Applying cognitive debriefing to pre-test patient-reported outcomes in older people with multiple sclerosis

Purpose The aim of this study was to reduce respondent burden and decrease data errors in patient-reported outcomes (PROs) intended for a postal survey measuring health and lifestyle factors that may affect quality of life in older people with multiple sclerosis (MS). Methods Participants (n = 18) were recruited from a database of outpatient visits. Using the qualitative diagnostic method, cognitive debriefing, participants completed five standardized questionnaires; Frenchay Activities Index, Barthel Index, Simple Lifestyle Indicator Questionnaire, EuroQoL EQ-5D and Personal Resources Questionnaire 2000. PRO item issues and respondent behaviors such as skipping items were recorded. Data collection was an iterative process whereby difficulties experienced by 2-3 subjects were used to modify the survey for following respondents until data saturation was reached. Results Most respondents had serious difficulties with at least one PRO item. Response errors fell into three main categories: (1) respondents did not read instructions and completed the item incorrectly, (2) respondents did not understand the question and required examples to clarify and (3) respondents felt that the pre-determined response options did not apply to them. PRO reformatting, minor modifications to item wording and addition of item examples improved precision and reduced respondent burden. Conclusion Our findings support the notion that methods such as cognitive debriefing help improve precision of selfreported measures in a special population such as ours.