STANDARDIZING CRITERIA FOR COGNITIVE ASSESSMENT OF PAPER-TO-ELECTRONIC EQUIVALENCE OF PROS
Author(s)
Martin ML, McCarrier KP
Health Research Associates, Inc., Seattle, WA, USA
Presentation Documents
AIMS: Use of electronically administered outcome measures is increasing in clinical trial data collection. The 2009 ISPOR ePRO Task Force Report recommends cognitive assessment as sufficient evidence of equivalence when only minor changes have been made to the measure. No consensus exists, however, for how cognitive assessment results should be evaluated to determine if equivalence has been established. METHODS: Existing literature around equivalence assessment for PROs was combined with firsthand experience in conducting over 500 cognitive interviews aimed at assessing paper and electronic equivalence of 55 different PRO instruments. Using these two resources, we developed suggested criteria to cognitively assess equivalence between the two modes, and present a practical process for meeting these criteria. RESULTS: The criteria for determining equivalence between paper and ePRO formats should focus primarily on whether or not the ePRO is likely to produce data substantially different if administered via one method versus the other. To determine whether such a risk to ePRO data exists, we propose a three-step process for interpreting cognitive interview responses. 1) Determine whether a patient perceives a cognitive difference between paper PRO and ePRO, and whether that difference represents a variation in the understanding of the item or simply a recognition of different appearance; 2) Determine whether any difference in patient understanding has a meaningful impact on their response to the item, and 3) determine whether or not that difference presents a significant risk to the data that justifies modification of the ePRO. CONCLUSIONS: In the years since the ISPOR ePRO Task Force issued their recommendations, data capture technology has continued to develop while clarity on methods to evaluate the cognitive impact of this technology have lagged. The steps we present provide a standardized system for ensuring that ePROs measure what is intended and the risk to research data is minimized.
Conference/Value in Health Info
2014-05, ISPOR 2014, Palais des Congres de Montreal
Value in Health, Vol. 17, No. 3 (May 2014)
Code
PRM134
Topic
Methodological & Statistical Research
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
Multiple Diseases