Endocrine Abstracts

Background: Sick euthyroid syndrome is a well-known condition and no additional treatment is often required. However various authors have postulated that there is an effect of abnormal thyroid function on the outcomes of hospitalized patients. In our study, we chose to study the effect of thyroid function abnormalities and their independent effect on mortality of critically ill patients who had no known prior thyroid diseases.

Methods: Data was obtained from ‘Medical Information Mart for Intensive Care III’ database between 2001 and 2012. All patients whose Thyroid Stimulating hormone (TSH) and Thyroxine (T4) were tested were considered. Patients who had prior thyroid diseases were removed. The rest were split into multiple groups based on their TSH and T4 levels during their ICU stay. The groups are outlined in Table 1. Base outcome group was those with normal TSH and normal T4 (Group Zero - G0). Multinomial logistic regression with length of stay (LOS) and Simplified Acute Physiology Score (SAPS) II as co-variables were run to assess the Relative Risk Ratios (RRR) of 30-day mortality with altered TSH and T4 levels.

Results: Total of 2970 adult ICU admissions who had no underlying thyroid disease were considered for the study. Among them, two groups, G1 (High TSH and High T4) and G8 (Low TSH and Low T4) when compared to the base outcome had a significantly increased RRR of mortality. The RRR for 30-day mortality, among G1 compared to G0 was 3.79 (P=0.021) and 3.67 (P=0.001) in G8 compared to G0.

Conclusion: Among the critically ill patients and specifically for those without prior thyroid illness, we postulate that T4 and TSH can be used as independent predictors of mortality. When both the values are altered, it might signify a more profound metabolic derangement, needing specific attention, as compared to the traditional subclinical hypo and hyperthyroidism.