BMC Geriatrics · Nov 18, 2026

Hospital Frailty Risk Score and treatment allocation, treatment intolerance, healthcare utilization, and survival in older adults with hepatocellular carcinoma in C​hina: a single-center retrospective cohort study

By Xu Y, Wen Y, Tan H et al.

Background Frailty may influence treatment decision-making and outcomes in older adults with hepatocellular carcinoma (HCC), but scalable measures are needed to support risk stratification and care planning. We aimed to determine whether an elevated Hospital Frailty Risk Score (HFRS) was associated with allocation to active tumor-directed therapy, treatment intolerance, unplanned healthcare utilization, and overall survival in older adults with HCC treated at a tertiary hospital in China. Methods We conducted a single-center retrospective cohort study of consecutive adults aged ≥ 65 years with HCC who had a treatment-planning encounter at Zhujiang Hospital, a tertiary hospital in Guangzhou, China, between January 2011 and December 2022. Frailty risk was measured using the Hospital Frailty Risk Score (HFRS) derived from diagnostic codes in the International Classification of Diseases, Tenth Revision (ICD-10) and dichotomized as low (< 5) vs. elevated (≥ 5). We evaluated four prespecified domains: (1) initial treatment allocation (active therapy vs. best supportive care), (2) treatment intolerance among treated patients, (3) unplanned healthcare utilization, and (4) overall survival (OS). Multivariable logistic regression assessed treatment allocation, cause-specific Cox models evaluated time to intolerance (with Fine–Gray competing-risk sensitivity analyses), negative binomial models evaluated unplanned admissions and inpatient bed-days, and Cox models evaluated OS. Incremental discrimination for OS was assessed by the change in Harrell’s C-statistic; additional sensitivity analyses examined treatment modality, systemic-regimen class, treatment era, comorbidity burden, body mass index (BMI), cirrhosis, and alpha-fetoprotein (AFP). Results Among 1,397 patients (median age 69 years; 29.1% women), 832 (59.6%) had elevated frailty risk. Compared with patients with low HFRS, those with elevated HFRS were less likely to receive active therapy (69.0% vs. 79.5%; adjusted odds ratio (aOR) 0.63, 95% confidence interval (CI) 0.49–0.81), had higher risk of treatment intolerance among treated patients (62.0% vs. 46.1%; adjusted hazard ratio (aHR) 1.71, 95% CI 1.42–2.07; Fine–Gray subdistribution hazard ratio (sHR) 1.63, 95% CI 1.35–1.98), and had greater unplanned admissions (adjusted incidence rate ratio (IRR) 1.47, 95% CI 1.29–1.67) and inpatient bed-days (adjusted IRR 1.41, 95% CI 1.23–1.63). Median OS was shorter in the elevated-HFRS group (21.3 vs. 35.4 months; aHR 1.54, 95% CI 1.32–1.80). Adding HFRS to the clinical survival model increased Harrell’s C-statistic from 0.731 to 0.742 (paired Δ = 0.011, 95% CI 0.004–0.019). Conclusions Elevated HFRS (≥ 5) was associated with less active treatment, earlier intolerance, greater unplanned healthcare utilization, and higher mortality in older adults with HCC. Adding HFRS to established clinical variables produced a statistically significant but modest increase in mortality discrimination. These findings do not establish that HFRS-guided care improves treatment decisions or outcomes; HFRS should be considered a candidate risk marker, and external validation and prospective implementation studies are required before clinical use. Clinical trial number Not applicable.

doi.org/10.1186/s12877-026-08630-0Biomarkers

AgingResearch.News