Multimodal Assessment of Hand Hygiene Quality Using ATP Bioluminescence, Microbiological Culture, and UV-Fluorescence Digital Imaging: A Prospective Before–After Study Across Intensive Care, Hematology, and Gynecology Departments

Background: Healthcare-associated infections (HAIs) remain a critical patient safety challenge. Hand hygiene is considered the most effective preventive measure, yet traditional monitoring captures only compliance, not technique quality. This prospective before–after study evaluated whether real-time visual feedback via the Semmelweis UV-fluorescence system is associated with improved hand hygiene quality, measured by ATP bioluminescence and microbiological culture. Methods: Three clinical departments (the Intensive Care Unit, Hematology, and Gynecology) at a Romanian tertiary hospital were purposively selected. Seventy-one healthcare workers (HCWs) were enrolled. The 12-week study comprised Phase 1 (baseline, weeks 1–4), Phase 2 (active intervention with Semmelweis feedback, weeks 5–8), a one-week washout (week 9), and Phase 3 (sustainability assessment, weeks 10–12). Paired ATP-CFU samples were collected weekly. Within-group comparisons used Kruskal–Wallis H tests with post hoc Dunn’s tests and Bonferroni correction. Secondary outcomes included Semmelweis global and zone-specific coverage and the correlation between subject-level Semmelweis coverage and ATP bioluminescence (Spearman’s rho). Results: A total of 781 paired ATP-CFU samples and 497 Semmelweis evaluations were analyzed. Mean ATP declined from 195.9 RLU at baseline to 148.2 RLU in Phase 2 (−24.4%) and 154.8 RLU in Phase 3 (−21.0%; Kruskal–Wallis H = 102.73, p < 0.001). CFU/mL declined from 84.8 to 66.2 (−21.9%) and 70.7 (−16.6%; H = 22.48, p < 0.001). Post hoc comparisons confirmed significant Phase 1 versus Phase 2 and Phase 1 versus Phase 3 differences for both markers (all p < 0.01), while Phase 2 versus Phase 3 was non-significant, indicating stabilization at an improved level. Subject-level Semmelweis coverage correlated negatively with ATP (rho = −0.665, 95% CI −0.778 to −0.510, p < 0.001), supporting construct validity at the operator level. Semmelweis global coverage was 93.1% (Phase 2) and 90.6% (Phase 3); interdigital spaces showed the highest inadequacy rate (73.9% protocol-based, 92.5% targeted). Conclusions: Real-time visual feedback via UV-fluorescence imaging was associated with significant and sustained improvements in hand hygiene quality beyond baseline. ATP, CFU, and Semmelweis assessments captured complementary, non-redundant dimensions, supporting multimodal evaluation. Interdigital spaces and fingertips remained persistent failure points requiring targeted educational reinforcement.