Modeling severe uncontrolled asthma: Transitioning away from health states

Abstract

Background

Models developed to date to simulate long-term outcomes of asthma have been criticized for lacking granularity and ignoring disease heterogeneity.

Objective

To propose an alternative approach to modeling asthma and apply it to model long-term outcomes in a population with moderate-to-severe type 2 asthma (patients with raised fractional exhaled nitric oxide or eosinophils) and treated with conventional therapy.

Methods

A discretely integrated condition event (DICE) approach was adopted, simulating individual profiles with asthma over patients’ lifetime in terms of exacerbations, asthma-related death, and death unrelated to asthma. The timing of these events is dependent on profile characteristics including lung function, asthma control, exacerbation history, and other baseline characteristics or contextual factors. Predictive equations were derived from a clinical trial to model time to exacerbation, change in asthma control, lung function, and utility. Real-world studies were used to supplement data gaps. Outcomes evaluated included life expectancy, quality-adjusted life-years (QALY), number of exacerbations, and lung function over time.

Results

Average annual rates of severe and moderate exacerbations were 1.82 and 3.08 respectively, with rates increasing over time. Lung function declined at a higher rate compared with the general population. Average life expectancy was 75.2 years, compared with 82.4 years in a matched general population. The majority of life-years were spent with uncontrolled asthma and impaired lung function.

Conclusion

Patients with moderate-to-severe type 2 asthma and a history of exacerbations suffer from frequent exacerbations and reduced lung function and life expectancy. Capturing multiple conditions to simulate long-term outcomes in patients with asthma may provide more realistic projections of exacerbation rates.