Unwarranted Between-hospital Variation in Mortality, Readmission, and Length of Stay of Urological Admissions: An Important Trigger for Prioritising Quality Targets

Abstract

BackgroundUnwarranted between-hospital variation is a persistent health care quality issue. It is unknown whether urology patients are prone to this variation. ObjectiveTo examine between-hospital variation in mortality, readmission, and length of stay for all 22 urological All Patient Refined Diagnosis Related Groups (APR-DRGs). Design, setting, and participantsThis study included administrative data from 320640 urological admissions in 99 (98%) Belgian acute-care hospitals between 2016 and 2018. Outcome measurements and statistical analysisWe used hierarchical mixed-effect logistic regression models to estimate hospital-specific and APR-DRG–specific risk-standardised rates for in-hospital mortality, 30-d readmission, and length of stay above the APR-DRG–specific 90th percentile. Between-hospital variation was assessed based on the estimated variance components. Associations of outcomes with patient and hospital characteristics and time trends were examined. Results and limitationsOur analysis revealed important between-hospital variation in mortality, readmission, and length of stay for urological pathologies, particularly for medical diagnoses. Significant variation was shown in all three outcomes for kidney and urinary tract infections; other kidney and urinary tract diagnoses, signs, and symptoms; urinary stones and acquired upper urinary tract obstruction; and kidney and urinary tract procedures for nonmalignancy. Lowering of mortality rates in upper-quartile hospitals to the median could potentially save 41.5% of deaths in these hospitals, with the largest absolute gain for kidney and urinary tract infections and kidney and urinary tract malignancy. Limitations included a likely underestimation of readmission rates. ConclusionsUrological patient outcomes are characterised by unwarranted between-hospital variation. We recommend improvement initiatives to prioritise kidney and urinary tract infections because of significant variation across the three outcomes and the largest potential gain in lives saved. Patient summaryWe found notable between-hospital variation in mortality, readmission, and length of stay for urological hospital admissions in Belgium. As much as 41.5% of deaths could potentially be avoided if underperforming hospitals improved. Targeting kidney and urinary tract infections could help reduce variation.