Evaluating longitudinal therapy effects via the North Star Ambulatory Assessment

INTRODUCTION/AIMS: In comparative studies, treatment effects are typically evaluated at a specific time point. When data are collected periodically, an alternative, clinically meaningful approach could be used to assess the totality of treatment effects. We applied a well‐developed analytical procedure for evaluating longitudinal treatment effects using North Star Ambulatory Assessment (NSAA) data for illustration. METHODS: The NSAA comprises 17 scorable items/outcomes that measure changes in motor function. Using NSAA data from the published ataluren phase 3, randomized, placebo‐controlled trial (NCT01826487), cumulative counts of failures to perform each item (transition from 2/1 [able/impaired] to 0 [unable]) were collected at specified time points for each patient over 48 wk. Treatment group‐wise mean cumulative item failure count curves were constructed, comparing ataluren versus placebo and deflazacort versus prednisone/prednisolone among placebo‐treated patients. The steeper the curve, the worse the outcome. A clinically meaningful summary of the between‐group difference was provided for each comparison. RESULTS: The curve was uniformly steeper for placebo than ataluren after 16 wk and for prednisone/prednisolone than deflazacort after 8 wk. The two curves in each comparison continued to diverge thereafter, indicating sustained treatment benefits over time. Using a unique analytical approach, cumulative failure rates were reduced, on average, by 27% for ataluren versus placebo (rate ratio, 0.73; 95% confidence interval [CI], 0.55–0.97; p = .027) and 28% for deflazacort versus prednisone/prednisolone (rate ratio, 0.72; 95% CI, 0.53–0.96; p = .028). DISCUSSION: Unlike fixed‐time analyses, this analytical approach enabled demonstration of cumulative, longitudinal treatment effects over time using repeatedly measured NSAA observations.