Automated Identification of Cardiopulmonary Disease Cases for Preoperative Risk Stratification Using Machine
Ishan Aggarwal1, Christopher Rhee, Mamta Chura
1From the Department of Anesthesiology and Perioperative Medicine, Medical College of Georgia, Augusta University, Augusta, Georgia.
Insights
A clinical insight bot efficiently extracts cardiovascular risk signals from preoperative notes, improving anesthetic planning. This AI tool offers a 100x efficiency gain, reducing manual review time while maintaining high precision and low false positive rates.
Area of Science:
- Artificial Intelligence in Medicine
- Clinical Informatics
- Natural Language Processing
Background:
- Preoperative chart review is time-consuming and error-prone, especially for cardiopulmonary conditions impacting anesthetic planning.
- Current methods struggle to efficiently identify critical cardiovascular risk signals from extensive documentation.
Purpose of the Study:
- To develop and evaluate a guideline-aligned "clinical insight bot" to automatically surface perioperative cardiovascular risk signals from free-text clinical notes.
- To improve the efficiency and accuracy of preoperative risk assessment for noncardiac surgery.
Main Methods:
- Analyzed 1000 de-identified cases from the PhysioNet MIMIC database.
- Utilized regex-based NLP for medical terminology extraction and TF-IDF/semantic embeddings for text feature encoding.
- Trained and compared four machine learning models (Logistic Regression, Random Forest, SVM, Naive Bayes) using stratified fivefold cross-validation.
Main Results:
- A linear Support Vector Machine (SVM) achieved the best performance (F1 score ≈ 0.71) in classifying "cardiopulmonary-only" versus "mixed/other" cases.
- The model demonstrated high precision (0.94) and a very low false positive rate (≈0.6%), with false negatives as the primary error.
- The pipeline processed documents rapidly, yielding an estimated 100x efficiency gain compared to manual review.
Conclusions:
- A guideline-aligned "clinical insight bot" can effectively transform unstructured preoperative notes into actionable prompts for cardiovascular risk signals.
- The bot's high precision and low false positive rate support safe integration into anesthesiology workflows, minimizing alert fatigue.
- Future work includes multicenter validation and structured data fusion to enhance sensitivity and evaluate clinical outcomes.
Background:
Preoperative chart review is time-consuming and prone to errors, particularly for cardiopulmonary conditions that impact anesthetic planning. We developed a guideline-aligned "clinical insight bot" that mines free-text documentation to surface perioperative cardiovascular risk signals relevant to the 2024 Mult Society perioperative guideline for noncardiac surgery.
Methods:
We analyzed 1000 de-identified medical cases from the PhysioNet MIMIC database. Medical terminology was extracted using regex-based NLP and categorized into 13 clinical specialties. Text features were encoded using TF-IDF vectorization and 1536-dimensional semantic embeddings stored in a PostgreSQL vector database (pgvector). Four machine learning models-Logistic Regression, Random Forest, Support Vector Machine (SVM), and Naive Bayes-were trained with stratified fivefold cross-validation to classify cases as "cardiopulmonary-only" versus "mixed/other." Performance was evaluated using accuracy, precision, recall, and F1 score, with statistical comparison via McNemar's test and bootstrap confidence intervals.
Results:
In a held-out test set of 200 notes (28 positive; 172 negatives; ~14% prevalence), a linear support vector machine achieved the best overall balance (F1 ≈ 0.71), with high precision (positive predictive value 0.94) and very low false positive rate (FPR) (1/172 ≈ 0.6%). False negatives were the dominant residual error class. The pipeline processed documents near-instantaneously and, when scaled to 1000 notes, replaced on the order of tens of clinician review hours (≈100× efficiency gain) while maintaining performance across common preoperative document types.
Conclusions:
A lightweight, guideline-aligned insight bot can transform unstructured preoperative notes into concise, stepwise prompts that flag cardiovascular risk signals before the day of surgery. High precision with a very low FPR supports safe integration with anesthesiology workflows by minimizing paging noise, whereas time savings create operational and financial value. Future work should emphasize multicenter validation, structured data fusion (including labs, imaging, and vitals) to improve sensitivity, and prospective evaluation of downstream clinical and operational outcomes.