Nationwide Comparison of ICU Procedure Frequencies in Japan Using a Public Open Database: A Cross-Sectional Study by ICU Admission Fee Type and Region

<b>Background/Objectives:</b> Publicly available open databases offer advantages in terms of accessibility and transparency. However, their application in intensive care research remains limited. Therefore, in this study, we examined whether simple nationwide comparisons of intensive care unit (ICU) practice patterns are feasible using an open database. <b>Methods:</b> A multicenter, cross-sectional study was conducted using data from the Bed Function Report. ICU wards reimbursed under ICU admission fee types were included and classified as high-acuity or standard ICUs. The ward-level procedure frequencies of procedures, including mechanical ventilation, were calculated. Comparisons were performed according to ICU admission fee type and geographic region. Quasi-Poisson regression models with offsets for annual ICU admissions were applied, accounting for overdispersion. <b>Results:</b> A total of 602 ICUs were included in the study. Non-metropolitan ICUs demonstrated higher procedural rates for mechanical ventilation compared with metropolitan ICUs (rate ratio [RR], 1.11; 95% confidence interval [CI], 1.02-1.21). Standard ICUs consistently had lower procedural rates for mechanical ventilation than high-acuity ICUs (RR, 0.74; 95% CI, 0.68-0.81). Group analyses indicated that regional differences in procedure frequencies were evident in standard ICUs, but not in high-acuity ICUs. <b>Conclusions:</b> This study demonstrated the feasibility of comparing ICU practice patterns across different regions and facility types in Japan using a nationwide open public database. This approach may serve as an initial step in a stepwise research framework that links open-database profiling to patient-level analysis using more detailed data sources.