Predicting asthma-related crisis events using routine electronic healthcare data: a quantitative database analysis...

Area of Science:

• Respiratory Medicine
• Health Informatics
• Clinical Prediction Modeling

Background:

• No existing algorithm predicts asthma crisis events (accident and emergency attendance, hospitalisation, or death) using routinely available electronic health record (EHR) data.
• Asthma exacerbations pose a significant burden on healthcare systems and patients.

Purpose of the Study:

• To develop and validate an algorithm for identifying individuals at high risk of asthma crisis events.
• To leverage routinely collected electronic health record (EHR) data for predictive modeling in asthma management.

Main Methods:

• A multivariable logistic regression model was developed using primary care EHR data from 61,861 asthma patients in England and Scotland.
• External validation was performed on a separate dataset of 174,240 patients from Wales.
• Outcomes included hospitalisation, accident and emergency attendance, or death within a 12-month period.

Main Results:

• Key risk factors identified include previous hospitalisation, older age, underweight, smoking, and blood eosinophilia.
• The algorithm demonstrated acceptable predictive ability in the validation dataset (ROC 0.71).
• The algorithm can identify individuals with a 6.0% event risk compared to 1.1% in the general asthma population.

Conclusions:

• An externally validated algorithm effectively predicts high-risk asthma patients using EHR data.
• The algorithm can identify patients at high risk of asthma crisis events and exclude those not at high risk.
• This tool supports targeted interventions for patients at risk of severe asthma exacerbations.