Serum albumin and the short-term mortality in individuals with congestive heart failure in intensive care unit: an analysis of MIMIC

Decreased albumin levels are common in congestive heart failure (CHF), but little is known about its role in mortality risk in CHF. This study developed a cohort prediction model based on 7121 individuals with heart failure to evaluate the short-term mortality and prognostic role of albumin in patients with CHF. The cohort was from intensive care unit between 2001 and 2012 in a publicly available clinical database in intensive care called MIMIC III. We used a generalized additive model to determine the nonlinear correlation between serum albumin and 14th day, 28th day and 90th day all-cause mortality in patients with heart failure. The results showed that serum albumin is an independent risk factor for 14th, 28th and 90th day all-cause mortality, and has a linear relationship with all-cause mortality in congestive heart failure. Cox regression analysis using restricted cubic spline with albumin as continuous parameter showed that the decrease of albumin level is directly related to the increase of mortality (14th day mortality: hazard ratio [HR], 0.65 [95% CI, 0.58 to 0.73]); 28th day mortality: HR, 0.56 [95% CI, 0.51 to 0.63]; 90th day mortality: HR, 0.52 [95% CI, 0.47 to 0.57]; P for trend < 0.001). The multivariate adjusted association between albumin (as a continuous variable) and all-cause mortality on the 90th days is mixed by ARDS [HR, 0.64, 95% CI (0.47-0.87), P = 0.005]. The all-cause mortality on the 90th day predicted better clinical results with the all-cause mortality on the 14th day.