-based devices, supports implementing self-sampling as a primary strategy to scale cervical cancer prevention in underserved populations.