Predictors of electronic patient-reported outcomes use in the survivorship setting.

Abstract

e14038 Background: Patient-reported outcomes (PRO) offer insight into patient perception of health and symptom burden. Despite a shift toward electronic PRO (ePRO), optimal administration methods are unclear. Our institution recently began ePRO collection in survivorship clinics: patients are invited via email to complete a health survey on our online patient portal prior to annual visits, enabling clinician review of symptoms in advance of the visit. Patients who do not complete an ePRO survey at home are offered an iPad or paper survey at visit check-in. In the first year of ePRO, 87 patients inadvertently submitted multiple responses to the questionnaire, across two modalities. This study aimed to 1) assess determinants of ePRO completion across modalities (portal, iPad, paper); and, 2) among patients who submitted multiple surveys, compare consistency of responses in surveys completed within 30 days of each other. Methods: We reviewed records for 10194 patients seen in breast, thoracic, colorectal, and gynecologic survivorship clinics over one year. Demographics, disease/treatment details, and PRO responses (symptoms, health behaviors, etc.) were extracted. For aim 1, we used multivariate regression to determine predictors of completion method. For aim 2, we calculated Cohen’s kappa coefficients to compare responses based on completion modality. Results: Most patients (65.6%) completed the survey on an iPad in clinic; 16.7% on the portal, 17.7% on paper in clinic. Younger age (p < .001), white race (p < .001), less fatigue (p = .01), and English as primary language (p < .001) were associated with portal use in multivariate analyses. In general, Cohen’s Kappa analyses revealed high agreement between surveys. Conclusions: Our findings highlight demographic gaps in ePRO acceptance. Although most patients completed an ePRO (portal or iPad), few completed it at home in advance of their visit, which has implications for clinic flow and clinician preparation for visits. However, our finding of consistent symptom reporting across mode and location of completion is reassuring. Future work should seek to improve comfort with ePRO completion at home among groups less likely to accept it and explore the implications of symptom burden on ePRO acceptance.