Factors influencing early and late readmissions in Australian hospitalised patients and investigating role of admission nutrition status as a predictor of hospital readmissions: a cohort study

Abstract

Objectives: Limited studies have identified predictors of early and late hospital readmissions in Australian healthcare settings. Some of these predictors may be modifiable through targeted interventions. A recent study has identified malnutrition as a predictor of readmissions in older patients but this has not been verified in a larger population. This study investigated what predictors are associated with early and late readmissions and determined whether nutrition status during index hospitalisation can be used as a modifiable predictor of unplanned hospital readmissions.

Design: A retrospective cohort study.

Setting: Two tertiary-level hospitals in Australia.

Participants: All medical admissions ≥18 years over a period of 1 year.

Outcomes: Primary objective was to determine predictors of early (0-7 days) and late (8-180 days) readmissions. Secondary objective was to determine whether nutrition status as determined by malnutrition universal screening tool (MUST) can be used to predict readmissions.

Results: There were 11 750 (44.8%) readmissions within 6 months, with 2897 (11%) early and 8853 (33.8%) late readmissions. MUST was completed in 16.2% patients and prevalence of malnutrition during index admission was 31%. Malnourished patients had a higher risk of both early (OR 1.39, 95% CI 1.12 to 1.73) and late readmissions (OR 1.23, 95% CI 1.06 to 128). Weekend discharges were less likely to be associated with both early (OR 0.81, 95% CI 0.74 to 0.91) and late readmissions (OR 0.91, 95% CI 0.84 to 0.97). Indigenous Australians had a higher risk of early readmissions while those living alone had a higher risk of late readmissions. Patients ≥80 years had a lower risk of early readmissions while admission to intensive care unit was associated with a lower risk of late readmissions.

Conclusions: Malnutrition is a strong predictor of unplanned readmissions while weekend discharges are less likely to be associated with readmissions. Targeted nutrition intervention may prevent unplanned hospital readmissions.

Trial registration: ANZCTRN 12617001362381; Results.

Keywords: epidemiology; internal medicine; quality in health care.

Figure 1

Distribution of readmissions.

Figure 2

Cumulative incidence estimate model for readmissions with death as a competing risk. Competing risk regression was used to estimate subdistribution HR (SHR), 1.17 (95% CI 1.06 to 1.28). Model adjusted for covariates—age, sex, Charlson Comorbidity Index and length of hospital stay.