Abstract 9926: A Controlled Study of Frailty and Clinical Outcomes in Acute Care Patients with Atrial Fibrillation

Abstract

Background: Frailty has multidimensional aetiology. AF may predict frailty and is associated with hospital mortality in epidemiological studies. In geriatric patients receiving acute hospital care, interrelationship between AF, frailty, a predisposition to subacute care and overall mortality is unclear. Aim: To investigate an association between frailty and in-hospital mortality in acute care patients with AF Methods: A retrospective study was conducted to investigate acute geriatric inpatients (>65 years, n=1518) with AF admitted under General Medicine units at three Monash Health affiliated hospitals between July and November 2020. Exclusion criteria were: non-English speaking background, MMSE<25, high-level care requiring residents. Frailty was measured by Rockwood clinical frailty scale. Primary endpoint was all-cause mortality. Results: Prevalence of AF was 19%. Of 294 with AF, 6% were new-onset. Mean age was 82±7 years and 43% were female. Mean left ventricular ejection fraction was 56±13 % and 30% had regional hypokinesis on transthoracic echocardiogram. Median Rockwood score was 5(IQR 3,6), acute hospital length of stay was 11±15days and 9% died during hospital stay. In comparison, Rockwood score was lower 4 (IQR 3,5) and only 2% died in a control group without AF (P=0.017). On univariate logistic regression, higher Rockwood score was associated with subacute admissions (OR=1.4,P=0.001) whereas, regional hypokinesis (OR=2.5,P=0.048), chronic kidney disease (OR=2.7,P=0.015) and higher Rockwood score (OR=2.9,P<0.0001) were associated with all-cause in-hospital mortality. After adjusting for age, premorbid status, and comorbidities, higher Rockwood score remained predictive of all-cause in-hospital mortality (OR=3.0,P<0.0001) in AF group. Conclusion: Frailty measured by Rockwood scale in acute care patients with AF is associated with in-hospital mortality. All-cause in-hospital mortality is significantly lower in a control group without AF. Prospective controlled studies are warranted to further strengthen the evidence.