Developing a nomogram for risk prediction of the low T3 syndrome

Abstract

Abstract The objective of this study is to establish a nomogram for predicting low triiodothyronine syndrome (LTS), and clarify the risk factors of LTS. In this study, 109 LTS patients and 331 non-LTS patients were included as the subjects. The general biochemical information of the subjects was collected, and the relevant factors of LTS were evaluated. Concordance index (C-index), Calibration plot, Hosmer-Lemeshow test (H-L test), and Receiver Operating Characteristic (ROC) curve were used to assess the nomogram model. The results of the multiple logistic regression analysis showed that the chronic heart failure (CHF) (OR: 1.978, 95% CI: 1.006–3.892), interleukin-6 (IL-6) (OR: 2.018, 95%CI: 1.043–3.901), infection (OR: 2.136, 95%CI: 1.154–3.955), C-reactive protein (CRP) (OR: 2.720, 95%CI: 1.311–5.641), and N-terminal pro-B-type natriuretic peptide (NT-proBNP) (OR: 3.121, 95%CI: 1.691–5.759) might be the relevant independent risk factors for LTS (P < 0.05), while serum albumin (Alb) might be a protective factor (P < 0.05). All the significant predictors were combined into a predictive nomogram, and the obtained C-index was 0.867 (95% CI: 0.824–0.910). The calibration plot showed good performance of the nomogram, and the predictive model passed H-L test (χ 2 = 8.431, P = 0.392). The ROC curve of the model analysis revealed a high sensitivity and specificity. Infection, CHF, serum CRP, NT-proBNP, and IL-6 were identified as independent risk factors for LTS, while serum Alb was a protective factor for LTS. This nomogram predicting LTS possessed good discrimination and accuracy, which could provide scientific guidance for individualized prevention in clinical.

Keywords

• Triiodothyronine syndrome
• Risk factors
• Prediction model
• Nomogram