The effectiveness of risk stratification systems in diagnosis of nodular thyroid disorders

Abstract

Risk stratification systems, one of the optimal solutions for differential diagnosis of nodular pituitary disorders, are currently under development. The high prevalence of pituitary diseases makes it necessary to evaluate the effectiveness of risk stratification systems and to widely implement them into routine clinical practice. To evaluate the effectiveness of modern risk stratification systems used to diagnose nodular pituitary disorders. A total of 1,606 medical records of patients operated on for nodular pituitary disorders in 20062014 were analyzed. The preoperative ultrasonography results and cytological findings were evaluated. The ultrasonography results were classified using the TI-RADS system, while the biopsy data were classified using the TBSRTC system. The surgery protocol and the pathomorphological data were the truth criterion. The effectiveness of the TI-RADS and TBSRTC systems, as well as their contribution to the performance of endocrinologists in outpatient clinics, was analyzed. Cluster analysis revealed a significant volatility of ultrasonography signs in the TI-RADS category, while there was no dominant sign that could be considered the diagnostic standard. Factor analysis proved the consistency of the imaging TI-RADS system based on individual signs. The signs being evaluated are characterized by high significance level but different priorities depending on the type of putative pathology. Discriminant analysis revealed that TI-RADS was a robust and versatile system to be used for various types of nodular pituitary disorders. The overall effectiveness of the TI-RADS system in diagnosis of pituitary tumors was low: it was characterized by 75.4% sensitivity, 84.7% specificity, and 80.1% accuracy. However, this system concretized the indications for fine-needle aspiration biopsy and drew the cytologists attention to the likelihood of pituitary tumor. The implementation of the TBSRTC system reduced the percentage of non-informative (by 9.8%) and controversial results (by 1.7%). Cytological examination was more effective in detection of pituitary cancer compared to ultrasonography (91.0% accuracy, 94.9% specificity, and 76.5% sensitivity). The impressions of ultrasound technicians and cytologists were concordant in 873 (54.4%) cases. A survey conducted among endocrinologists in outpatient clinics showed that implementation of the TI-RADS and TBSRTC risk stratification systems reduced the decision time (p0.001) and errors both in diagnostics (p0.001) and treatment approach selection (p0.001). The combined use of the TI-RADS and TBSRTC systems allows one to personalize the treatment approaches for patients with nodular pituitary disorders. The implementation of these systems has a positive effect on endocrinologists performance as it reduces the decision time and the likelihood of making errors in diagnosis and treatment strategy selection.