Geographic remoteness-based differences in in-hospital mortality among people admitted to NSW public hospitals with heart failure, 2002–21: a retrospective observational cohort study

Abstract

Objective

To examine associations between remoteness of region of residence and in-hospital mortality for people admitted to hospital with heart failure in New South Wales during 2002–21.

Study design

Retrospective observational cohort study; analysis of New South Wales Admitted Patient Data Collection data.

Setting, participants

Adult (16 years or older) NSW residents admitted with heart failure to NSW public hospitals, 1 January 2002 – 30 September 2021. Only first admissions with heart failure during the study period were included.

Main outcome measures

In-hospital mortality, by remoteness of residence (Australian Statistical Geography Standard), adjusted for age (with respect to median), sex, socio-economic status (Index of Relative Socioeconomic Advantage and Disadvantage [IRSAD], with respect to median), other diagnoses, hospital length of stay, and calendar year of admission (by 4-year group).

Results

We included 154 853 admissions with heart failure; 99 687 people lived in metropolitan areas (64.4%), 41 953 in inner regional areas (27.1%), and 13 213 in outer regional/remote/very remote areas (8.5%). The median age at admission was 80.3 years (interquartile range [IQR], 71.2–86.8 years), and 78 591 patients were men (50.8%). The median IRSAD score was highest for people from metropolitan areas (metropolitan: 1000; IQR, 940–1064; inner regional: 934; IQR, 924–981; outer regional/remote/very remote areas: 930; IQR, 905–936). During 2002–21, 9621 people (6.2%) died in hospital; the proportion was 8.0% in 2002, 4.9% in 2021. In-hospital all-cause mortality was lower during 2018–21 than during 2002–2005 (adjusted odds ratio [aOR], 0.52; 95% confidence interval [CI], 0.49–0.56); the decline was similar for all three remoteness categories. Compared with people from metropolitan areas, the odds of in-hospital death during 2002–21 were higher for people from inner regional (aOR, 1.12; 95% CI, 1.07–1.17) or outer regional/remote/very remote areas (aOR, 1.35; 95% CI, 1.25–1.45).

Conclusion

In-hospital mortality during heart failure admissions to public hospitals declined across NSW during 2002–21. However, it was higher among people living in regional and remote areas than for people from metropolitan areas. The reasons for the difference in in-hospital mortality should be investigated.