Cost effectiveness of fluticasone propionate and flunisolide in the treatment of corticosteroid-naive patients with moderate asthma

Objective: The aim of this study was to determine the cost effectiveness of 2 inhaled corticosteroids, fluticasone propionate and flunisolide, in the management of asthma from a third-party payer perspective in Germany (Germa n Sickness Fund). Design and setting: Direct treatment costs were retrospectively applied to 2 prospective randomised parallel group clinical trials conducted in German y comparing fluticasone propionate and flunisolide: one 6-week open-label s tudy (n = 332) and one 8-week double-blind study (n = 308) in corticosteroi d-naive patients with asthma of moderate severity aged between 18 and 70 ye ars. All costs were adjusted to 1997 Deutschmarks. Efficacy parameters incl uded changes in morning and evening peak expiratory flow rate (PEFR) measur ements, the number of successfully treated patients (defined as those with a PEFR improvement of greater than or equal to 10%) and proportion of sympt om-free days. Main outcome measures and results: The fluticasone propionate groups had hi gher respective proportions of successfully treated patients and symptom-fr ee days than the flunisolide groups in both the open-label (56.8 vs 39.6% a nd 36.4 vs 28.5%) and double-blind (55.3 vs 44.5% and 35.1 vs 31.1%) studie s. Improvements in both morning and evening PEFR measurements were also sig nificantly (p < 0.01) greater with fluticasone propionate than with fluniso lide. Although average daily treatment costs were slightly higher in the fl uticasone propionate groups than in the flunisolide groups, all cost-effect iveness ratios (daily cost per successfully treated patient and daily cost per symptom-free day) favoured fluticasone propionate. Sensitivity analysis showed that these results were robust over a wide range of assumptions. Conclusion: In these patients, management with fluticasone propionate was m ore cost effective than with flunisolide in the German healthcare setting.