Predictive Value of Comprehensive Geriatric Assessment Scores for Mortality in Patients With Hip Fracture: A Retrospective Cohort Study

Objective To assess the predictive value of three scoring systems, namely the American Society of Anesthesiologists (ASA) classification, the Clinical Frailty Scale (CFS), and the Nottingham Hip Fracture Score (NHFS), in predicting mortality among patients with hip fractures. Materials and methods This retrospective cohort study included 628 participants aged 60 years and above who sought treatment at a UK hospital between January 2018 and December 2018. Data on age, gender, mortality, and assessment scores were collected. The area under the curve was calculated for each receiver operator characteristic (ROC). Cross-tabulation was performed to examine the association between various assessment scores and mortality using the chi-square test. Results The mean age was 80.80±11.18 years. Females were 408 (64.97%). Higher CFS (p<0.001) and NHFS (p<0.001) scores were significantly associated with mortality, while the ASA score did not show a significant association (p=0.225). The calculated area under the curve (AUC) values were as follows: 0.71 (95% CI: 0.65 to 0.76) for CFS, 0.46 (95% CI: 0.39 to 0.53) for NHFS, and 0.41 (95% CI: 0.34 to 0.48) for the ASA score. Utilizing a cut-off of ≥6 for CFS, 57 individuals (98.3%) in the 30-day mortality group were correctly identified. Similarly, the ROC analysis determined a ≥5 cut-off for NHFS accurately predicting 50 patients (86.2%) who deceased within 30 days. Applying an ASA ≥3 cut-off resulted in a predictive mortality rate of 56 (96.6%). The NHFS score demonstrated the highest predictive capability for mortality, with patients scoring ≥5 having a significantly higher risk of mortality compared to those with a score <5. Conclusion This study showed robust correlations between high CFS (≥6) and NHFS (≥5), and mortality within the hip fracture patient cohort.