Abstract Introduction Frailty is common in hospitalised older adults. This study compared efficacy of a modified Hospital Frailty Risk Score (mHFRS) to standard HFRS and Clinical Frailty Scale (CFS) to determine whether mHFRS can be used to identify frail hospitalised patients. Methods Anonymised retrospective review of Electronic Health Records was undertaken in patients = > 65 years old attending the Emergency Department (ED) and admitted to hospital 1st July 2022 to 31st March 2023. mHFRS utilises 2 prior emergency admissions within 2 years to generate a frailty risk score, whereas HFRS requires an index admission plus 2 prior emergency admissions. Hospitalisation outcomes and predictive models were evaluated with correlation and measures of agreement between CFS and HFRS, CFS and mHFRS using Spearman’s rank correlation and Cohen’s kappa. Results Of 3042 patients, CFS categorised 1635 patients as non-frail (CFS 1–4) and 1407 as frail (CFS 5–9). Using mHFRS, only 1623 patients could be categorised and of these, 608 were deemed low, 657 intermediate and 358 high risk of frailty. Frail patients were older (81.8 years, SD 8.41 vs 75.3 years, SD 7.20, p = <.001), had significantly longer LOS (52.5% % vs 31.5%, p = <0.001), higher 30-day unplanned hospital readmissions (18.5% vs 9.9%, p = <0.001), and higher in-patient (6.1% vs 2.0%, p < 0.001), 30-day (9.1% vs 2.3%, p < 0.001), and 90-day (15.8% vs 5.1%, p < 0.001) mortality. mHFRS achieved comparable association with hospitalisation outcomes compared to CFS & HFRS. Cohens’s kappa, showed fair agreement across HFRS and mHFRS, κ of 0.235 0.243 respectively. mHFRS was less sensitive at identifying frail patients but had better specificity to identify non frail patients. Conclusion mHFRS is a comparable frailty screening tool that doesn’t require clinical assessment but is standardised and easy to use in those who can be scored.
Read full abstract