Development and evaluation of a simple predictive model for falls in acute care setting

Area of Science:

• Gerontology and Geriatric Medicine
• Patient Safety and Quality Improvement
• Clinical Nursing Research

Background:

• Patient falls in acute care settings pose significant risks, leading to injuries, prolonged hospital stays, and increased healthcare costs.
• Existing fall prediction tools often lack the simplicity and reliability required for practical application in busy acute care environments.
• A need exists for a straightforward and dependable assessment tool to identify patients at high risk of falling.

Purpose of the Study:

• To develop and validate a simple, reliable assessment tool for predicting falls among patients in acute care settings.
• To identify key predictors of falls that can be easily incorporated into routine clinical practice.

Main Methods:

• A retrospective cohort study was conducted using data from a Japanese teaching hospital.
• The modified Japanese Nursing Association Fall Risk Assessment Tool (initially 50 variables) was utilized, with variables reduced to 26.
• Stepwise logistic regression analysis selected six key variables. The dataset was split (7:3) for model derivation and validation, with performance evaluated using sensitivity, specificity, and ROC curve analysis.

Main Results:

• A six-item predictive model was developed, incorporating age (>65 years), impaired extremities, muscle weakness, need for mobility assistance, unstable gait, and psychotropic medication use.
• A cut-off score of 2 (one point per item) was established for the model.
• The model demonstrated high performance in the validation dataset, with sensitivity and specificity exceeding 70% and an area under the curve greater than 0.78.

Conclusions:

• A simple and reliable six-item model has been successfully developed to predict high fall risk in acute care patients.
• The model's performance was robust, even with non-random data partitioning by time, suggesting its potential for real-world application.
• This validated tool is poised to enhance fall prevention strategies and improve patient safety in clinical practice.