ABSTRACT
BACKGROUND: Thyroid hormones (THs) and thyroid-stimulating hormone (TSH) seem to show high potential in predicting the clinical death outcome of patients admitted to the intensive care unit (ICU). However, diverse studies on this topic are conflicting. METHODS: A search was conducted by two investigators involved in this research in the PubMed, Embase, and Cochrane databases (all last launched on July 12, 2021). The quality of the included studies was evaluated using the Newcastle-Ottawa Quality Assessment Scale (NOS). Subgroup analyses were performed to determine the sources of heterogeneity. Sensitivity and publication bias analyses were also assessed. RESULTS: A total of 27 studies (4970 participants) were included based on the eligibility criteria. Compared with survivors, nonsurvivors were found to have lower levels of THs (T3, T4, fT3, and fT4), whereas no significant difference was found in TSH levels (13 studies for T3: standardized mean differences [SMD], -0.78; 95% CI, -1.36 to -0.20; I2 = 96%; p = 0.008; 11 studies for T4: SMD = -0.79; 95% CI, -1.31 to -0.28; I2 =95%; p = 0.0002; 14 studies for fT3: SMD = -0.76; 95% CI, -1.21 to -0.32; I2 = 95%; p = 0.0008; 17 studies for fT4: SMD = -0.60; 95% CI, -0.99 to -0.22; I2 = 95%; p = 0.002; 20 studies for TSH: SMD = 0.00; 93% CI, -0.29 to 0.29; I2 = 93%; p = 0.98). CONCLUSION: Nonsurvivors were associated with lower levels of THs (T3, T4, fT3, and fT4) than survivors. THs show great application potential in predicting ICU patients' death outcomes and improving already widely used prognostic scores in the ICU (ie, Acute Physiological and Chronic Health Evaluation [APACHE] II and Therapeutic Intervention Scoring System).
