Related Experiment Videos
Risk Factors for Postoperative Hemorrhage Following Thyroid Surgery: Results of a Case-Control Study and Development
Constantin Smaxwil1, Ali Naddaf1, Mirjam Busch1
1Department of Endocrine Surgery, Endocrine Centre Stuttgart, Diakonie-Klinikum Stuttgart, 70176 Stuttgart, Germany.
Journal of Clinical Medicine
|July 28, 2026
Summary
A new risk score identifies patients at high risk for postoperative hemorrhage (POH) after thyroid surgery. This tool combines traditional statistics and machine learning to aid perioperative care and surgical planning.
Area of Science:
- Surgery
- Public Health
- Machine Learning
Background:
- Postoperative hemorrhage (POH) is a rare but serious complication of thyroid surgery.
- Early identification of high-risk patients is crucial for effective perioperative management.
- Increasing outpatient thyroid procedures necessitate improved risk assessment tools.
Purpose of the Study:
- To develop and validate a quantitative risk score for predicting POH after thyroidectomy.
- To identify key predictors of POH using both univariate and machine learning methods.
- To enable risk stratification for individualized patient care and resource allocation.
Main Methods:
- Retrospective, single-centre case-control study of 9158 thyroidectomies (2012-2019).
- 104 POH cases were matched 1:4 with controls based on age, sex, procedure type, and surgical year.
- Univariate analysis and Random Forest machine learning were used to identify risk factors and develop a scoring system.
Main Results:
- Key predictors of POH included reoperative surgery, smoking, comorbidities, hyperthyroid treatment, and advanced age (3 points).
- Moderate predictors included alcohol consumption, Graves' disease, hyperthyroid state, procedure duration, and thyroid weight (1 point).
- The risk score differentiated between patients with and without POH, forming the basis for a 'traffic light' risk stratification model.
Conclusions:
- A simple, interpretable point-based scoring system effectively stratified POH risk in the study population.
- Combining statistical and machine learning approaches enhances the score's predictive capability.
- The exploratory score requires external validation for generalizability and clinical utility in predicting postoperative hemorrhage after thyroid surgery.