Data-Driven Strategies for Carbimazole Titration: Exploring Machine Learning Solutions in Hyperthyroidism Control
The Journal of Clinical Endocrinology & Metabolism · 2025
DOI: 10.1210/clinem/dgae642 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of Clinical Endocrinology & Metabolism, 2025. Verfasst von Thilo Reich und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computing & Informatics, Bournemouth University, Talbot Campus, Poole BH12 5BB, UK. Hauptthemen laut MEDLINE: Machine Learning, Carbimazole, Hyperthyroidism, Antithyroid Agents. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 38 Literaturangaben. Gefördert von: Bournemouth University Higher Education Innovation Funding 6 (2017-2022).
Abstract (Originalsprache)
Background
University Hospitals Dorset (UHD) has more than 1000 thyroid patient contacts annually. These are primarily patients with autoimmune hyperthyroidism treated with carbimazole titration. Dose adjustments are made by a healthcare professional (HCP) based on the results of thyroid function tests, who then prescribes a dose and communicates this to the patient via letter. This is time consuming and introduces treatment delays. This study aimed to replace some time-intensive manual dose adjustments with a machine learning model to determine carbimazole dosing. This can in the future serve patients with rapid and safe dose determination and ease the pressures on HCPs.
Methods
Data from 421 hyperthyroidism patients at UHD were extracted and anonymized. A total of 353 patients (83.85%) were included in the study. Different machine learning classification algorithms were tested under several data processing regimes. Using an iterative approach, consisting of an initial model selection followed by a feature selection method, the performance was improved. Models were evaluated using weighted F1 scores and Brier scores to select the best model with the highest confidence.
Results
The best performance is achieved using a random forest (RF) approach, resulting in good average F1 scores of 0.731. A model was selected based on a balanced assessment considering the accuracy of the prediction (F1 = 0.751) and the confidence of the model (Brier score = 0.38).
Conclusion
To simulate a use-case, the accumulation of the prediction error over time was assessed. It was determined that an improvement in accuracy is expected if this model was to be deployed in practice.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Computing & Informatics, Bournemouth University, Talbot Campus, Poole BH12 5BB, UK.
- Innovation Department, Poole Hospital, Poole BH15 2JB, UK.
- Department of Medical Science & Public Health, Bournemouth University, Poole BH12 5BB, UK.
- Bournemouth Diabetes and Endocrine Centre, University Hospitals Dorset, Bournemouth BH7 7DW, UK.
ORCID-Profile: Thilo Reich, Rashid Bakirov, Marcin Budka
Förderung
- Bournemouth University Higher Education Innovation Funding 6 (2017-2022)
Zitieren
APA
Reich, T., Bakirov, R., Budka, D., Kelly, D., Smith, J., Richardson, T., & Budka, M. (2025). Data-Driven Strategies for Carbimazole Titration: Exploring Machine Learning Solutions in Hyperthyroidism Control. The Journal of Clinical Endocrinology & Metabolism. https://doi.org/10.1210/clinem/dgae642
Vancouver
Reich T, Bakirov R, Budka D, Kelly D, Smith J, Richardson T, et al. Data-Driven Strategies for Carbimazole Titration: Exploring Machine Learning Solutions in Hyperthyroidism Control. The Journal of Clinical Endocrinology & Metabolism. 2025. doi:10.1210/clinem/dgae642
BibTeX
@article{Reich2025,
author = {Reich, Thilo and Bakirov, Rashid and Budka, Dominika and Kelly, Derek and Smith, James and Richardson, Tristan and Budka, Marcin},
title = {Data-Driven Strategies for Carbimazole Titration: Exploring Machine Learning Solutions in Hyperthyroidism Control},
journal = {The Journal of Clinical Endocrinology & Metabolism},
year = {2025},
doi = {10.1210/clinem/dgae642}
}
EUVIMED