Objectives To examine variation in lengths of stay between hospitals supplying renal transplant services in England, accounting for hospital and patient characteristics to establish whether there are statistically significant differences that would justify paying different local prices. Design Analysis of retrospective administrative data using multi-level models including hospital specific and patient specific characteristics. Graphical presentation of adjusted provider-level lengths of stay and associated confidence intervals. Setting Renal transplants undertaken in the financial year 2019/20 across 19 providers in England. Population 3,444 patients who were subject to renal transplants for which 58% were male and whose average age was 51 years. Data Hospital Episode Statistics (Admitted Patient Care) and Civil Registration Deaths data for England 2019/2020. Main Outcome Measures Estimated hospital level lengths of stay for patients and associated standard errors, after accounting for a suite of hospital level and patient level variables. Results After controlling for hospital characteristics and patient characteristics there is little statistically significant variation in length of stays between hospitals. Only 3 hospitals exhibit lengths of stay that are significantly longer or shorter than the overall average. Hence, most observed variability can be attributed to factors that are external to a hospital or outside of its control. Conclusions There is little support from these data for the use of different locality specific prices with a view to compensating for (or providing incentives to align) differences is efficiency across locations in respect of renal transplant services. Any discretion in terms of price setting will need to be supported by data supplementary to the available administrative data analysed in this study.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)