Related Experiment Video
Updated: Jul 2, 2025
07:15
A Mouse Model of Incompletely Resected Soft Tissue Sarcoma for Testing Neoadjuvant Therapies
Published on: July 28, 2020
9.6K
Explainable Model Using Shapley Additive Explanations Approach on Wound Infection after Wide Soft Tissue Sarcoma
Ji-Hye Choi1,2, Yumin Choi3, Kwang-Sig Lee4
1Department of Orthopedic Surgery, Anam Hospital, Korea University College of Medicine, 73 Goryeodae-ro, Seongbuk-gu, Seoul 02841, Republic of Korea.
Medicina (Kaunas, Lithuania)
|February 24, 2024
Summary
Perioperative blood transfusions significantly increase the risk of wound infections in soft tissue sarcoma patients undergoing wide resection. Male sex, older age, and lower socioeconomic status also contribute to infection risk.
Area of Science:
- Oncology
- Surgical Infections
- Data Science in Healthcare
Background:
- Soft tissue sarcomas are rare but aggressive cancers posing challenges due to their nature and infection risk.
- Limited large-scale studies exist due to the low prevalence of these sarcomas.
- Perioperative wound infections are a significant concern for orthopedic surgeons managing these patients.
Purpose of the Study:
- To analyze the incidence and risk factors of post-operative wound infections in soft tissue sarcoma patients.
- To leverage big data analytics for a comprehensive understanding of infection determinants.
- To identify key predictors associated with wound infections after wide sarcoma resection.
Main Methods:
- Utilized the Health Insurance Review and Assessment Service (HIRA) database in South Korea (approx. 50 million individuals).
- Included patients who underwent wide excision of soft tissue sarcomas between 2010 and 2021.
- Employed random forest models and Shapley Additive Explanations (SHAP) for determinant analysis and predictor importance.
Main Results:
- A total of 10,969 patients were analyzed; 8.08% experienced post-operative infections.
- The transfusion rate was 20.67%, with blood transfusions identified as a major risk factor for infection.
- SHAP values indicated a positive association between blood transfusions and wound infection likelihood.
Conclusions:
- Perioperative blood transfusion, male sex, advanced age, and low socioeconomic status are significant predictors of wound infection in soft tissue sarcoma patients.
- Machine learning models (random forest) and SHAP values effectively identified these key risk factors.
- Findings highlight the need for careful management of transfusions and patient factors to reduce infection rates.
