Evaluation of Length of Stay, Care Volume, In-Hospital Mortality, and Emergency Readmission Rate Associated With Use of Diagnosis-Related Groups for Internal Resource Allocation in Public Hospitals in Hong Kong

Abstract

Hong Kong's internal resource allocation system for public inpatient care changed from a global budget system to one based on diagnosis-related groups (DRGs) in 2009 and returned to a global budget system in 2012. Changes in patient and hospital outcomes associated with moving from a DRG-based system to a global budget system for inpatient care have rarely been evaluated. To examine associations between the introduction and discontinuation of DRGs and changes in length of stay, volume of care, in-hospital mortality rates, and emergency readmission rates in the inpatient population in acute care hospitals overall, stratified by age group, and across 5 medical conditions. This cross-sectional study included data from patients aged 45 years or older who were hospitalized in public acute care settings in Hong Kong before the introduction (April 2006 to March 2009), during implementation (April 2009 to March 2012), and after discontinuation (April 2012 to November 2014) of the DRG scheme. Data analysis was conducted from January to June 2021. Public hospitals transitioned from a global budget payment system to a DRG-based system in April 2009 and returned to a global budget system in April 2014. The main outcome was the association of use of DRGs with patient-level length of stay, in-hospital mortality rate, 1-month emergency readmission rate, and population-level number of admissions per month. An interrupted time series design was used to estimate changes in the level and slope of outcome variables after introduction and discontinuation of DRGs, accounting for pretrends. This study included 7 604 390 patient episodes. Overall, the mean (SD) age of patients was 68.97 (13.20) years, and 52.17% were male. The introduction of DRGs was associated with a 1.77% (95% CI, 1.23%-2.32%) decrease in the mean length of stay, a 2.90% (95% CI, 2.52%-3.28%) increase in the number of patients admitted, a 4.12% (95% CI, 1.89%-6.35%) reduction in in-hospital mortality, and a 2.37% (95% CI, 1.28%-3.46%) decrease in emergency readmissions. Discontinuation of the DRG scheme was associated with a 0.93% (95% CI, 0.42%-1.44%) increase in the mean length of stay and a 1.82% (95% CI, 1.47%-2.17%) reduction in the number of patients treated after adjusting for covariates; no statistically significant change was observed in in-hospital mortality (-0.14%; 95% CI, -2.29% to 2.01%) or emergency readmission rate (-0.29%; 95% CI, -1.30% to 0.71%). In this cross-sectional study, the introduction of DRGs was associated with shorter lengths of stay and increased hospital volume, and discontinuation was associated with longer lengths of stay and decreased hospital volume. In-hospital mortality and emergency readmission rates did not significantly change after discontinuation of DRGs.