Quality of life measures predict mortality in patients with cirrhosis and severe ascites.

Background Severe ascites is associated with both a poor health-related quality of life (HRQL) and a mortality in excess of that captured by current prognostic clinical scores. Aim To determine the association between HRQL and mortality in patients with severe ascites. Methods The HRQL data from previously published randomised controlled trials examining the efficacy of satavaptan in ascites were retrospectively evaluated. Results Of the 496 patients randomised who completed the SF-36, 405 patients had complete datasets and were included in the analysis (difficult-to-treat ascites, n = 164 or refractory ascites, n = 241). Overall, patients reported poor HRQL, in particular the physical component score (PCS) of SF-36. The physical component score (PCS) correlated with the mental component score (MCS) of SF-36 (Spearman rank correlation = 0.68) but not with markers of severity of liver disease. The PCS, but not the MCS, was significantly lower in patients who died (P = 0.01 and P = 0.84, respectively). After confounder-adjustment, the hazard ratio for a 10-point increase in the physical component score was 0.83 (95% CI; 0.72-0.97) for all-cause mortality and 0.84 (95% CI; 0.71-0.99) for cirrhosis-related deaths only, indicating that patients with better physical HRQL live longer on average. Conclusions Poor physical component score (PCS) of SF-36 is an independent predictor of 12-month mortality in patients with severe ascites independent of current prognostic clinical scores. It holds promise not only in prognostic modelling but also as an endpoint in the evaluation of therapies targeting ascites.