Randomized Controlled Trial of Screening, Risk Modification, and Physical Therapy to Prevent Falls Among the Elderly Recently Discharged From the Emergency Department to the Community: The Steps to Avoid Falls in the Elderly Study

Objective To evaluate the effectiveness of a multifactorial, tailored program of physical therapy to reduce the occurrence of falls among a heterogeneous group of high-risk elderly Singaporeans recently discharged from the emergency department (ED). Design Randomized controlled trial. Setting Communities. Participants Adults (N=354) aged ≥65 years who were seen in the ED for a fall or fall-related injuries and discharged home. Interventions The intervention primarily consisted of a tailored program of physical therapy focused on progressive training in strength, balance, and gait for a period of 3 months. Participants in the intervention group also received screening and follow-up for vision, polypharmacy, and environmental hazards. Participants in the control group received usual care prescribed by a physician and educational materials on falls prevention. Main Outcome Measures The primary outcome measure was experiencing at least 1 fall during the 9-month study period (a 3-mo active intervention phase and a 6-mo maintenance phase). Secondary outcome measures were the occurrence of at least 1 injurious fall during the study period and a change in the Short Physical Performance Battery (SPPB) score. Participants were assessed both after 3 and 9 months. Results During the 9-month study period, 37.8% of the control group and 30.5% of the intervention group fell at least once, which was not statistically significantly different (odds ratio [OR]=.72; 95% confidence interval [CI], .46–1.12; P=.146). The intervention group had statistically significantly fewer individuals with injurious falls (OR=.56; 95% CI, .32–.98; P=.041) and less deterioration in physical performance, reflected by a mean difference of 0.6 in SPPB scores (P=.029). Multivariate analyses indicated a strong interaction effect between the intervention and the presence of 2 or more major comorbidities; after accounting for this effect, the intervention program reduced the number of people experiencing at least 1 fall (OR=.34; 95% CI, .17–.67; P=.002). Conclusions We observed that in this heterogeneous population, the proportion of participants experiencing at least 1 fall during the study period was not statistically significantly lower in the intervention group compared with the control group. Secondary analyses strongly suggest that individuals with 2 or more major comorbidities do not benefit from a tailored physical therapy program; however, individuals with less comorbidity may substantially benefit.