External validation of the hospital frailty risk score for predicting clinical outcomes in patients hospitalised with community acquired pneumonia in two Australian hospitals

Abstract Background Community-acquired pneumonia (CAP) is common globally, disproportionately affecting older, frailer patients. Frailty is linked to poorer clinical outcomes. Therefore, improved risk stratification is essential for optimal care. The Hospital Frailty Risk Score (HFRS) assesses frailty using administrative data, but its performance in evaluating clinical outcomes among hospitalised Australian CAP patients has not been assessed. Aims To validate the HFRS in Australian CAP patients. Methods This retrospective observational study examined all adult emergency admissions for CAP between 01/01/2018 and 31/12/2023 at two Australian hospitals. Data were captured from electronic medical records using the International Classification of Diseases. Patients were classified as non-frail (HFRS < 5) or frail (HFRS ≥ 5). Multilevel multivariable regression analysis determined whether the HFRS predicts clinical outcomes: length of stay (LOS), risk of intensive care unit (ICU) admission, in-hospital, 30-day, 365-day mortality, and 30-day readmission risk. Results There were 6,885 patients with a mean (SD) age of 75.4 (17.4) and 54.7% were male. The mean (SD) HFRS was 5.3 (4.8) with 41.7% classified as frail. Increasing HFRS values significantly predicted longer hospital LOS (IRR 1.07), greater likelihood of ICU admission (aOR 1.12), higher in-hospital mortality (aOR 1.06), higher 30-day mortality (aOR 1.05), and higher 365-day mortality (aOR 1.02; p ≤ 0.001 for all outcomes). By contrast, the HFRS was not significantly associated with 30-day readmission risk. Conclusion The HFRS is a practical, low-cost tool which could have far-reaching implications for streamlining clinical care worldwide. The evidence for its use is becoming more robust with each subsequent validation study.