Machine-Learning Model for the Prediction of Hypoxaemia during Endoscopic Retrograde Cholangiopancreatography under Monitored Anaesthesia Care
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Purpose: Hypoxaemia is a significant adverse event during endoscopic retrograde cholangiopancreatography (ERCP) under mon itored anaesthesia care (MAC); however, no model has been developed to predict hypoxaemia. We aimed to develop and compare logistic regression (LR) and machine learning (ML) models to predict hypoxaemia during ERCP under MAC. Materials and Methods: We collected patient data from our institutional ERCP database. The study population was randomly divid ed into training and test sets (7:3). Models were fit to training data and evaluated on unseen test data. The training set was further split into k-fold (k=5) for tuning hyperparameters, such as feature selection and early stopping. Models were trained over k loops; the i-th fold was set aside as a validation set in the i-th loop. Model performance was measured using area under the curve (AUC). Results: We identified 6114 cases of ERCP under MAC, with a total hypoxaemia rate of 5.9%. The LR model was established by combining eight variables and had a test AUC of 0.693. The ML and LR models were evaluated on 30 independent data splits. The average test AUC for LR was 0.7230, which improved to 0.7336 by adding eight more variables with an l1 regularisation-based selec tion technique and ensembling the LRs and gradient boosting algorithm (GBM). The high-risk group was discriminated using the GBM ensemble model, with a sensitivity and specificity of 63.6% and 72.2%, respectively. Conclusion: We established GBM ensemble model and LR model for risk prediction, which demonstrated good potential for pre venting hypoxaemia during ERCP under MAC.
Abstract
Purpose: Hypoxaemia is a significant adverse event during endoscopic retrograde cholangiopancreatography (ERCP) under mon itored anaesthesia care (MAC); however, no model has been developed to predict hypoxaemia. We aimed to develop and compare logistic regression (LR) and machine learning (ML) models to predict hypoxaemia during ERCP under MAC. Materials and Methods: We collected patient data from our institutional ERCP database. The study population was randomly divid ed into training and test sets (7:3). Models were fit to training data and evaluated on unseen test data. The training set was further split into k-fold (k=5) for tuning hyperparameters, such as feature selection and early stopping. Models were trained over k loops; the i-th fold was set aside as a validation set in the i-th loop. Model performance was measured using area under the curve (AUC). Results: We identified 6114 cases of ERCP under MAC, with a total hypoxaemia rate of 5.9%. The LR model was established by combining eight variables and had a test AUC of 0.693. The ML and LR models were evaluated on 30 independent data splits. The average test AUC for LR was 0.7230, which improved to 0.7336 by adding eight more variables with an l1 regularisation-based selec tion technique and ensembling the LRs and gradient boosting algorithm (GBM). The high-risk group was discriminated using the GBM ensemble model, with a sensitivity and specificity of 63.6% and 72.2%, respectively. Conclusion: We established GBM ensemble model and LR model for risk prediction, which demonstrated good potential for pre venting hypoxaemia during ERCP under MAC.
참고문헌 (40)
- Hu LH, "ERCP development in the largest developing country : a national survey from China in 2013", Gastrointest Endosc, 84, pp.659~666 (2016)
- Coelho-Prabhu N, "Endoscopic retrograde cholangiopancreatography : utilisation and outcomes in a 10-year population-based cohort", BMJ Open, 3, pp.e002689~ (2013)
- 안동원, "Practice of Endoscopic Retrograde Cholangiopancreatography in Korea: Results from a National Survey", 대한췌담도학회지, 24(1), pp.21~30 (2019), 대한췌장담도학회[DOI]
- Wadhwa V, "Monitoring standards in sedation and analgesia : the odyssey of capnography in sedation for gastroenterology procedures", Curr Opin Anaesthesiol, 32, pp.453~456 (2019)
- Wehrmann T, "Sedation with propofol for interventional endoscopic procedures : a risk factor analysis", Scand J Gastroenterol, 43, pp.368~374 (2008)
- Sudheer PS, "Haemodynamic effects of the prone position : a comparison of propofol total intravenous and inhalation anaesthesia", Anaesthesia, 61, pp.138~141 (2006)
- Edgcombe H, "Anaesthesia in the prone position", Br J Anaesth, 100, pp.165~183 (2008)
- Early DS, "Guidelines for sedation and anesthesia in GI endoscopy", Gastrointest Endosc, 87, pp.327~337 (2018)
- Klare P, "Capnographic monitoring of midazolam and propofol sedation during ERCP : a randomized controlled study(EndoBreath study)", Endoscopy, 48, pp.42~50 (2016)
- Gerstenberger PD, "Capnography and patient safety for endoscopy", Clin Gastroenterol Hepatol, 8, pp.423~425 (2010)
- Yang JF, "Efficacy and safety of propofol-mediated sedation for outpatient endoscopic retrograde cholangiopancreatography(ERCP)", Dig Dis Sci, 61, pp.1686~1691 (2016)
- Park CH, "Efficacy and safety of etomidate-based sedation compared with propofol-based sedation during ERCP in low-risk patients : a double-blind, randomized, noninferiority trial", Gastrointest Endosc, 87, pp.174~184 (2018)
- Smith ZL, "A randomized controlled trial evaluating general endotracheal anesthesia versus monitored anesthesia care and the incidence of sedation-related adverse events during ERCP in high-risk patients", Gastrointest Endosc, 89, pp.855~862 (2019)
- Hormati A, "Prevalence of anesthetic and gastrointestinal complications of endoscopic retrograde cholangiopancreatography", Anesth Pain Med, 9, pp.e95796~ (2019)
- Qadeer MA, "Capnographic monitoring of respiratory activity improves safety of sedation for endoscopic cholangiopancreatography and ultrasonography", Gastroenterology, 136, pp.1568~1576 (2009)
- Kim SH, "The addition of capnography to standard monitoring reduces hypoxemic events during gastrointestinal endoscopic sedation : a systematic review and meta-analysis", Ther Clin Risk Manag, 14, pp.1605~1614 (2018)
- Coté GA, "Incidence of sedation-related complications with propofol use during advanced endoscopic procedures", Clin Gastroenterol Hepatol, 8, pp.137~142 (2010)
- Berzin TM, "A prospective assessment of sedation-related adverse events and patient and endoscopist satisfaction in ERCP with anesthesiologist-administered sedation", Gastrointest Endosc, 73, pp.710~717 (2011)
- Wani S, "Obesity as a risk factor for sedation-related complications during propofol-mediated sedation for advanced endoscopic procedures", Gastrointest Endosc, 74, pp.1238~1247 (2011)
- Buxbaum J, "Anesthetist-directed sedation favors success of advanced endoscopic procedures", Am J Gastroenterol, 112, pp.290~296 (2017)
- Müller S, "Predictive factors of oxygen desaturation of patients submitted to endoscopic retrograde cholangiopancreatography under conscious sedation", Arq Gastroenterol, 41, pp.162~166 (2004)
- Perbtani YB, "Impact of endotracheal intubation on interventional endoscopy unit efficiency metrics at a tertiary academic medical center", Am J Gastroenterol, 111, pp.800~807 (2016)
- Martindale SJ, "Anaesthetic considerations during endoscopic retrograde cholangiopancreatography", Anaesth Intensive Care, 34, pp.475~480 (2006)
- Allyn J, "A comparison of a machine learning model with EuroSCORE II in predicting mortality after elective cardiac surgery : a decision curve analysis", PLoS One, 12, pp.e0169772~ (2017)
- Heo J, "Machine learning–based model for prediction of outcomes in acute stroke", Stroke, 50, pp.1263~1265 (2019)
- Ming C, "Machine learning techniques for personalized breast cancer risk prediction : comparison with the BCRAT and BOADICEA models", Breast Cancer Res, 21, pp.75~ (2019)
- Friedman JH, "Greedy function approximation : a gradient boosting machine", Ann Stat, 29, pp.1189~1232 (2001)
- Chen X, "EGBMMDA : extreme gradient boosting machine for MiRNA-disease association prediction", Cell Death Dis, 9, pp.3~ (2018)
- Yeboah J, "Association between sleep apnea, snoring, incident cardiovascular events and all-cause mortality in an adult population : MESA", Atherosclerosis, 219, pp.963~968 (2011)
- Tibshirani R, "Regression shrinkage and selection via the Lasso", J R Stat Soc Series B Methodol, 58, pp.267~288 (1996)
- Chen T, "Xgboost: a scalable tree boosting system", Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.785~794 (2016), KDD
- Sorser SA, "Complications of ERCP in patients undergoing general anesthesia versus MAC", Dig Dis Sci, 59, pp.696~697 (2014)
- Smith ZL, "Anesthesia-administered sedation for endoscopic retrograde cholangiopancreatography : monitored anesthesia care or general endotracheal anesthesia?", Curr Opin Anaesthesiol, 32, pp.531~537 (2019)
- Smith ZL, "Type of sedation and the need for unplanned interventions during ERCP : analysis of the clinical outcomes research initiative national endoscopic database(CORI-NED)", Frontline Gastroenterol, 11, pp.104~110 (2020)
- Thosani N, "Deep sedation or general anesthesia for ERCP?", Dig Dis Sci, 58, pp.3061~3063 (2013)
- Barnett SR, "Deep sedation without intubation for ERCP is appropriate in healthier, non-obese patients", Dig Dis Sci, 58, pp.3287~3292 (2013)
- Kuzhively J, "Anesthesia and airway management for gastrointestinal endoscopic procedures outside the operating room", Curr Opin Anaesthesiol, 32, pp.517~522 (2019)
- Baron TH, "Quality indicators for endoscopic retrograde cholangiopancreatography", Gastrointest Endosc, 63(4 Suppl), pp.S29~S34 (2006)
- Triantafillidis JK, "Sedation in gastrointestinal endoscopy : current issues", World J Gastroenterol, 19, pp.463~481 (2013)
- Hoka S, "Propofol-induced increase in vascular capacitance is due to inhibition of sympathetic vasoconstrictive activity", Anesthesiology, 89, pp.1495~1500 (1998)