Novel score predicts short-term mortality of patients with cirrhosis
Click Here to Manage Email Alerts
Researchers developed a model from the Karnofsky Performance Status score and accurately predicted mortality of patients with cirrhosis 3 months posthospitalization, according to published findings.
“This easy-to-use assessment is strongly and independently associated with an increased risk of mortality and could be adopted in practice to guide postdischarge early interventions, including the integrated provision of active and palliative management strategies,” Puneeta Tandon, MD, of the division of gastroenterology, University of Alberta, Edmonton, Alberta, Canada, and colleagues wrote.
The researchers — on behalf of the North American Consortium for the Study of End-Stage Liver Disease — prospectively enrolled and evaluated 954 middle-aged patients hospitalized at 16 different centers for cirrhosis between July 2010 and September 2015. Each patient was followed through hospitalization for at least 3 months post-discharge via phone calls, medical records evaluation and interviews.
“Hospitalization is a marker for a variety of poor outcomes including readmission and death. … we lack a practical prognostic model to identify those at highest risk for 3-month postdischarge mortality. Identifying this group would facilitate interventions that include more intensive follow-up and consideration of early liver transplant when available,” the researchers wrote.
The goal of this study was to determine and evaluate outcomes “within the discharging health system or elsewhere,” according to the researchers. Karnofsky Performance Status (KPS) score was categorized as low (score 10-40), intermediate (50-70) or high (80-100).
Among the patients (63% male), the median MELD score was of 17 (interquartile range 13-21). Low KPS performance status was observed in 17% of patients, intermediate performance in 51% and high performance in 32%. The mortality rates by KPS performance status group were 23% for low (36/150), 11% for intermediate (55/489) and 5% for high (15/306; P < .001).
Analysis showed low performance status was associated with older age, dialysis, hepatic encephalopathy, longer hospitalization and higher white blood cell count or MELD score at discharge.
The researchers used multiple variables on univariate analysis to determine which were predictors of 3-month mortality. Three variables remained significant on multivariate analysis (KPS, age and MELD score), which were then used to develop a prognostic model known as the KAM model.
This model showed better discrimination (area under the receiver operating characteristic curve [AUROC] = 0.74) compared with a model using MELD (AUROC = 0.62) or MELD and age (AUROC = 0.67) to predict 3-month mortality rates.
“Accurate prognostication is essential in cirrhosis,” the researchers wrote. “It guides the type and frequency of clinical care interventions we prescribe and informs our discussions with patients and families about anticipated outcomes,” the researchers wrote.
Disclosures: Tandon reports no relevant financial disclosures. Please see the study for a list of all other authors’ relevant financial disclosures.