Discharge functional status and predictors of all-cause geriatric trauma readmission across a mature trauma network.

Abstract

Readmission after trauma remains a significant challenge in the geriatric population. Few studies have looked at geriatric trauma readmission (GTR) across a mature trauma network including at level IV centers. Our objective was to determine if discharge functional status predicts GTR across all levels of trauma centers. Secondary objective was to determine incidence and reason for GTR. Institutional trauma registries were queried for all geriatric trauma admissions across our network (2018-2023). These data were merged with all-cause network readmission data. Demographics, injury characteristics, trauma center level, frailty, discharge functional status (functional independence measure [FIM]), disposition, and payor status were compared between non-GTR and GTR patients. Univariate, followed by multivariate, logistic regression was used to identify predictors of readmission. Reason for GTR and time to GTR were examined. A total of 11,270 patients were admitted across the network with a median age of 81 years (interquartile range, 74-88 years) and a median Injury Severity Score of 5 (4-9), while 6.6% (n = 741) had an Injury Severity Score of >16. All-cause GTR rate was 6.2% (n = 700). On multivariate analysis, FIM score (odds ratio [OR], 0.99 [0.95-1.02]; p = 0.60), treatment at a level IV center, and disposition were not predictors of GTR. The results were similar after adjusting for frailty, with three or more comorbidities (OR, 1.708 [1.16-2.51]; p < 0.01) and hospital length of stay (OR, 1.05 [1.01-1.09]; p < 0.01) representing the highest predictors of GTR. Fourteen percent (n = 101/700) of GTR patients were readmitted for a trauma complication, 11% (n = 75) for new injury, 63% (n = 439) for a medical condition, and 12% (n = 84) because of prior refusal for rehab. The median time to GTR was 14 days (interquartile range, 6-21 days). Fifty-seven percent (n = 396) had incomplete follow-up. Hospital length of stay and more than three comorbidities but not FIM score predict GTR with most GTR for a new or preexisting medical condition. These variables represent suitable targets for GTR reduction. Clinical Research, Retrospective Study; Level IV.