Age does not improve the predictive ability of the Hospital Frailty Risk Score for length of stay.

Abstract

The Hospital Frailty Risk Score (HFRS) has been widely used to identify patients at high risk of poor outcomes and to predict poor outcomes for older people. Although poor health outcomes are associated more with frailty than age, HFRS has been validated only for older people. This study aimed to explore for the first time whether age influences the predictive power of Hospital Frailty Risk Score to predict a long length of stay. A retrospective cohort study analysing data collected at Queen Alexandra Hospital in Portsmouth, UK, between January 1, 2010 and December 31, 2019. Data included people aged ≥16 years. We assessed the correlation between the HFRS and age using Pearson's correlation coefficient. We used logistic regression models to develop prediction models (HFRS alone and HFRS +age) for nine periods of length of stay, in nine age groups data to assess association and influences age on Hospital Frailty Risk Score. The correlation between Hospital Frailty Risk Score and age was weak in eight age groups, correlation coefficient ranged from 0.01 to 0.17. In each age groups, the proportion of the intermediate and high risk of frailty increased with a longer length of stay. Adjusted models (HFRS+age) did not show a better discriminative power compared with HFRS alone in all age groups (AUROCs ranging from 0.772 to 0.932). The odds ratio values from HFRS alone models were higher than adjusted models. This study concluded that in this patients' population, age does not improve the power of Hospital Frailty Risk Score to predict long length of stay.