Optimal assessment of baseline treadmill walking performance in claudication clinical trials

Treadmill testing is frequently used to assess the functional capacity of patients with claudication, but the optimal application of treadmill testing in the setting of multicenter clinical trials remains uncertain. The current study used data from a recent clinical trial of the drug NM-702, which employed three baseline assessments of peak walking time (PWT) using a graded treadmill. These data were used to describe the different methods of defining the baseline peak treadmill performance with respect to reproducibility, stability over time and detection of treatment effect. A series of baseline definitions (first test only, last test only, highest PWT of the three tests, arithmetic mean of the three tests, mean of the first two tests, median of the three tests and a reproducibility-based criterion) were used to calculate the population (n = 386) variability in baseline testing, the placebo response over the 24 weeks of treatment, and the effect size of NM-702. Placebo responses and NM-702 effect sizes were not substantively affected by the method used to calculate baseline PWT. Changes in PWT on placebo were less than 25% for all methods of baseline quantitation. No method yielded an NM-702 effect size quantitatively greater than that obtained using only the first baseline test in the analysis for either PWT or claudication onset time. The graded treadmill test quantifies PWT with high reproducibility and stability over time. These characteristics may obviate the need for multiple treadmill tests, potentially saving study costs and improving patient acceptance of trial participation.