Serum Albumin Level as a Predictor of Outcome in Patients Admitted to Pediatric Intensive Care Units

Objective The aim of this study was to assess serum albumin level on admission to the pediatric intensive care unit (PICU) as a prognostic indicator. Methods A prospective study was conducted in Fayoum University Children's Hospital. The study subjects' demographics and clinical and laboratory data were recorded. Pediatric Risk of Mortality III (PRISM-III) score was calculated. Serum albumin level was assessed within 24 hours from admission. Outcomes included mortality, PICU and hospital stay, need and duration of mechanical ventilation, and inotrope use. Results The incidence of admission hypoalbuminemia was 26%. The study subjects had a significantly higher mortality rate than subjects with normal albumin levels (42.3% vs 17.6%, respectively, P = 0.011). Each unit of increase in serum albumin decreased the risk of mortality by 28.9% (odds ratio, 0.289; confidence interval, 0.136-0.615, P = 0.001). Serum albumin showed a fair discriminatory power (area under the curve, 0.738). At a cutoff point of <= 3.7 g/dL, albumin had a 79.2% sensitivity, 67.1% specificity, 43.2% positive predictive value, and 91.1% negative predictive value. Incorporation of serum albumin with PRISM-III score was more predictive of mortality than either predictors alone (area under the curve, 0.802). No significant difference was found between the 2 groups regarding either PICU and hospital stay as well as the need and duration of ventilation. Conclusions In PICUs, admission hypoalbuminemia is a good predictor of mortality. Further studies to confirm the value of adding serum albumin to PRISM-III score are recommended.