# Development of a multivariable prognostic prediction model for skin tears in older nursing home residents

## TL;DR

This study develops a model to predict skin tears in older nursing home residents based on factors like BMI and skin dryness.

## Contribution

The novel contribution is a validated predictive model for skin tear risk in elderly nursing home residents.

## Key findings

- Lower BMI and Barthel Index scores are significant risk factors for skin tears.
- Xerosis cutis on the legs and regular corticosteroid use also increase risk.
- The model shows good predictive accuracy with an area under the curve of 0.823.

## Abstract

Skin tears are traumatic wounds and are among the most prevalent skin conditions in older adults, particularly those in long-term care facilities. These injuries can lead to complications such as infection, pain, reduced quality of life, and increased healthcare costs. This study aimed to identify risk factors for skin tear development in nursing home residents aged 65 years or older and to develop a predictive prognostic model. A secondary data analysis was performed on long-term care nursing home residents ≥ 65 years who participated in a cluster-randomized controlled clinical trial conducted in Berlin, Germany, from April 2019 to June 2021. A total of 101 residents were included. At week 12, 19 residents (18.8%) developed at least one skin tear. The best-fit predictive model identified lower Body Mass Index, lower Barthel Index scores, presence of xerosis cutis on the legs, and regular corticosteroid use as significant risk factors for skin tear development. The model demonstrated good discriminatory ability (area under the curve: 0.823), with sensitivity and specificity rates of 73.7% and 74.4%, respectively. These risk factors could help identify at-risk individuals, enabling targeted preventive measures. However, the model requires validation in a prospective cohort to confirm its applicability in clinical practice.

The online version contains supplementary material available at 10.1038/s41598-025-95944-5.