Background: Patient falls are one of the most serious problems facing health care institutions in Hong Kong. Fall can result in lengthened hospital stay, increased medical costs, and threats to the well being of patients. A review of fall risk assessment tools which are currently being used in different hospitals in Hong Kong showed that basically they were either developed or adopted by staff members after literature review. The value of these tools was often questionable since they were neither scientifically tested nor validated. A review of fall literature has shown that quite a large number of fall predicting factors have been identified by various authors and there was no locally validated fall assessment tool available in Hong Kong. Therefore, the objectives of our study were to develop a fall risk assessment tool and to test its validity in predicting falls in a regional acute hospital in Hong Kong.
Method & Result: A prospective case-control study was conducted in the 1,405 beds regional acute hospital. An instrument developed by the researcher that included information of patient demographics, fall incidences, and 26 key predictors identified in the literature review was used for data collection. Within 48 hours of each fall, the patient and the primary nurse were interviewed and the case notes reviewed. The sample consisted of 98 cases (fall) and 98 controls (nonfall), matched for the duration of stay and specialty unit. Logistic regression was used to develop a multivariate risk factor model with four risk factors. The significant risk factors were confusion, depression, dizziness, and mobility. The adjusted relative risks of these significant risk factors were converted to risk points to be used for assessing patients’ fall risk. Within the data set, a sensitivity of 93%, a specificity of 37%, and a predictive value of 59% were calculated.
Back to Accepted Posters
Back to 37th Biennial Convention - Clinical Sessions
Sigma Theta Tau International