ABSTRACT
Elderly patients have low surgical tolerance, weak postoperative basic metabolism and recovery ability, often accompanied by chronic diseases, which not only increases the risk of perioperative complications, but also improves the mortality of patients. Clinically, APACHE II, possum, ASA, SRS, TS, MODS, NNIS and other scales are usually used for preoperative risk assessment. The existing quantitative evaluation methods are easily affected by the patients' own diseases, fail to focus to the elderly patients, and cannot consider all the influencing factors before, during and after operation. Machine learning model has the advantages of strong robustness and excellent generalization ability, which has been widely used in medical aided diagnosis. Therefore, based on the literature analysis and combined with the existing quantitative scoring methods and statistical analysis methods, the study extracts and screens out the elements of the clinical evaluation model, then constructs the perioperative risk evaluation model of elderly patients by the machine learning method, which can refer to not only the operation implementation process and recent post-operative factors, but also the basic state of patients, chronic diseases history, operation difficulty, degree of anesthesia, degree of recovery, etc. In test, the ANN model has the highest accuracy and recall in the prediction of perioperative complications in elderly patients, 96.77 percent and 98.33 percent respectively, which is better than 85.48 percent and 68.33 percent of the traditional possum method. The accuracy and recall rate of SVM model in predicting perioperative death of elderly patients are the highest, which can reach 84.56 percent and 90.48 percent respectively, which are better than APACHE II, ASA and P-POSSUM evaluation models. In study, the machine learning model has a good prediction effect in the perioperative risk assessment for elderly patients, and can also be used as an auxiliary decision-making tool for clinical perioperative risk assessment.
- Sessler DI, Bloomstone JA, Aronson S, Pressure Quality Initiative consensus statement on intraoperative blood pressure, risk and outcomes for elective surgery. Br J Anaesth, 2019, vol. 122, pp. 563-574.Google ScholarCross Ref
- Kaiser HA, Saied NN, Kokoefer AS, Incidence and prediction of intraoperative and postoperative cardiac arrest requiring cardiopulmonary resuscitation and 30-day mortality in non-cardiac surgical patients. PLoS One, 2020, vol. 15, pp. 225939.Google ScholarCross Ref
- Kisa NG, Kisa E, Cevik BE. Prediction of Mortality in Patients After Oncologic Gastrointestinal Surgery: Comparison of the ASA, APACHE II, and POSSUM Scoring Systems. Cureus, 2021, vol. 13, pp. e13684.Google Scholar
- Vyas N L , Chavan S , Gandhi A , Predictability of Morbidity and Mortality using POSSUM Score in post-operative patients. IOSR Journal of Dental and Medical Sciences, 2021, vol. 20, pp.12-16.Google Scholar
- Li G, Walco J P, Mueller D A, Reliability of the ASA Physical Status Classification System in Predicting Surgical Morbidity: a Retrospective Analysis. Journal of Medical Systems, 2021, vol. 45, pp.1-8.Google ScholarDigital Library
- Sutiono A B, Arifin M Z, Adhipratama H, The utilization of APACHE II score to predict the incidence of ventilator-associated pneumonia in patients with severe traumatic brain injury: A single-center study.Interdisciplinary Neurosurgery, 2022, vol. 28, pp.101457.Google ScholarCross Ref
- Toker A K, Kose S, Turken M. Comparison of SOFA Score, SIRS, qSOFA, and qSOFA+L Criteria in the Diagnosis and Prognosis of Sepsis. Eurasian Journal of Medicine, 2021, vol. 53, pp.40-47.Google ScholarCross Ref
- Kanake V, Kale K, Mangam S, Thorax trauma severity score in patient with chest trauma: study at tertiary-level hospital. Indian Journal of Thoracic and Cardiovascular Surgery, 2022, vol. 38, pp.149-156.Google ScholarCross Ref
- Sharif A F, Fayed M M. Evaluation of Multiple Organ Dysfunction Score (MODS) and the Sequential Organ Failure Assessment (SOFA) score as in-hospital outcome predictors among cases of hydrogen cyanamide exposure: a cross-sectional study. Environmental Science and Pollution Research, 2021, pp.1-16.Google Scholar
- Fujii M, Shirakawa T, Nakamura M, Factors influencing the injury severity score and the probability of survival in patients who fell from height. Scientific Reports, 2021, vol. 11, pp.15561.Google ScholarCross Ref
- Wen X M, Ren N, Hua X X, Distribution and Antibacterial Resistance of Nosocomial Infection Pathogens from National Nosocomial Infection Surveillance System. Chinese Journal of Nosoconmiology, 2002, vol. 12, pp.241-244.Google Scholar
- Wilcox T, Smilowitz NR, Xia Y, Cardiovascular Risk Scores to Predict Perioperative Stroke in Noncardiac Surgery. Stroke, 2019, vol. 50, pp.2002-2006.Google ScholarCross Ref
- Egbuta C, Mason K P. Recognizing Risks and Optimizing Perioperative Care to Reduce Respiratory Complications in the Pediatric Patient. Journal of Clinical Medicine, 2020, vol. 9, pp.1942.Google ScholarCross Ref
- Jones T J, Bhattacharya B, Davis K A. Optimizing Cardiac Performance During the Perioperative Period. Current Geriatrics Reports, 2022, vol. 10, pp.175-181.Google ScholarCross Ref
- Wright DE,Knuesel SJ,Nagarur A,et al. Examining Risk:A Systematic Review of Perioperative Cardiac Risk Prediction Indices. Mayo Clin Proc,2019, vol. 94, pp.2277-2290.Google Scholar
- Bpr A, Jgp A, Pmm B. Mortalidad tras fractura de cadera: modelos predictivos - ScienceDirect. Medicina Clínica, 2020, vol. 154, pp.221-231.Google ScholarCross Ref
- Jones T J, Bhattacharya B, Davis K A. Optimizing Cardiac Performance During the Perioperative Period. Current Geriatrics Reports, 2022, vol. 10, pp.175-181.Google ScholarCross Ref
- Strasberg S M, Linehan D C, Hawkins W G. The accordion severity grading system of surgical complications. Annals of Surgery, 2009, vol. 250, pp.177.Google ScholarCross Ref
- Lindroth H, Bratzke Lisa, Twadell S, Predicting postoperative delirium severity in older adults: The role of surgical risk and executive function. Int J Geriatr Psychiatry,2019, vol. 34, pp.1018-1028.Google Scholar
- Rozzini R. Patients' preferences and "paternalistic approach" in elderly patients with atrial fibrillation. BMJ, 2021, vol. 7246, pp.1380-1384.Google Scholar
- Michel K F, Patel H D, Ziemba J B. Emergency department and hospital revisits after ambulatory surgery for kidney stones: an analysis of the Healthcare Cost and Utilization Project. Urolithiasis, 2021, pp.1-9.Google Scholar
- Chen L, Wang YB, Zhang YH, Effective prediction of postoperative complications for patients after open hepatectomy: a simplified scoring system based on perioperative parameters. BMC Surg,2019, vol. 19, pp.128.Google Scholar
- Construction of Perioperative Risk Assessment Model for Elderly Patients based on Machine Learning
Recommendations
Potentially Inappropriate Medication Prescription in Elderly Outpatients
ICMHI '17: Proceedings of the 1st International Conference on Medical and Health Informatics 2017Potentially inappropriate medication (PIM) is defined as medication which may result in higher than expected risk when used by elderly patients. Its risk of occurrence may be increased through the effects of drug interactions, adverse drug reactions (...
Influencing factors analysis and modeling of hospital-acquired infection in elderly patients
Hospital-acquired infection threatens the patients' health and life and also impacts medical quality by decreasing the bed turnover rate, prolonging hospitalization, increasing hospital costs and bringing the patients the huge economic losses. Therefore,...
Integrating machine learning predictions for perioperative risk management: Towards an empirical design of a flexible-standardized risk assessment tool
Graphical abstractML-augmented risk assessment tool based on feedback from clinicians. Across all displays, red indicates high or increasing risk values, while green suggests lower or decreasing risk values. Each display is sorted in descending order by ...
Abstract BackgroundSurgical patients are complex, vulnerable, and prone to postoperative complications that can potentially be mitigated with quality perioperative risk assessment and management. Several institutions have incorporated machine learning (...
Comments