Comparative prognostic contributions of chronological age and comorbidity burden to mortality after geriatric hip fracture

Chronological age and comorbidity burden are established prognostic factors after geriatric hip fracture, but their relative contributions are difficult to compare directly. Age-adjusted comorbidity indices incorporate chronological age within the score itself, and the two predictors are measured on different scales, limiting comparison of per-unit hazard ratios. This study assessed the independent and relative prognostic contributions of chronological age and comorbidity burden, quantified using the non-age-adjusted Charlson Comorbidity Index (CCI), to mortality after geriatric hip fracture. This retrospective cohort study included 648 patients aged ≥ 65 years who underwent hemiarthroplasty for a fragility hip fracture between January 2017 and December 2020. Survival status was verified through the National Death Notification System. Age and CCI were modeled as continuous variables; proportional hazards assumptions and functional form were assessed using covariate × log(time + 1) interactions and centered quadratic terms, respectively. Clinically comparable effects were estimated across the interquartile range (IQR) of each predictor, and incremental contributions to model fit were compared using hierarchical Cox models with equal-complexity two-degree-of-freedom blocks. The median age was 82 years (IQR 76–87) and the median CCI was 2 (IQR 1–3). During a median follow-up of 27 months (IQR 5–52), 477 patients (73.6%) died; 30-day mortality was 10.3% and cumulative one-year mortality was 35.0%. Median survival was 44.0 months (95% CI 37.8–50.2) for CCI 0–2 versus 6.0 months (95% CI 3.5–8.5) for CCI ≥ 3 (log-rank p < 0.001). An increase in age across its IQR (76 to 87 years) was associated with higher mortality (HR 1.532; 95% CI 1.329–1.766; p < 0.001), whereas an increase in CCI across its IQR (1 to 3 points) was associated with a greater increase in hazard (HR 3.436; 95% CI 2.741–4.307; p < 0.001). Both predictors improved model fit when added to a model already containing the other (age: LR χ²=36.314; CCI: LR χ²=159.916; both df = 2, p < 0.001). Chronological age and comorbidity burden provide independent and complementary prognostic information after geriatric hip fracture, with CCI showing a stronger association with mortality and a greater incremental contribution to model fit. Risk stratification should incorporate objective comorbidity assessment alongside, rather than instead of, chronological age. Not applicable.