Cardiac Operative Risk in Latin America: A Comparison of Machine Learning Models vs EuroSCORE-II

Molina, Raúl Santiago; Molina-Rodríguez, María Alejandra; Rincón, Francisco Mauricio; Maldonado, Javier

doi:10.1016/j.athoracsur.2021.02.052

Cited by 9 publications

(12 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among these 5 articles, 4 [26][27][28]35 had wide confidence intervals ranging from 0.15 to 0.35, and 1, 37 which used waveform input features, had a narrow confidence interval of 0.01 (albeit with moderate performance in predicting postoperative deterioration, AUROC 0.71). Among all 36 articles, 5 31,33,35,39,40 (13.8%) performed external validation, 2 9,29 (5.6%) performed real-time validation, and 29 (80.6%) performed internal validation only. None of the articles described equity analyses in which model performance was stratified by sex or race.…”

Section: Resultsmentioning

confidence: 99%

Artificial Intelligence–enabled Decision Support in Surgery

et al. 2023

View full text Add to dashboard Cite

Objective: To summarize state-of-the-art artificial intelligence-enabled decision support in surgery and to quantify deficiencies in scientific rigor and reporting. Background: To positively affect surgical care, decision-support models must exceed current reporting guideline requirements by performing external and real-time validation, enrolling adequate sample sizes, reporting model precision, assessing performance across vulnerable populations, and achieving clinical implementation; the degree to which published models meet these criteria is unknown. Methods: Embase, PubMed, and MEDLINE databases were searched from their inception to September 21, 2022 for articles describing artificial intelligence-enabled decision support in surgery that uses preoperative or intraoperative data elements to predict complications within 90 days of surgery. Scientific rigor and reporting criteria were assessed and reported according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines. Results: Sample size ranged from 163-2,882,526, with 8/36 articles (22.2%) featuring sample sizes of less than 2000; 7 of these 8 articles (87.5%) had below-average (< 0.83) area under the receiver operating characteristic or accuracy. Overall, 29 articles (80.6%) performed internal validation only, 5 (13.8%) performed external validation, and 2 (5.6%) performed real-time validation. Twenty-three articles (63.9%) reported precision. No articles reported performance across sociodemographic categories. Thirteen articles (36.1%) presented a framework that could be used for clinical implementation; none assessed clinical implementation efficacy. Conclusions: Artificial intelligence-enabled decision support in surgery is limited by reliance on internal validation, small sample sizes that risk overfitting and sacrifice predictive performance, and failure to report confidence intervals, precision, equity analyses, and clinical implementation. Researchers should strive to improve scientific quality.

show abstract

Section: Resultsmentioning

confidence: 99%

Artificial Intelligence–enabled Decision Support in Surgery

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Compared to traditional cardiac surgical risk models limited to electronic health record data and other machine learning-based prediction models developed for anesthesiology and critical care settings, [17][18][19][20]37,38 waveform features. Our study also leverages tensor decomposition, a method of parsing multidimensional arrays of waveform data in a computationally efficient manner, for feature reduction.…”

Section: Discussionmentioning

confidence: 99%

“…14,15 Data science approaches may overcome such issues through improved synthesis of diverse, complex health data for detecting digital signatures of early-stage clinical deterioration. [16][17][18][19][20] Through this observational study of high-fidelity electronic health record and physiologic waveform ICU data from an academic quaternary care hospital, we leveraged machine learning techniques for early detection of postoperative deterioration among patients undergoing cardiac surgical procedures. We hypothesized that patterns exist within both electronic health record data and physiologic waveform data predictive of hemodynamic deterioration and that the performance of models using both electronic health record and physiologic waveform data to predict postoperative deterioration is superior to models using either modality alone.…”

Section: What We Already Know About This Topicmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of Postoperative Deterioration in Cardiac Surgery Patients Using Electronic Health Record and Physiologic Waveform Data

et al. 2022

View full text Add to dashboard Cite

Background Postoperative hemodynamic deterioration among cardiac surgical patients can indicate or lead to adverse outcomes. Whereas prediction models for such events using electronic health records or physiologic waveform data are previously described, their combined value remains incompletely defined. The authors hypothesized that models incorporating electronic health record and processed waveform signal data (electrocardiogram lead II, pulse plethysmography, arterial catheter tracing) would yield improved performance versus either modality alone. Methods Intensive care unit data were reviewed following elective adult cardiac surgical procedures at an academic center between 2013-2020. Model features included electronic health record features and physiologic waveforms. Tensor decomposition was used for waveform feature reduction. Machine learning-based prediction models included a 2013-2017 training set and a 2017-2020 temporal holdout test set. The primary outcome was a postoperative deterioration event, defined as a composite of low cardiac index <2.0mL/min/m2, mean arterial pressure <55mmHg sustained ≥120 minutes, new or escalated inotrope/vasopressor infusion, epinephrine bolus ≥1mg, or intensive care unit mortality. Prediction models analyzed data 8 hours prior to events. Results Among 1,555 cases, 185 (12%) experienced 276 deterioration events, most commonly including low cardiac index (7.0% of patients), new inotrope (1.9%), and sustained hypotension (1.4%). The best performing model on the 2013-2017 training set yielded a c-statistic of 0.803 (95% CI 0.799-0.807) although performance was substantially lower in the 2017-2020 test set (0.709, 0.705-0.712). Test set performance of the combined model was greater than corresponding models limited to solely electronic health record features (0.641, 0.637-0.646) or waveform features (0.697, 0.693-0.701). Conclusions Clinical deterioration prediction models combining electronic health record data and waveform data were superior to either modality alone, and performance of combined models was primarily driven by waveform data. Decreased performance of prediction models during temporal validation may be explained by dataset shift, a core challenge of healthcare prediction modeling.

show abstract

“…14 Recent literature has demonstrated that ML models outperform simple statistical models in predicting outcomes from datasets. 15,16 In this study, we have employed a supervised ML approach to train models to predict binary class classifications of complications in Indian patients undergoing coronary artery bypass grafting (CABG). Thus, the occurrence of a complication was classed as ‘1’ and its absence was recorded as ‘0’.…”

Section: Introductionmentioning

confidence: 99%

Machine learning algorithms for population-specific risk score in coronary artery bypass grafting

Swamy

Rajagopal

Krishnan

et al. 2023

Asian Cardiovasc Thorac Ann

View full text Add to dashboard Cite

Background The aim of this study was to develop a new risk prediction score (NH Score) for patients undergoing coronary artery bypass grafting (CABG) specific to the Indian population and compare it to the Society of Thoracic Surgeon (STS) Score and the EuroSCORE II. Method The baseline features of adult patients who underwent CABG between the years 2015 and 2021 ( n = 6703) were taken and split into training data (2015–2020; n = 5561) and validation data (2020–2021; n = 1142). The CatBoost algorithm was trained to predict risk scores (NH score), and the performance was tested on the validation set by Precision-Recall Curve and F1 Score. Model calibration was measured by the Brier Score, Expected Calibration Error and Maximum Calibration Error. Results The NH score outperformed both the STS and EuroSCORE II for all outcomes. For mortality, the PR AUC for NH Score was (0.463 [95% confidence interval [CI], 0.28–0.64]) compared to 0.113 [95% CI, 0.04–0.22] for the STS score and 0.146 [95% CI, 0.06–0.31] for the EuroSCORE II ( p ≪ 0.0001). With respect to morbidity NH Score was superior to the STS score (0.43 [95% CI, 0.33–0.50]) vs. (0.229 [95% CI, 0.18–0.3, p < 0.0001). The observed to the predicted ratio for NH score was superior to the STS Score and similar to EuroSCORE II. NH Score was also more accurate at predicting the risk of prolonged ventilation compared to the STS Score. Conclusion NH score shows an excellent improvement over the performance of STS score and EuroSCORE II for modelling risk predictions for patients undergoing CABG in Indian population. It warrants further validation for larger datasets.

show abstract

Cardiac Operative Risk in Latin America: A Comparison of Machine Learning Models vs EuroSCORE-II

Cited by 9 publications

References 22 publications

Artificial Intelligence–enabled Decision Support in Surgery

Artificial Intelligence–enabled Decision Support in Surgery

Prediction of Postoperative Deterioration in Cardiac Surgery Patients Using Electronic Health Record and Physiologic Waveform Data

Machine learning algorithms for population-specific risk score in coronary artery bypass grafting

Contact Info

Product

Resources

About