Under-reporting of falls in hospitals: a multisite study in South Korea

BACKGROUND: Inpatient falls are adverse events that often result in injury due to complex interactions between the hospital environment and patient risk factors and remain a significant problem in clinical settings. OBJECTIVES: This study aimed to identify (1) practice variations and key issues ranging from hospital fall management protocols to incident detection, and (2) potential approaches to address these challenges. DESIGN: Retrospective cohort study. SETTING: Four general hospitals in South Korea. METHODS: Qualitative and quantitative data were analysed using the Donabedian quality outcomes model. Data were collected retrospectively during 2015-2023 from four general hospitals on local practice protocols, patient admission and nursing data from electronic records, and incident self-reports. Content analysis of practice protocol and manual chart reviews for hospital falls incidents was conducted at each site. Quantitative analyses of nursing activities and analysis of patient falls prevention interventions were also conducted at each site. RESULTS: There were variations in fall definitions, risk-assessment tools and inclusion and exclusion criteria among the local fall management protocols. The original and modified versions of the heuristic tools performed poorly to moderately, with areas under the receiver operating characteristic curve of 0.54~0.74 and 0.59~0.80, respectively. Preventive intervention practices varied significantly among the sites, with risk-targeted and tailored interventions delivered to only 1.15%~49.5% of at-risk patients. Fall events were not recorded in self-reporting systems and nursing notes for 29.5%~90.6% and 4.4%~17.1% of patients, respectively. CONCLUSION: Challenges in fall prevention included weaknesses in the design and implementation of local fall protocols and low-quality incident self-reporting systems. Systematic and sustainable solutions are needed to help reduce hospital fall rates and injuries.