Patient surveys for cancer populations: Initial experiences with Cancer CAHPS.

281 Background: Patient experiences are an important component of quality, but information on how well oncology providers meet the specific needs of patients with cancer is limited. This study reports on the patient experiences among patients undergoing drug therapy treatment for cancer in diverse outpatient settings.

METHODS: We assessed patient experiences using a modified version of the CAHPS Survey for Cancer Caredeveloped by the American Institutes for Research.The survey has five composites and an overall rating of providers. We used claims data from an insurance plan in Southeastern Pennsylvania to identify all patients who had a paid claim for drug therapy treatment at 26 practices from January 1 through July 31, 2014. The survey was sent to 2,302 patients who met this criteria and 701 patients returned a completed survey for a response rate of 31%. We calculated scores of 0-100 on the composites after adjusting for age, education, self-rating of mental health and response tendency (a measure of level of optimism/pessimism).

RESULTS: In general, cancer patients rated their providers highly on patient experience, particularly on access and communication. However, cancer patients rated their providers lower on shared decision-making and helping patients manage their symptoms and care at home. The scaled, adjusted scores on the five composites were: Access (score=91; 95% CI=90-92); Affective communication (score=90; 95% CI=88-91); Exchanging information/keeping patients up to date (score=89; 95% CI=88-90); Shared decision-making (score=78; 95% CI=76-80); Enabling patient self-management (score=69; 95% CI=67-71). We encountered several challenges in fielding the survey related to attributing patients to specific practice sites, identifying patients who had died, and concerns raised by patients/family respondents about the length or content of the survey.

CONCLUSIONS: Our findings suggest the need for strong efforts to engage cancer patients in decision-making and provide supportive care. Methodology for sampling patients using claims data needs to be more accurate and the survey could be edited for length and comprehension.