# Psychotropic medication and in-hospital falls in older adults: a cohort-based secondary analysis with exploratory stratification among users

## TL;DR

Older adults taking psychotropic medications are at higher risk of falling in hospitals, with factors like age and eyesight playing a role.

## Contribution

This study identifies psychotropic medication use as a significant risk factor for in-hospital falls and develops a predictive model incorporating clinical indicators.

## Key findings

- Patients using psychotropic medications had a significantly higher fall rate (7.1%) compared to non-users (2.1%).
- Age, eyesight, and admission days were identified as independent risk factors for falls.
- A predictive model with an AUC of 0.811 was developed to assess fall risk in psychotropic medication users.

## Abstract

People using psychotropic medication have a higher risk of falls in the hospital environment. This study aims to investigate and assess the risk of falls in the people using psychotropic medication in hospitals. A secondary analysis was conducted to study and analyze the factors related to the risk of falls in the people using psychotropic medication collected from the Department of Health and Preventive Medicine in Fukushima City, Japan, between August 2008 and September 2009. Clinical records collected from structured questionnaires administered through face-to-face interviews conducted by nurses and doctors were used, and fall events were collected from the clinical records. Univariate and multivariate regression analyses were used to evaluate fall risk factors. ROC curves were used for predictive analysis of factors influencing falls. A total of 606 patients using psychotropic medications, of whom 43 (7.1%) experienced falls. Among the remaining 8863 patients, 187 (2.1%) experienced falls, and there was a significant difference between the two groups (P < 0.001). Multivariate regression analysis showed that age, eyesight, and admission days were independent risk factors for falls (all P < 0.05). The Model constructed with 11 indicators showed a good ROC curve (AUC = 0.811), with a sensitivity of 90.48% and specificity of 64.86%. Patients receiving psychotropic medications had a higher risk of in-hospital falls than those not receiving such medications. Within this subgroup, a multivariable model incorporating age, planned surgery, ADL (evacuation), and length of stay was used to quantify the association and showed acceptable internal performance; we suggest interpreting the model as an analytical aid.

The online version contains supplementary material available at 10.1038/s41598-025-31320-7.