Errors in answering migraine patient-reported outcome measure questions: patients consider training necessary in clinical trials

Abstract

<p class="abstract"><strong>Background:</strong> In migraine clinical trials, patients’ understanding of the terminology used in patient-reported outcome (PRO) measures is important as variability in completing PRO measures can reduce the power to detect treatment efficacy. This study examines patients’ understanding of how to complete PRO measures in the absence of training, if minimal training can improve the accuracy of answering PRO items, and patients’ opinion on the necessity of training and their preference for the method of training.</p><p class="abstract"><strong>Methods:</strong> Participants reporting a diagnosis of migraine completed online surveys. Participants were given scenarios of how to report headache days and pain severity. Respondents were asked about their opinions on the necessity of training, and their preference for the method of training. In a second study, participants were given a hypothetical scenario on how to report pain severity before and after a short training.</p><p class="abstract"><strong>Results:</strong> The majority of participants had different criteria to interpret PRO questions and provided incorrect answers to our scenarios. In the second study, with minimal training, errors were reduced by 7.5%. Over 90% of participants viewed educational materials and training as necessary and preferred electronic modes of training with the ability to review training materials as needed for the duration of the trial.</p><p class="abstract"><strong>Conclusions: </strong>Patient training may improve data quality and inter-rater reliability in clinical trials. Electronic interactive training could be used as an approach to reduce inconsistencies in PRO measures and improve data quality.</p>