Abstract [from journal]
Background: With a rising burden of cirrhosis surgeries, understanding risk factors for post-operative mortality are more salient than ever. The role of baseline frailty has not been assessed in this context. We aimed to evaluate the association between patient frailty and post-operative risk among diverse patients with cirrhosis, and to determine if frailty improves prognostication of cirrhosis surgical risk scores.
Methods: This was a retrospective cohort study of U.S. Veterans with cirrhosis identified between 2008 and 2016 who underwent non-transplant major surgery. Frailty was ascertained using the Hospital Frailty Risk Score (HFRS). Cox regression analysis was used to investigate the impact of patient frailty on post-operative mortality. Logistic regression was used to identify incremental changes in discrimination for post-operative mortality when frailty was added to risk prediction models including the MELD, MELD-Na, Child-Turcotte-Pugh (CTP), Mayo, and VOCAL-Penn scores.
Results: A total 804 cirrhosis surgeries were identified. The majority of patients (48.5%) had high-risk frailty at baseline (HFRS >15). In adjusted Cox regression models, increasing frailty category was associated with poorer post-operative survival. For example, intermediate-risk frailty (HFRS 5-15) conferred a 1.77-fold increased hazard relative to low-risk frailty (HFRS <5, 95% confidence interval [CI] 1.06-2.95, p=0.029). High-risk frailty demonstrated a similarly increased hazard (hazard ratio 1.74, 95% CI 1.05-2.88, p=0.032), suggesting a threshold effect of frailty on post-operative mortality. Incorporation of frailty improved discrimination of MELD, MELD-Na, and CTP for post-operative mortality, but did not do so for Mayo risk or VOCAL-Penn scores.
Conclusion: Patient frailty as an additional important predictor of cirrhosis surgical risk. Incorporation of pre-operative frailty assessments may help to risk-stratify patients, especially in settings where MELD-Na and CTP are commonly applied.