Aim
To determine the incidence and predictors of 28 day unplanned hospital readmission (28D-UHR) in patients with delirium during the index hospital admission
Methods
A retrospective longitudinal cohort study was conducted over a 12-month period (July 2020 to June 2021) at a Melbourne hospital on all hospital presentations with delirium. Univariate and multivariate analysis was performed to assess predictors of 28D-UHR. 12-month mortality was the secondary outcome.
Results
9.8% (160 of 1634) of patients identified with delirium had a 28D-UHR. In the readmissions group, univariate analysis found a higher percentage of patients from residential aged care facilities (RACF), shorter length of stay (LOS), higher 12-month mortality and higher number of admission and discharge medications during the index admission when compared to the non-readmission group.
Patients discharged to RACF had shorter LOS compared to those discharged to the community (8 v 14 days, p <0.001).
Multivariate logistic regression model adjusting for these variables found a shorter LOS (odds ratio [OR] 0.98, p<0.001), higher number of discharge medications (OR 1.09 p<0.001) and residing in RACF pre-admission (OR 1.34, p = 0.028) to be associated with 28D-UHR.
Conclusions
Admission from RACF was associated with a higher risk of 28D-UHR while discharge to RACF was associated with shorter LOS during index admission. Further research is required into hospital based post-discharge interventions to support RACFs in managing patients discharged with delirium. Potential hospital pharmacy interventions to support de-prescribing prior to discharge should also be considered.