A health policy model for COPD

Objectives: 1) To improve an existing COPD model by incorporating the distinction between mild, moderate, severe and very severe COPD and by quantifying the progression of COPD over these stages 2) To use the improved model to estimate the potential impact of smoking cessation programs offered to COPD patients and project their effect on the future burden of COPD. Methods: An existing population model for COPD, which is a module of the RIVM Chronic Disease model, was extended with disease progression over time. Prevalent cases in the starting year were distributed over 4 severity stages mild (28%), moderate (54%), severe (15%) and very severe (3%) (GOLD-classification). The severity distribution was based on data from GP registrations. The COPD incidence was 41% in mild, 55% in moderate and 4% in severe. Disease progression was modelled as annual decline in lung function in FEV1% predicted. The Lung Health Study was used to estimate gender, age, smoking and baseline FEV1% predicted dependent values of lung function decline and one-time increase in lung function associated with smoking cessation. A meta-analysis was done to obtain severity stage specific mortality rates. The new model was used to project COPD prevalence, mortality and costs by COPD severity stage over the period 2000-2025 (the base-case scenario). A series of sensitivity analyses was performed to assess the robustness of the results to changes in input data and assumptions. The new model was used to compare two scenarios on increased implementation of two smoking cessation interventions, minimal counselling by the general practioner (H-MIS) and intensive counselling with bupropion (IC+Bupr). They were compared to the base-case scenario in terms of life-years, QALYs, interventions costs and savings of COPD-related costs. In the scenarios H-MIS or IC+Bupr was implemented for a period of either 1 year, 10 years or 25 years and reached 25% of the smokers. Smoking cessation results in a one-time increase in lung function and a lower annual decline in FEV1% predicted, which results in less disease progression and less mortality among COPD patients who quit smoking. Future costs and effects of these scenarios were discounted at 4%. Incremental cost-effectiveness ratios were calculated as (additional intervention costs minus the savings in COPD-related health care costs)/ gain in health outcomes. Results: In the base-case scenario, the total number of COPD patients increases from 300 thousand in 2000 to 490 thousand patients in 2025. Between 2000 and 2025 the prevalence rate of mild COPD increases from 5 to 11 per 1000 inhabitants. The prevalence rate of moderate COPD increases from 11 to 14. For severe COPD the rate increases from 3.0 to 3.9 and for very severe COPD the rate increases from 0.5 to 1.3. In absolute numbers the increase is highest in mild COPD, but the largest relative increase in prevalence rate is seen in very severe COPD. As a result of the increase in COPD prevalence and aging of the COPD population, all-cause mortality rates per 1000 inhabitants increase in all severity stages. In 2000, total COPD-related health care costs are estimated to be 280 million Euros. In 2025 total costs are projected to be 495 million Euros. Costs for very severe COPD have the highest relative increase. The sensitivity analyses show that the model projections were most sensitive to assumptions about the severity distribution of incidence. Implementation of H-MIS and IC+Bupr results in more mild and moderate and less severe and very severe COPD patients compared to the base-case scenario after 25 years. Costs per additional quitter are 700 for H-MIS and 2700 for IC+Bupr. Irrespective of the duration of implementation, H-MIS generates net savings, which indicates that the intervention costs of H-MIS are offset by the savings in COPD-related costs. For IC+Bupr savings do not outweigh the interventions costs. For the years 2000 to 2025 the costs per life-year gained of implementing IC+Bupr for 10 years are estimated to be 12000 Euros. Conclusions: Modelling COPD progression over time proves feasible. The model showed that implementation of H-MIS among COPD patients results in better health outcomes and is cost saving. Implementation of IC+Bupr has higher costs than savings, but is still cost-effective with costs per life-year ranging from 10600 to 24500 depending on the duration of implementation.