Setting
Cedars-Sinai Medical Center is a tertiary-care, 886-bed hospital and associated integrated healthcare system in Los Angeles with 50,000 admissions per year.
Population
Adults age ≥ 65 admitted as inpatients or placed under observation between January 1, 2016 to December 31, 2021.
Data
We obtained all diagnoses and medications during the admission from the electronic health record data warehouse.
Primary independent variable
Our primary independent variable was the number of psychotropic sedative medications (benzodiazepines, z-drugs, antipsychotics, barbiturates) on the patient’s medication list on admission. To capture only current medications, we excluded medications prescribed > 31 days before admission. We excluded benzodiazepines prescribed for alcohol withdrawal treatment.
Outcomes
We used ICD-10 codes from Moreland et al(2), who used guidance from the National Center for Health Statistics and the National Center for Injury Prevention and Control, to create separate dummy variables for injury and falls diagnoses.
Analyses
We used bivariate statistics (chi-square, ANOVAs) to examine patients with different numbers of psychotropic medications on admission. We estimated a negative binomial regression to examine the association between the number of sedative medications on admission and either a fall or injury diagnosis, including an interaction term for the fall and injury diagnoses. We used a generalized estimating equations approach to account for multiple hospitalizations by the same patients. We controlled for age, sex, length of stay, behavioral health diagnoses, race, ethnicity, interpreter needed, marital status, and Charlson comorbidity score (3, 4). All analyses were performed using Stata version 17. We also estimated marginal effects using the Stata margins command. The study was approved by the health system’s Institutional Review Board.