The Added Value of Serum Random Cortisol and Thyroid Function Tests as Mortality Predictors for Critically Ill Patients: A Prospective Cohort Study

Abstract

Background: Thyroid hormone and cortisol levels can change during a course of illness. Our study was conducted to assess the ability of the level of these hormones to predict mortality among intensive care unit (ICU) patients. The added predictive value of these hormones with APACHE II scores was also evaluated. Methods: Thyroid hormones and random cortisol levels in adult ICU patients were collected on admission. Multivariate logistic regression analysis was used to assess the relationship between hormone levels and mortality. The added value of the mortality predictive ability was determined by area under the receiver operating characteristic (AuROC). Results: A total of 189 patients were included in the study. Free T3 and serum random cortisol levels were statistically significantly related to ICU mortality with OR 0.51 (0.28, 0.97), p = 0.047 and OR 1.02 (1.01, 1.04), p < 0.002, respectively. Free T3 and serum random cortisol significantly enhanced the predictive performance of APACHE II scores with an AuROC of 0.656 (non-added value model) versus 0.729 (added value model), p = 0.009. The scoring system was created with a total score that ranged from 1 to 14. A score above 7.0 indicated a high mortality rate with a sensitivity of 81.5% and a specificity of 33%. Conclusions: Serum free T3 and cortisol levels are significantly associated with ICU mortality and can enhance the ability of APACHE II scores to predict ICU mortality.