Analytical approach:
This economic evaluation was based on two decision tree models, which captured the clinical and economic impact of different diagnosis and management or treatment options for asthma over a one-year time horizon. The authors stated that the analysis was carried out from the perspective of the German health care payer.
Effectiveness data:
The clinical data appear to have been derived from a selection of known, relevant published studies. The data on test accuracy were obtained from a diagnostic study, while randomised controlled trials were used to assess the impact of FENO on the management of patients with asthma. The authors provided a few details on the sample sizes in these studies, but no other information was provided. The key clinical endpoint was the accuracy of the diagnostic tests.
Monetary benefit and utility valuations:
The utility estimates were derived from a study, which elicited preferences for health conditions from a sample of 228 adult asthma patients using the European Quality of life (EQ-5D) questionnaire.
Measure of benefit:
Quality-adjusted life-years (QALYs) were used as the summary benefit measure, in the asthma management model.
Cost data:
The health service costs were those of diagnostic tests, in-patient and out-patient visits to health care professionals, hospitalisations, and drug therapies. The unit costs were reported and were based on official reimbursement rates. The resource use was based on published studies and authors’ assumptions. The calculation of drug costs was based on averages across recommended dosages and typical treatments. All costs were in Euros (EUR) and the price year was not explicitly reported but may have been 2006.
Analysis of uncertainty:
A series of univariate sensitivity analyses was carried out on the key model inputs to determine whether the model findings were robust. The alternative ranges of values for these were defined by the authors. A second-order sensitivity analysis was also undertaken by conducting 1,000 simulations for the relevant model inputs.