Development and validation of an asthma exacerbation prediction model using electronic health record (EHR) data

Area of Science:

• Pulmonary Medicine
• Health Informatics
• Biostatistics

Background:

• Asthma exacerbations lead to significant morbidity, mortality, and healthcare costs.
• Accurate identification of high-risk asthma patients is crucial for proactive management.
• Electronic Health Records (EHRs) offer a valuable data source for risk prediction.

Purpose of the Study:

• To develop and validate a risk prediction tool for asthma exacerbations.
• To identify key factors associated with asthma exacerbations using EHR data.
• To assess the clinical utility of a predictive model for asthma exacerbations.

Main Methods:

• Retrospective analysis of structured EHR data from 37,675 asthma patients.
• Identification of exacerbation predictors through univariable and multivariable statistical analysis.
• Development and testing of a risk prediction model using prescription fill data.

Main Results:

• Smoking, allergy testing, obesity, and uncontrolled asthma (low ACT score) were associated with increased exacerbation risk.
• The developed risk prediction model achieved an Area Under the Curve (AUC) of 0.67.
• A significant proportion (54.8%) of the study cohort experienced an asthma exacerbation.

Conclusions:

• The developed asthma exacerbation risk prediction model did not achieve sufficient accuracy for clinical use.
• Despite rigorous methodology, the predictive performance was limited.
• Further research is needed to develop clinically useful tools for identifying high-risk asthma patients.