背景：世界衛生組織（WHO）提出的一個目標，讓70%符合資格的婦女在35歲與45 歲間，使用高效能的檢測方法接受子宮頸癌篩檢，以期在2030年前消除子宮頸癌。人類乳突病毒檢測(HPV testing) 的敏感度比抹片高，WHO推薦為第一線篩檢工具。況且提供自我採檢樣本更可以提高婦女接受子宮頸癌篩檢的比例。我國於2025年起，將HPV testing納入健保給付，目的為增加篩檢率與迎合WHO在2030年前清除子宮頸癌的目標。前段排出尿液檢體(first void urine, FVU)可用於HPV testing，但準確性正探討中。FVU含較高濃度的HPV DNA，推測可使用FVU提高篩檢敏感度。因此本研究目的為驗證HPV testing伴隨FVU篩檢子宮頸癌的準確性。材料與方法：於PubMed，Embase與Cochrane library資料庫中，以(cervical intraepithelial neoplasia or CIN2 or CIN3 or HSIL or high-grade squamous intraepithelial lesion or Uterine Cervical Dysplasia) and (Hu-man Papillomavirus DNA Tests or HPV testing) and (urine or Urine Specimen Col-lection)關鍵字搜尋。納入條件為：研究FVU偵測高度子宮頸癌前期及更嚴重病灶(high-grade squamous intraepithelial lesion or worse, HSIL+)的文獻；以histopathology 或 colposcopy 為黃金標準之文獻。統合分析的分法以random-effects model。結果：共納入20篇文章，樣本數為6150。統合分析結果顯示，HPV testing使用FVU偵測 HSIL+ 的 pooled sensitivity為 84.0% (95% CI: 78.6% to 88.2%)，pooled specificity為 45.2% (95% CI: 37.9% to 52.6%)。子族群分析指出，使用標準化容器採集FVU，HPV testing的pooled sensitivity為 88% (95% CI: 83.1%至91.6%)。結論：HPV testing伴隨FVU篩檢子宮頸癌具有高敏感度。HPV testing使用標準化容器採集的FVU樣本，可略為提升敏感度。

Background: Self-collection methods are regarded as a potential strategy to promote HPV testing for cervical cancer screening. HPV testing using urine speci-mens offers benefits, including its non-invasiveness. This characteristic has the po-tential to remove barriers associated with the procedure, such as discomfort. Never-theless, concerns regarding the accuracy of this method persist. The present me-ta-analysis aims to verify the accuracy of HPV testing with first void urine (FVU). Materials and methods: A literature search was conducted in the PubMed, Embase, and Cochrane Library databases to identify studies that evaluated the performance of HPV testing with FVU. The inclusion criteria were as follows: studies that evaluated the diagnostic accuracy of HPV testing for high-grade squamous intraepithelial le-sion or worse (HSIL+) with FVU. Additionally, studies that provided sufficient data for conducting a meta-analysis were assessed. The meta-analysis was conducted us-ing the bivariate random-effects model. Results: Twenty studies with 6150 samples were identified. The meta-analysis yielded a pooled sensitivity of 84.0% (95% CI: 78.6% to 88.2%) of HPV testing with self-collected FVU for HSIL+. For specimen collection, pooled sensitivity was 88% (95% CI: 83.1% to 91.6%) in the subgroup of papers using FVU via a standardized device to evaluate HPV testing. With regard to DNA-based testing, the subgroup analysis yielded a pooled sensitivity of 86.3% (95% CI: 84% to 88.4%). Conclusion: This meta-analysis indicates that HPV testing with FVU achieves high sensitivity for detecting HSIL+. Furthermore, HPV testing employing FVU collected with a standardized device has the potential to yield mar-ginally elevated sensitivity.