Findings from three methods to identify falls in hospitals: Results from the Ambient Intelligent Geriatric Management system fall prevention trial.

OBJECTIVE: To (a) compare characteristics of patients who fall with those of patients who did not fall; and (b) characterise falls (time, injury severity and location) through three fall reporting methods (incident system reports, medical notes and clinician reports). METHODS: A substudy design within a stepped-wedge clinical trial was used: 3239 trial participants were recruited from two inpatient Geriatric Evaluation and Management Units and one general medicine ward in two Australian states. To compare the characteristics of patients who had fallen with those who had not, descriptive tests were used. To characterise falls through three reporting methods, bivariate logistic regressions were used. RESULTS: Patients who had fallen were more likely than patients who had not fallen to be cognitively impaired (51% vs. 29%, p < 0.01), admitted with falls (38% vs. 28%, p = 0.01) and have poor health outcomes such as prolonged length of stay (24 [16-34] vs. 12 [8-19] days [IQR], p < 0.01) and less likely to be discharged directly to the community (62% vs. 47%, p < 0.01). Most falls were captured from medical notes (93%), with clinician (71%) and incident reports (68%) missing 21%-25% of falls. The proportion of injurious falls identified through incident reports was higher than medical records or clinician reports (40% vs. 34% vs. 37%). CONCLUSIONS: This study reaffirms the need to improve reporting falls in incident systems and at clinical handover to the team leader. Research should continue to use more than one method of identifying falls, but include data from medical records. Many falls cause injury, resulting in poor health outcomes.

Low cost and batteryless sensor-enabled radio frequency identification tag based approaches to identify patient bed entry and exit posture transitions.

INTRODUCTION: Falls in hospitals and residential care facilities commonly occur near the bed. The aim of this study was to investigate the accuracy of a continuously wearable, batteryless, low power and low cost monitoring device (Wearable Wireless Identification and Sensing Platform) with a single kinematic sensor capable of real-time monitoring to automatically detect bed entry and exit events. MATERIALS AND METHODS: Three dimensional acceleration readings and the strength of the transmitted signal from the WISP was interpreted to identify bed exit events and sensitivity, specificity and Receiving Operator Curves (ROC) were determined. RESULTS: The sensor located over sternum method performed best with sensitivity and specificity values of 92.8% and 97.5% respectively for detecting bed entry and values of 90.4% and 93.80% respectively for bed exit. On the other hand, the sensor-on-mattress algorithm achieved sensitivity and specificity values of 84.2% and 97.4% respectively for bed entry and 79% and 97.4% for bed exit detection. CONCLUSION: The WISP located over the sternum method is the preferred method to detect bed entry and exit. However, further work in frail older people is required to confirm the performance of this method.