Understanding disparities and organisational variation in cancer patient experience: Lessons from the English National Cancer Patient Experience survey.

Abstract

181 Background: Surveys of the experience of cancer patients are currently being introduced in several countries, including the U.S. and Canada. Insights to inform the development and use of such surveys can be acquired from the English Cancer Patient Experience Survey programme. Methods: Each of the three national surveys (2010, 2011/12 and 2012/13) had a responder sample of ~70,000 patients (response rate 64-67%) and included about 60 evaluative questions on all domains of experience, from cancer diagnosis to care after hospital treatment. Anonymous data were analysed, using appropriate logistic regression models for positive/negative experience outcomes, based on public reporting conventions. We illustrate a research programme using the survey data focusing on two questions. 1. Who are the patients at greater risk of a negative experience? 2. Does case-mix explain apparent variation between hospitals? Results: There was evidence of socio-demographic variation in cancer patient experience, consistently across questions: younger and very old patients, ethnic minorities, and women reported poorer experience. Regarding variation between patients with 36 common and rarer tumours, those with hepato-biliary and thyroid cancers were most likely to be less satisfied with their overall experience, whereas patients with breast cancer were least likely (top-to-bottom odds ratio 3.7, p<0.0001). There were disparities in experience between patients treated by the same specialty for 5/9 services (p<0.0001). For example, patients with ovarian, multiple myeloma, anal and renal cancer reported worse experiences than patients with other gynaecological, haematological and urological malignancies, respectively. There was high concordance between crude and adjusted ranks of hospital scores (median Kendall's τ=0.84; interquartile range: 0.82-0.88). Conclusions: Initiatives to improve cancer patient experience may be suitably targeted on patients at higher risk of poorer experience. Evidence about disparities in experience and the effect of case-mix provide useful priors to motivate similar inquiries in other healthcare settings.