Abstract
Purpose
To study 6-months mortality predictors upon inpatient geriatric rehabilitation admission.
Methods
A single-center retrospective study of all patients admitted for inpatient geriatric rehabilitation during 2018–2020. Logistic regression analysis and Chi-squared automatic interaction detection (CHAID) decision tree were used to study which variables were independently associated with 6-months mortality upon rehabilitation admission including: age, gender, chronic co-morbidities, prior functional status, and rehabilitation indication.
Results
The final cohort included 993 patients: 588 (59.2%) females, median age 82.0 years (IQR 75.0–87.0 years). Overall, 101 (10.2%) patients died within six months of rehabilitation admission. Logistic regression analysis showed that mortality was independently associated with male gender (OR 1.9, 95%CI 1.2–3.1), symptomatic severe aortic stenosis (OR 3.6, 95%CI 1.4–9.6), chronic kidney disease (OR 2.0, 95%CI 1.3–3.3), active cancer (OR 3.3, 95%CI 1.8–6.1), prior mild dependency (OR 5.3, 95%CI 2.3–12.4), prior severe dependency (OR 5.8, 95%CI 2.4–14.0), and deconditioning indicating rehabilitation (OR 2.2, 95%CI 1.2–3.9). CHAID decision tree showed that chronic kidney disease was the first splitting variable (p < 0.001) followed by male gender (p < 0.001) and age > 81 years (p = 0.027).
Conclusion
Symptomatic severe aortic stenosis and chronic kidney disease are presented herein, for the first time, as 6-months mortality predictors upon geriatric rehabilitation admission.