Development of a Predictive Model and Nomogram for Thyroid Dysfunction Following Thermal Ablation of Papillary Thyroid Carcinoma

Background Thyroid dysfunction (TD) can occur after thermal ablation of papillary thyroid carcinoma (PTC), yet no dedicated clinical prediction tool has been available. To develop and internally validate a multivariable logistic regression model for predicting post-ablation TD. Methods In this single-center retrospective cohort, 295 patients with PTC who underwent ultrasound-guided thermal ablation and completed 12-month follow-up were randomly split 80:20 into training (n = 236) and validation (n = 59) sets. Nineteen candidate variables spanning demographics, serology, imaging, and procedural parameters were screened by univariable analysis; significant factors entered a multivariable logistic model via bidirectional stepwise selection (AIC). Discrimination, calibration, and clinical utility were assessed by AUC, calibration curves with the Hosmer–Lemeshow test, and decision curve analysis (DCA). The model was presented as a nomogram and a web-based calculator. Results Three independent predictors were retained: preoperative TSH (OR per mIU/mL, 1.91; 95% CI, 1.31–2.78), ablation time (OR per second, 1.01; 95% CI, 1.01–1.01), and nodule location (bilateral vs unilateral: OR, 3.85; 95% CI, 1.58–9.43; lobe+isthmus vs unilateral: OR, 7.95; 95% CI, 1.40–45.03). Overall, 83/295 (28.1%) patients developed TD within 12 months. Discrimination was good (AUC 0.72 training; 0.81 validation). Calibration was acceptable (Hosmer–Lemeshow P = 0.950 training; P = 0.251 validation). Across a wide range of threshold probabilities, DCA showed higher net benefit than treat-all or treat-none strategies. A nomogram and online calculator were derived for individualized risk estimation. Conclusion An internally validated multivariable model using preoperative TSH, ablation time, and ablation extent (bilateral or lobe+isthmus) provides individualized risk prediction of post-ablation TD and may support pre-procedure counseling, peri-procedural planning, and tailored follow-up. External, multicenter validation and prospective evaluation are warranted.