Electronic Synoptic Reporting of Thyroid Nodules: Potential for Reduction in Number of Patients Undergoing Thyroid Nodule Biopsies

Abstract

Purpose: The objective of the study was to design and implement an electronic synoptic report for thyroid sonography that incorporates the thyroid imaging reporting and data system (TIRADS) and assess potential for reducing unnecessary fine needle aspiration biopsies (FNAB) of thyroid nodules. Methods: The electronic synoptic report was developed using a relational database based on elements from TIRADS and a multidisciplinary consensus statement for thyroid reporting. A retrospective analysis of 138 patients with previously reported thyroid sonographic exams was evaluated for the presence of these elements. The electronic synoptic report calculates the TIRADS score and generates a formal report. Using the TIRADS score the potential decrease in unnecessary FNAB was estimated. Results: Key TIRADS elements were variously reported ranging from 43% for the thyroid nodule's architecture as solid or cystic. Thyroid nodule echogenicity and calcification was commented in 27% and 23%, respectively. Other features of the TIRADS score were commented in 0% to 8% of the official reports. Estimated reduction for potentially reduced need for FNAB was 34.5%. Conclusions: This study is the first implementation of synoptic reporting using a relational database for sonography of thyroid nodules. Implementation of an electronic standardized synoptic reporting system may facilitate more accurate, and more comprehensive reporting for thyroid ultrasound scanning of thyroid nodules. The use of TIRADS was estimated to be able to potentially reduce the need for FNAB which was significant.

Ambiguous and Incomplete: Natural Language Processing Reveals Problematic Reporting Styles in Thyroid Ultrasound Reports

“…Prior studies have demonstrated a lack of completeness in thyroid US reporting. 18 28 29 30 Like Gamme et al, we found that nodule size parameters were most likely to be included, but that other nodule characteristics were not consistently included. In addition, our findings are similar to the results of others demonstrating that only 31 to 43% of US reports incorporated nodule contents and that inclusion of borders, echogenicity, and calcifications was even lower.…”

“…28 32 33 Many institutions have adopted TI-RADS reporting of thyroid nodules. Although TI-RADS reporting may reduce unnecessary biopsies, 29 inclusion of the TI-RADS score alone to describe a nodule without cataloging nodule characteristics may limit NLP utility. Future work will be critical in determining the efficacy of NLP tools in annotating structured or synoptic thyroid US reports with detailed TI-RADS features and thus the applicability of NLP tools in thyroid US reports for clinical decision-making or research purposes.…”

“…3 Adoption of TI-RADS after our study period is likely to have resulted in more structured US reporting and reduced unnecessary FNA biopsies at some institutions. 29 34 However, inclusion of the TI-RADS score alone without the TI-RADS lexicon for each nodule scoring element likely limits the function of NLP systems like cTAKES and necessitates additional logic with regular expressions as was used in this study. If the US reports used in this study included clinical risk assessment or a summative score as well as a synoptic reporting system that incorporated elements of nodule features, these features likely would have improved capture of our target information and reduced the amount of missing data.…”

“…In addition, our findings are similar to the results of others demonstrating that only 31 to 43% of US reports incorporated nodule contents and that inclusion of borders, echogenicity, and calcifications was even lower. 28,29 Of note, the academic medical center demonstrated a slightly lower overall rate of completeness in reporting thyroid US parameters compared with the regional health care system. These findings, which are in contrast to the work of Inman et al, may be due to the tendency of radiologists at the academic medical center in our study to include recommendations for clinical intervention.…”

Online TI-RADS Calculator

“…Selected ultrasound features of thyroid nodules are combined into a score to identify nodules that warrant biopsy or sonographic follow-up. The use of TI-RADS to risk stratify incidental nodules may result in fewer unnecessary biopsies [3]. In order to facilitate the use of TI-RADS, an online calculator was developed.…”