Clinical validation臨床驗證

Feasibility of unstaffed operation. A fully automated uroflowmetry unit ran in an existing public toilet at the Tainan City Public Health Bureau from March 2025, with venue, utilities and cleaning provided from existing non-clinical staff at no new public cost. Over 3,000 uses were recorded by August 2026 (as of 31 July 2026: over 3,000 uses and over 300 consented records).無人值守營運的可行性。全自動尿流速裝置自 2025 年 3 月起在台南市衛生局既有公廁運作，場地、水電與清潔由既有非臨床人力提供，未新增公共支出。截至 2026 年 8 月已逾 3,000 人次（以 2026/7/31 為基準日：逾 3,000 人次、逾 300 筆同意留存紀錄）。

Reach. 71 % of consenting users were aged 50 or under — close to the inverse of the age profile of clinic-based lower-urinary-tract care in national claims data, where only 21 % were under 50.觸及對象。同意留存的使用者中 71% 在 50 歲以下——幾乎是全國健保下泌尿道就醫人口（50 歲以下僅 21%）的反轉。

Signal. Among 221 interpretable records (voided volume ≥ 150 mL), maximum flow was below 15 mL/s in 41–47 % of men in every age band. This is a screening signal in an unselected population, not a prevalence estimate: the threshold was designed for symptomatic men in clinics, and single uncontrolled measurements inevitably yield high abnormal rates.訊號。221 筆可判讀紀錄（排尿量 ≥ 150 mL）中，各年齡層均有 41–47% 的男性最大尿流速低於 15 mL/s。這是未經篩選人群中的篩檢訊號，不是盛行率估計：該門檻原為診所中有症狀的男性設計，單次、非受控的量測本來就會出現偏高的異常率。

Limitations we state openly. Nine tenths of users did not complete the optional form; the design cannot distinguish no phone, unfamiliarity with scanning or unwillingness. The consented subsample favours smartphone users. Anonymity prevented follow-up. Full detail is in the GHF 2026 abstract.我們公開承認的限制。九成使用者未填寫選填表單；設計上無法區分沒有手機、不熟悉掃碼或不願意。同意留存的子樣本偏向智慧型手機使用者。匿名使追蹤不可行。完整內容見 GHF 2026 摘要。