Pilot implementation and evaluation of a social risk screening tool for hospitalised patients in Singapore.

Abstract

Routine in-hospital screening of social risk factors linked with poor health outcomes is not done in Singapore, where an aging population pays for 25% healthcare costs out-of-pocket despite universal coverage. We developed and implemented a social risk screening tool for hospitalised patients and reported the association between higher social risk and length of stay (LOS) and hospitalisation costs. The tool aimed to reveal variation in social risk among patients admitted to medical wards between October and December 2022. Sociodemographic data, medical co-morbidities, and frailty scores were recorded. Multivariable negative binomial regression and Poisson regression assessed the association between higher social risk (risk score above cohort median), and LOS and LOS > 21 days, respectively. A parsimonious multivariable generalised linear model (GLM) modeled the association with direct health costs. Among 320 patients (median age 75, interquartile range [IQR]: 65-83), 80% were Chinese and 65% female. A total of 33% had higher social risk (>2 social risk factors). They were more likely to be male (50% vs. 27%), frail (66% vs. 50%), and have ≥4 chronic conditions (54% vs. 32%). Higher social risk was associated with a 48% longer LOS (incidence rate ratio: 1.48, confidence interval [CI]: 1.20-1.81), a relative risk of 1.69 (CI: 0.95-3.00) for LOS > 21 days, and $3545 (range: $397-16,718) in added hospitalisation costs per admission. Patients admitted to our unit with higher social risk cost an additional $25 million (range: 2.8-119 million) in 2022. Prospective screening of social risk for hospitalised patients can help health systems identify patients with higher social needs and prioritise equitable resource allocation.