Predictors of treatment outcome in intermittent claudication.

OBJECTIVE To derive formulae to predict the likely 12-month health-related quality of life outcome following different treatments for intermittent claudication (IC). DESIGN A prospective, randomized, controlled study. MATERIALS One hundred and seventy-one unselected patients with stable IC were sequentially randomized to invasive therapy, supervised physical training or observation. Hierarchical analysis was used to identify significant predictors of outcome. RESULTS The strongest outcome predictors were baseline values of the respective outcome variables in all groups. No more than two significant secondary predictors were identified for each outcome variable and no outcome variable was a predictor of any other outcome variable. Resulting prediction equations achieved between 61 and 90% concordance with improvement (75% considered adequate), with best prediction for invasive therapy and poorest for observation. Suggested cutpoints for the various endpoints in the three groups had sensitivities ranging between 65 and 100% and false positive rates between 5 and 50%. CONCLUSIONS The derived equations adequately predicted improvement on the various outcome variables in invasive therapy and supervised physical training, and may serve as aids in selecting patients likely to benefit most from a particular treatment strategy. The uniqueness of the outcome variables underscores the importance of implementing a comprehensive set of endpoints relevant to the impacts of the condition.