Unwarranted clinical variation in the care of children and young people hospitalised for injury: a population-based cohort study

Abstract

IntroductionInjury is a leading cause of death and disability among children and young people. Recovery may be negatively affected by unwarranted clinical variation such as representation to an emergency department (ED), readmission to a hospital, and mortality. The aim of this study was to examine unwarranted clinical variation across providers of care of children and young people who were hospitalised for injury in New South Wales (NSW). Materials and MethodsRetrospective population-based cohort study using linked ED, hospital, and mortality data of all children and young people aged ≤25 years who were injured and hospitalised during 1 January 2010–30 June 2014 in NSW. Unwarranted clinical variation across providers was examined using three indicators. That is, for each hospital that treated ≥100 cases per year, risk standardised ratios were calculated with 95% and 99.8% confidence limits using the number of observed and expected events of (1) representations to ED within 72 h, (2) unplanned readmissions to hospital within 28 days, and (3) all-cause mortality within 30 days. ResultsThere were 189,990 injury-related hospitalisations of children and young people. Of these, 4.4% represented to an ED, 8.7% were readmitted to hospital, and 0.2% died. Of the 45 public hospitals that treated ≥100 cases per year, higher than expected rates of ED representations, hospital readmissions, and mortality were observed in eleven, six, and two hospitals, respectively. ConclusionThe rates of ED representations, hospital readmissions, and mortality among children and young people hospitalised for injury in NSW were similar to the rates reported in other countries. However, unwarranted clinical variation across public hospitals was observed for all three indicators. These findings suggest that by improving routine follow-up support services post-discharge for children and young people and their families, it may be possible to reduce unwarranted clinical variation and improve health outcomes.