Analyzing a Multifactorial Fall Prevention Program Using ARIMA Models

Abstract

Background: Preventing inpatient falls is challenging for hospitals to improve and often leads to patient injury. Purpose: To describe multifactorial patient-tailored interventions and to evaluate whether they were associated with a sustained decline in total and injury falls. Methods: A multifactorial fall prevention program was instituted over the course of several years. An interrupted time series design was used to assess the effect of each intervention on total and injury fall rates. ARIMA models were built to assess the step and ramp change. Results: Total fall rates decreased from 4.3 to 3.6 falls per 1000 patient days (16.28% decrease), and injury fall rates decreased from 1.02 to 0.8 falls per 1000 patient days (21.57% decrease). All the interventions contributed to fall reduction, with specific interventions contributing more than others. Conclusions: Using multiple interventions that are sustained long enough to demonstrate success reduced the total fall rate and injury fall rate.