78 | Chapter 5 therapist was included as random effect in the models. Despite the standardized training of therapists, the context of different sites of a widely implemented intervention may affect outcomes. To check for this, we calculated the Intraclass Correlation Coefficient (ICC) between therapists. In case the ICC was 0.2 or higher, therapist was included as random effect in the models. However, we found a low ICC (about 0.01), indicating that the data is largely independent of the therapist, and therefore correction for therapist was not needed (236). The mean number of falls and fall-related injuries, incidence rate ratios (IRR) and 95% confidence intervals (CI) were reported. Faller status (no falls or injuries vs at least one fall or injury) and number needed to treat are also described. The Timed Up and Go test, Four Stage Balance test, and the Physical Functioning domain and Emotional Wellbeing domain of the SF-36 were analysed with linear mixed-effect models (149, 238, 239). We included participant as a hierarchical level in the models to account for the longitudinal nature of the data. The mean scores of the secondary outcome measures, Relative Effects, and 95% confidence intervals were reported. To identify possible differences in intervention effects between non-frail and pre-frail phenotypes, we performed an a priori defined analysis, stratified for frailty status (non- and pre-frail). Last, a per-protocol analysis was done comparing individuals of the In Balance group who attended at least 75% of all In Balance sessions with those of the control group who did not participate in any fall prevention intervention during the follow-up (242). RESULTS Between June 2021 and January 2023, we screened 849 potential participants (Figure 5.1). Of these, 264 people were included in the study between August 2021 and January 2023 and randomly assigned to the intervention group (n = 131) or the control group (n = 133). Follow-up finished in January 2024. We had complete fall and injuries data of 117 persons the intervention group, and 106 persons in the control group. The In Balance program was provided by 15 therapists in 9 municipalities.
RkJQdWJsaXNoZXIy MjY0ODMw