Abstract
ObjectiveThe ability to predict impending asthma exacerbations may allow better utilization of healthcare resources, prevention of hospitalization and improve patient outcomes. We aimed to develop models using machine learning to predict risk of exacerbations. MethodsData from 29,396 asthma patients was collected from electronic medical records and national registers covering clinical and epidemiological factors (e.g. comorbidities, health care contacts), between 2000 and 2013. Machine-learning classifiers were used to create models to predict exacerbations within the next 15 days. Model selection was done using the mean cross validation score of area under precision-recall curve (AUPRC). ResultsThe most important predictors of exacerbation were comorbidity burden and previous exacerbations. Model validation on test data yielded an AUPRC = 0.007 (95% CI: ± 0.0002), indicating that historic clinical information alone may not be sufficient to predict a near future risk of asthma exacerbation. ConclusionsSupplementation with additional data on environmental triggers, (e.g. weather, pollen count, air quality) and from wearables, might be necessary to improve performance of the short-term predictive model to develop a more clinically useful tool.
Highlights
Asthma affects approximately 300 million individuals worldwide, including around 800,000 individuals in Sweden [1,2]
Exacerbations are of significant concern in asthma since they are the major cause of asthma related morbidity and mortality, as well major contributors of the asthma-related healthcare utilization [8,9]
Most existing models for predicting asthma exacerbations have been based on conventional logistic regression methodologies, which are known to perform poorly in situations where there is an imbalance in outcomes[19]
Summary
Asthma affects approximately 300 million individuals worldwide, including around 800,000 individuals in Sweden (approximately 8% of the Swedish population) [1,2]. The risk of exacerbations has epidemiologically been linked to multiple factors including history of previous exacerbations, asthma severity, respiratory viruses and aeroallergens, sex and co-morbidities[10,11,12,13,14]. Several pre diction models have been proposed to assess a risk of exacerbations in the asthma patients based on a variety of clinical and patient charac teristics [15]. Previous healthcare-utilization, symptoms, and spirom etry values have consistently been described as important predictors of asthma exacerbations. Most existing models for predicting asthma exacerbations have been based on conventional logistic regression methodologies, which are known to perform poorly in situations where there is an imbalance in outcomes (e.g. when exacerbations are rare events)[19]. More recently developed analytical methodologies with an ability to capture complex, non-linear relationships in data and interactions be tween predictors may achieve more accurate predictions[20], this is not always the case
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have