Abstract
OBJECTIVES: Patient-reported outcome (PRO) measures have been used in epilepsy outpatient clinics in Denmark since 2011. The patients' self-reported PRO data are used by clinicians as a decision aid to support whether a patient needs contact with the outpatient clinic or not based on a PRO algorithm. Validity and reliability are fundamental to any PRO measurement used at the individual level in clinical practice. The aim of this study was to evaluate the test-retest reliability of the PRO algorithm used in epilepsy outpatient clinics and to analyse whether the method of administration (web and paper) would influence the result. DESIGN AND SETTING: Test-retest reliability study conducted in three epilepsy outpatient clinics in Central Denmark Region, Denmark. PARTICIPANTS: A total of 554 epilepsy outpatients aged 15 years or more were included from August 2016 to April 2017. The participants completed questionnaires at two time points and were randomly divided into four test-retest groups: web-web, paper-paper, web-paper and paper-web. In total, 166 patients completed web-web, 112 paper-paper, 239 web-paper and 37 paper-web. RESULTS: Weighted kappa with squared weight was 0.67 (95% CI 0.60 to 0.74) for the pooled PRO algorithm, and perfect agreement was observed in 82% (95% CI 78% to 85%) of the cases. There was a tendency towards higher test-retest reliability and agreement estimates within same method of administration (web-web or paper-paper) compared with a mixture of methods (web-paper and paper-web). CONCLUSIONS: The PRO algorithm used for clinical decision support in epilepsy outpatient clinics showed moderate to substantial test-retest reliability. Different methods of administration produced similar results, but an influence of change in administration method cannot be ruled out.