Fall Ascertainment and Development of a Risk Prediction Model Using Electronic Medical Records

Area of Science:

• Gerontology
• Health Informatics
• Epidemiology

Background:

• Falls are a significant health concern for older adults, leading to injury and reduced quality of life.
• Electronic Medical Records (EMR) offer a rich data source for studying health outcomes and developing predictive models.
• Accurate identification of fall events and prediction of fall risk are crucial for effective prevention strategies.

Purpose of the Study:

• To investigate the utility of Electronic Medical Record (EMR) data for identifying falls in an elderly population.
• To develop and validate a fall risk prediction model using EMR data.
• To assess the performance of the developed fall risk prediction model.

Main Methods:

• A retrospective longitudinal study analyzed 10 years of EMR data from Kaiser Permanente Hawaii (2004-2014).
• Fall events were identified using International Classification of Diseases, Ninth Revision (ICD-9) codes and/or primary reason for visit documentation.
• Logistic regression and LASSO regression were employed to build and select predictors for the fall risk model.

Main Results:

• A total of 57,678 adults aged 60 years and older were included in the study.
• The final fall risk prediction model, utilizing 13 key predictors, achieved a sensitivity of 67%, specificity of 69%, and an area under the curve of 0.74.
• Key predictors included age, comorbidities, female sex, walking issues, and polypharmacy.

Conclusions:

• EMR data can be effectively utilized to ascertain falls and develop a fall risk prediction model with moderate predictive capabilities.
• Enhancing fall documentation within EMR systems through collaboration with clinical providers is recommended to improve fall capture and model accuracy.
• The developed model demonstrates the potential of EMR data in identifying older adults at risk of falls, paving the way for targeted interventions.