Validation of the Hospital Frailty Risk Score in older surgical patients: A population-based retrospective cohort study.

Abstract

There is a need for standardized and cost-effective identification of frailty risk. The objective was to validate the Hospital Frailty Risk Score which utilizes International Classification Diagnoses in a cohort of older surgical patients, assess the score as an independent risk factor for adverse outcomes and compare discrimination properties of the frailty risk score with other risk stratification scores. Data were analysed from all patients ≥65years undergoing primary surgical procedures from 2006-2018. Patients were categorized based on the frailty risk score. The primary outcomes were 30-day mortality and 180-day risk of readmission. Of 16793 patients evaluated, 7480 (45%), 7605 (45%) and 1708 (10%) had a low, intermediate and high risk of frailty. There was a higher incidence of 30-day mortality for individuals with intermediate (2.9%) and high (8.3%) compared with low (1.4%) risk of frailty (P<.001 for both comparisons). Similarly, the hazard of readmission within the first 180days was higher for intermediate (HR 1.25; 95% CI: 1.16-1.34) and high (HR 1.84; 95% CI: 1.66-2.03) compared with low (HR 1.00, P<.001 for both comparisons) risk of frailty. The hazard of long-term mortality was higher for intermediate (HR 1.70; 95% CI: 1.61-1.80) and high (HR 4.16; 95% CI: 3.84-4.49) compared with low (HR 1.00, P<.001 for both comparisons) risk of frailty. Finally, long length of primary hospitalization occurred for 9.3%, 15.0% and 27.3% of individuals with low, intermediate and high frailty risk (P<.001 for all comparisons). A model including age and ASA classification had the best discrimination for 30-day mortality (AUC 0.862; 95% CI: 0.847-0.877). Our findings suggest that the Hospital Frailty Risk Score might be used to screen older surgical patients for risk of frailty. While only slightly improving prediction of 30-day mortality using the ASA classification, the Hospital Frailty Risk Score can be used to independently classify older patients for the risk of important outcomes using pre-existing readily available electronic data.