Quality of Urolithiasis Information on Wikipedia and Baidu Baike: Cross-Language Comparative Study Using Nurse and Patient Ratings (Preprint)

BACKGROUND Online encyclopedic platforms are a common source of health information, but cross-language evaluations of urolithiasis content rated by both clinically informed and lay users remain limited. OBJECTIVE This study compared the quality of urolithiasis information on Wikipedia and Baidu Baike using ratings from urology nurses and patients. METHODS We conducted a cross-sectional comparison of 30 urolithiasis webpages (15 Wikipedia and 15 Baidu Baike). Pages were identified using clinically relevant English and Chinese search terms and were matched by exact topic or the nearest clinically equivalent concept. Two urology nurses and 2 patients independently rated each page using DISCERN and QUEST. For the main analysis, the 2 ratings within each evaluator group were averaged for each webpage. Primary outcomes were DISCERN and QUEST total scores. Platform differences were tested using the Mann-Whitney U test, interrater reliability was assessed with ICC(2,k), and mixed-effects models were used as sensitivity analyses. RESULTS Interrater agreement was good to excellent. Wikipedia scored higher than Baidu Baike on DISCERN total in nurse ratings (median 59.50 vs 50.50; P=.002) and patient ratings (61.50 vs 56.00; P=.020). Wikipedia also scored higher on QUEST total in nurses (18.50 vs 15.50; P=.038) and patients (19.00 vs 17.00; P=.016). After false discovery rate adjustment, only DISCERN reliability remained higher for Wikipedia in both evaluator groups, whereas no individual QUEST domain remained significant. Mixed-effects analyses were consistent with the primary findings. CONCLUSIONS In this sampled cross-language comparison, Wikipedia received higher overall DISCERN and QUEST scores than Baidu Baike in both nurse and patient ratings. The clearest domain-level difference was in DISCERN reliability. Other domain-specific differences, especially within QUEST, were less consistent after adjustment for multiple testing.