PERFORMANCE OF SIX PREDICTIVE MODELS OF DEATH OF PATIENTS HOSPITALIZED FOR DECOMPENSATED CIRRHOSIS: A MULTICENTER...

Area of Science:

• Hepatology
• Internal Medicine
• Medical Statistics

Background:

• Cirrhosis progresses from compensated to decompensated stages, impacting patient survival.
• Accurate prediction of mortality is vital for clinical decisions, including liver transplantation.
• Several prognostic models exist, but their comparative performance needs evaluation.

Purpose of the Study:

• To compare the predictive performance of six mortality scoring systems for hospitalized patients with decompensated cirrhosis.
• To assess the accuracy of models like MELD 3.0 and MELD-Plus in predicting 90-day mortality.

Main Methods:

• A retrospective, multicenter study included 481 patients with decompensated cirrhosis.
• Six prognostic models (MELD-Na, MELD-Plus, MELD 3.0, ReMELD, Refit MELD, Refit MELD-Na) were calculated on admission.
• Model accuracy was evaluated using the area under the receiver operating characteristic (AUROC) curve and compared using the DeLong test.

Main Results:

• All evaluated scores demonstrated moderate discrimination ability, with AUROC values ranging from 0.703 to 0.758.
• No single scoring system demonstrated unequivocal superiority over the others in predicting 90-day mortality.
• The models showed a good ability to discriminate between survivors and deceased patients.

Conclusions:

• The evaluated scoring systems have limited predictive power for 90-day mortality in decompensated cirrhosis.
• This study could not establish a definitively superior prognostic scoring system for this patient population.
• Further research may be needed to refine or identify more accurate predictive models.