Development and pilot of brief surveys to evaluate the quality of patient-centered cancer care.

Abstract

202 Background: Oncology practices have increasingly engaged in patient-centered systematic quality improvement efforts. One test of their effectiveness is patient feedback, yet many existing instruments are unwieldy or insensitive to change. Here we describe the development and pilot testing of short patient surveys used to assess change in patient experience of the key domains of access, communication, and coordination. Methods: Domains and corresponding questions of patient-centered care were informed by existing literature and tools available online, with a focus on domains relevant to cancer patients. Questions were modified from validated or semi-validated instruments to have a consistent 4 or 5-category Likert response scale, put into four unique short instruments that included at most ten questions each and piloted in cancer clinic waiting rooms. Input from patients was sought to assess: 1) relevance and clarity of questions; 2) whether important concepts were missing; 3) acceptability to patients regarding place and timing of surveys and; 4) estimate top box scores. Internal consistency of the final question sets for each domain was examined using Cronbach’s alpha. Results: Patient feedback from the pilot was positive with respect to all above categories. The following modifications were made based on top box score analysis: 1) created a common 5-point Likert scale for all questions and 2) excluded questions with a top box score higher than 80%. An open comment field was also added to complement the quantitative responses. Cronbach’s alpha for all domains was ≥ 0.80 expect for the access domain which was 0.51. The final 4 complementary instruments have 7-8 questions each and take only a few minutes for patients to complete. Each return patient receives one of the instruments randomly at clinic visits so information on all domains is continuously collected and a given patient has the opportunity to provide feedback on all domains over time; response rates are approximately 60%. Conclusions: The short surveys are being used to assess the effectiveness of interventions. Other cancer centers and their patients may find similar short surveys an improvement over general patient satisfaction surveys e.g., Press-Ganey.