Rescuing Suboptimal Patient-Reported Outcome Instrument Data in Clinical Trials: A New Strategy

Background: Psychometric instruments such as the Repeated Battery for the Assessment of Neuropsychological Status (RBANS) are commonly used under conditions for which they were not developed or validated. They may then generate troublesome data that could conceal potential findings. Methods: Based on a previously published refinement of the RBANS, we reanalyzed the data on 303 patients from two National Institutes of Health (NIH) trails in Parkinson's disease and contrasted the results using the original versus refined scores. Results: Findings from the original RBANS scores were inconsistent; however, use of the refined scores produced potential findings that were in agreement with independent reports. Conclusion: This study demonstrates that, for negative trials using instrument scores as primary outcomes, it is possible to rescue potential findings. The key to this new strategy is to validate and refine the instrument for the specific disease and conditions under study and then to reanalyze the data. This study offers a demonstration of this new strategy for general approaches.