Philosophies of Surgical Care Are Embedded in Outcome Studies: An Illustrative Reanalysis of the Cartiva MOTION Trial

Abstract

Background:

Subjective assumptions on the definition of surgical success are inherent to the design of clinical trials with a categorial outcome. The current study used reasonable alternative assumptions about surgical care to reassess data for the randomized controlled Cartiva trial (MOTION).

Methods:

Data from the published study were augmented by publicly accessible internal US Food and Drug Administration documents. As in the published report, 1-sided lower bound 95% CIs (LBCI95) for the difference of proportions were calculated for a series of alternative scenarios in which the assumptions underlying what constitutes surgical success were altered.

Results:

Using a noninferiority margin of −15%, the MOTION trial reported success based on a 1-sided LBCI95 of −10.9%. Each of the 3 independent alternative scenarios analyzed yielded results that altered the primary outcome of the trial: (1) eliminating failures based solely upon radiographs findings, thereby considering a painless pseudarthrosis as a success (1-sided LBCI95 of −15.9%), (2) considering only major surgical revision as a failure and discounting isolated hardware removal (1-sided LBCI95 of −15.1%), and (3) using a visual analog scale (VAS) pain threshold of <30 as the success criterion rather than a 30% reduction in VAS pain score (1-sided LBCI95 of −15.8%).

Conclusion:

In this reanalysis, applying any of 3 reasonable alternative assumptions about the definition of surgical success to the data resulted in failure to prove noninferiority of Cartiva over arthrodesis, a reversal of the reported trial result. These results highlight the effect of subjective assumptions in the design of clinical trials with a categorical outcome and illustrate how differing philosophies about what constitutes surgical success can be pivotal in determining the final result.

Level of Evidence:

Level II, prospective comparative study.