A predictive nomogram incorporating gait speed for all-cause mortality in hospitalized cirrhotics

Abstract

ABSTRACT Objectives: No tailored model incorporating physical frailty for 2-year mortality in cirrhosis is available for practitioners in general practice. Thus we aimed to develop a model based on laboratory results and physical frailty allowing clinicians for stratifying cirrhotics by using individual estimate. Methods: One hundred and thirteen cases were assigned to the primary cohort, and all other 76 patients were regarded as the validation cohort. Multivariate Cox regression was performed, and a nomogram including five-meter gait speed (5MGS) were generated. The performance of the proposed model was assessed by C-index, calibration curve, and decision curve analysis (DCA). Results: On multivariate analysis, the Model for End-Stage Liver Disease-Sodium, albumin and 5MGS were independent predictors for 2-year mortality in cirrhosis. A nomogram incorporating all these parameters achieved a C-index of 0.804 (95%CI, 0.731–0.877). The calibration curve implied optimal correspondence between the predicted survival and actual outcomes. Our model is useful in the clinical settings based on DCA. Similar results were observed in the validation cohort with a C-index of 0.796 (95%CI, 0.689–0.899). Moreover, 5MGS, as a surrogate of physical performance, significantly correlated with multiple domains of general frailty according to Frailty Index (our published data), including instrumental activities of daily living, self-reported health, social activity and falls. Conclusion: In conclusion, the nomogram incorporating 5MGS may represent an individualized tool for predicting mortality in cirrhosis for primary care physicians.