Background: A "survival" model offers certain ethical and practical advantages over alternative experimental designs if used to compare antiasthmatic inhaled steroid formulations. The model requires an objective daily measure of therapeutic effect (e.g., peak expiratory flow rate, which may be expressed as the lower of two daily measurements (LPF) or as diurnal variability (DVPF). The relative efficiency of these two measures is unknown.
Objective: This study was conducted to determine the relative efficiency of LPF and DVPF.
Methods: We analyzed data from a placebo-controlled comparison of an active inhaled formulation of budesonide versus an inactive oral formulation. The primary outcome measure in this design is the number of days from the time the test treatments start until a statistically significant deterioration occurs from an optimal asthma control value established at baseline.
Results: DVPF proved less sensitive than LPF; that is, fewer patients relapsed during the 8-week trial period: 32 versus 41, respectively. Also, the median interval until relapse was longer: 24 versus 9 days. With LPF, inhaled budesonide proved more effective than placebo or oral budesonide (p = 0.03), whereas DVPF failed to discriminate among the test treatments (p = 0.38). LPF correlated with all three symptom indices (p > or = 0.003) and two of three spirometric indices measured concomitantly (p < or = 0.04). DVPF did not correlate with any index (p > or = 0.10).
Conclusion: In this experimental model, LPF proved more sensitive and valid than DVPF as an indicator of differences in antiasthmatic potency between two inhaled steroid formulations.