Regional variation in post‐operative mortality in New Zealand

• Published in: ANZ journal of surgery 2022-05, Vol.92 (5), p.1015-1025
• Main Authors: Gurney, Jason K., McLeod, Melissa, Stanley, James, Robson, Bridget, Campbell, Douglas, Dennett, Elizabeth, Ongley, Dick, Rumball‐Smith, Juliet, Sarfati, Diana, Koea, Jonathan
• Format: Article
• Language: English
• Subjects: Best practice; Comorbidity; Deprivation; disparity; Ethnic factors; Ethnicity; Humans; inequity; Mathematical analysis; Minority & ethnic groups; Mortality; New Zealand - epidemiology; Patients; perioperative mortality; Population studies; Postoperative Period; post‐operative mortality; regional variation; Regions; Regression analysis; Regression models; Surgical Outcomes and Audit; Variation
• ISSN: 1445-1433; 1445-2197; 1445-2197
• DOI: 10.1111/ans.17510

Background There is a growing body of evidence that access to best practice perioperative care varies within our population. In this study, we use national‐level data to begin to address gaps in our understanding of regional variation in post‐operative outcomes within New Zealand. Methods Using National Collections data, we examined all inpatient procedures in New Zealand public hospitals between 2005 and 2017 (859 171 acute, 2 276 986 elective/waiting list), and identified deaths within 30 days. We calculated crude and adjusted rates per 100 procedures for the 20 district health boards (DHBs), both for the total population and stratified by ethnicity (Māori/European). Odds ratios comparing the risk of post‐operative mortality between Māori and European patients were calculated using crude and adjusted Poisson regression models. Results We observed regional variations in post‐operative mortality outcomes. Māori, compared to European, patients experienced higher post‐operative mortality rates in several DHBs, with a trend to higher mortality in almost all DHBs. Regional variation in patterns of age, procedure, deprivation and comorbidity (in particular) largely drives regional variation in post‐operative mortality, although variation persists in some regions even after adjusting for these factors. Inequitable outcomes for Māori also persist in several regions despite adjustment for multiple factors, particularly in the elective setting. Conclusions The persistence of variation and ethnic disparities in spite of adjustment for confounding and mediating factors suggests that multiple regions require additional resource and support to improve outcomes. Efforts to reduce variation and improve outcomes for patients will require both central planning and monitoring, as well as region‐specific intervention. In this study, we used national‐level data to begin to address gaps in our understanding of regional variation in post‐operative outcomes within New Zealand. After adjusting for confounding, we found regional variation in post‐operative mortality and in equity of outcomes between Māori and European patients. Regional variation in patterns of age, deprivation and comorbidity (in particular) helps to drive regional variation in post‐operative mortality, although variation persists in some regions even after adjusting for these factors.