OBJECTIVES : Traditionally, economic models developed for the health technology appraisal of new COPD interventions have focussed on tracking disease progression according to lung function through GOLD stages I-IV. In this study, a new model structure was developed, focussing on the benefit of preventing acute exacerbation events, independent of lung function. For patients who have severe airflow obstruction and an exacerbating phenotype their disease progression is not adequately captured by GOLD stages, as they are unlikely to experience improvements in lung function, are already maximally bronchodilated, and are in need of new interventions to reduce the risk of future exacerbations and the downward spiral in overall health status.

METHODS A de novo model was conceptualised to capture the long-term sequelae of COPD exacerbations, including the downstream patient and healthcare system impact. To capture these long-term outcomes, it was necessary to source data with sufficient events and follow-up to correctly characterise pay-offs and risk by time since an exacerbation. A meta-analysis of individual patient-level data (IPD) from historical randomised controlled trials and observational COPD studies was conducted to obtain these model inputs.

RESULTS A Markov model with tunnel states was developed in MS Excel. Health states were defined by exacerbation type and frequency (none, moderate, severe, frequent severe) and tunnel states depicting time since an exacerbation in three-monthly increments for 12 months were included. Frequent severe exacerbations were defined as those occurring within 12 months of a previous severe exacerbation. The model has the flexibility to include treatment discontinuation/stopping rules. The biggest drivers of the model results are treatment efficacy and treatment duration.

CONCLUSIONS By leveraging IPD available through historical COPD studies, it was possible to develop a Markov-based economic model to robustly examine the long-term occurrence and consequences of exacerbations, independent of lung function, for patients with severe airflow obstruction.