Investigating the effectiveness of mobilisation alarms to prevent hospital falls using disinvestment: A randomised clinical trial.

BACKGROUND: Mobilisation alarms are commonly used in hospitals to prevent falls in patients who are at high risk for falls, yet the evidence for their effectiveness is uncertain. OBJECTIVE: To investigate the effectiveness of mobilisation alarms to prevent falls in hospitals. DESIGN: This was a 3-group, concurrent, non-inferiority, stepped wedge, clinical trial with an embedded parallel, cluster randomised design that adopted a disinvestment approach. Disinvestment from the intervention was carried out from 1st April 2023 to 31st January 2024. SETTING(S): This study was conducted in one private health service and four public health services in Metropolitan Melbourne, Australia. PARTICIPANTS: Acute and sub-acute hospital wards with at least a 3 % rate of use of mobilisation alarms. Twenty-two wards were screened and found eligible for the trial, 18 wards were recruited and randomised. A random subsample of patients in the recruited wards completed questionnaires. METHODS: Two conditions were evaluated against the "current" condition of high alarm use (>3 %), a "reduced" rate of use of mobilisation alarms (<3 % but >0 %) and "eliminated" use of alarms (0 %). Rate of falls was the primary clinical outcome; data for a range of other measures were collected for secondary clinical and intervention outcomes. RESULTS: There were 11 acute wards and 7 sub-acute wards with 157,037 occupied bed days observed; 1319 individual patients completed questionnaires. Non-inferiority of the "reduced" condition compared to the "current" condition was demonstrated [95 %, one-tailed, non-inferiority confidence limit of 2.00 falls/1000 occupied bed days (OBDs) increase] but not for the "eliminated" condition (3.68 falls/1000 OBDs increase). Superiority of any intervention condition was not demonstrated over another ["current" vs "reduced": 0.22 falls/1000 OBDs (two-tailed, 95 % CI: -1.89 to 2.34), "current" vs "eliminated": 0.90 (-2.41 to 4.21), "reduced" vs "eliminated": 0.67 (-2.12 to 3.47)]. Patients' frequency of sleep disturbance due to alarms was reduced in the "eliminated" condition [ordered logistic coefficient 0.47 (0.08 to 0.87)]. CONCLUSIONS: Reduction in use of mobilisation alarms was not inferior to usual care, but complete elimination was uncertain. REGISTRATION: The trial was registered with the Australian New Zealand Clinical Trials Registry, trial ID: ACTRN12621000823875p (https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?ACTRN=12621000823875p). Registered on 28/06/2021, first enrolment on 26/10/2022.