Main Article Content
Measures that assess how healthcare affects health outcomes make an important contribution to efforts to inform local efforts to improve care for patients in NSW.
Objectives and Approach
The Bureau of Health Information (BHI) in New South Wales, Australia produces independent reports and information about the performance of the healthcare system. Our reports on readmission and returns to acute care following hospitalisation from eight clinical conditions investigated variation across more than 70 NSW public hospitals, with information available for 2000 to 2018.
Readmissions included both returns to acute care from non-acute inpatient settings and readmissions to any public or private hospital within 30-days or 60-days following hospital discharge. Risk standardised readmission ratios (RSRR) were calculated following linkage of admitted patient and mortality data, to also account for competing risk of death. Hospital outliers were identified using funnel plots. A sensitivity analysis explored the effect of residence in an aged care facility.
In the most recent period, the unadjusted rate of readmission ranged from 10% for ischaemic stroke to 22% for congestive heart failure with rates remaining relatively stable over time for most of the condition. The proportion of readmissions that were returns to acute care from a non-acute hospitalisation ranged from 2% for chronic obstructive pulmonary disease (COPD) to 42% for hip fracture surgery. Across all eight conditions majority of hospitals had results lower or no different than expected. Number of hospitals with higher than expected readmissions ranged from none for total knee replacement to six for COPD. Taking into account residence in an aged care facility did not impact the results in a meaningful way.
The RSRR method compares a hospital’s readmissions given its case-mix with an average NSW hospital with the same case-mix. In NSW, this method is a screening tool, identifying areas of excellence and of potential improvement.
This work is licensed under a Creative Commons Attribution 4.0 International License.