Issue 2 (188), article 5
Kibern. vyčisl. teh., 2017, Issue 2 (188), pp.
Rudenko А.V.1, Professor, Corresponding member NAS of Ukraine,
First deputy of director
Nastenko I.А.1,2, Doctor (in Biology), PhD (in Technics),
Head of the Department of Information technologies and mathematical modeling
of physiological processes, Head of the Department of Biomedical Cybernetics
Zhurba O.А.1, Cardiovascular surgeon
Nosovets О.K.2, (in Technics),
Senior lecturer of the Department of Biomedical Cybernetics
Shardukova Y.V.1, Researcher at the Department of Information technology
and mathematical modeling of physiological processes
Lasoryshinets V.V.1, Professor, (in Medicine), Director
1 National Technical University of Ukraine «Igor Sikorsky Kyiv Polytechnic Institute», 37, Pobedi st., Kyiv, 03056, Ukraine
2 Government Facility “M.M. Amosov National Institute of Cardiovascular Surgery of National Academy of Medical Sciences of Ukraine”, 6, Amosov st., Kyiv, 03110, Ukraine
EVALUATION OF RISK FACTORS FOR OPERATIONS CORONARY BYPASS SURGERY ON A BEATING HEART
Introduction. The planned beating heart coronary aortic bypass grafting operations (BH CABG) prepares with and without parallel perfusion circulatory support. In second case the necessity of emergent use of circulatory support can appear. In these situations, the frequency of postoperative complications in circulatory system increases.
The decision about the planned use of circulatory support makes with analysis of big number of clinical data. This causes necessity to create the computer decision support systems to minimize the risk of emergent use of parallel perfusion.
The purpose of the article is to analyse statistically the risk factors for BH CABG operations on a working heart with the aim to minimize the risk of circulatory support emergent use.
Clinical material. 972 patients which undergone the BH CABG without circulatory support, 178 patients with planned use of circulatory support and 90 patients with emergent use of circulatory support. The 67 clinical parameters with package IBM SPSS Statistics 21.0 were analyzed.
Methods. The algorithm of binary logistic regression (BLR) for decision support systems development were used.
Results and discussion. The association between clinical values in groups without circulatory support, their planned and emergent use as well as their odds ratios and confidence intervals were analyzed. Then BLR algorithm to build the models for estimation of probability of planned and emergent use of circulatory support was used. The strategies of compulsory and stepwise inclusion of parameters were applied. The sensitivity, specificity and accuracy of the models obtained for learning and examination samples. The best models were chosen were calculated.
Conclusions. The created predictive models have a high sensitivity, specificity and accuracy, and can reduce the influence of subjective factors on medical decisions making regarding the use of the artificial circulatory support during off-pump coronary artery bypass surgery. Analysis of the variables included in the model, can contribute to a better understanding of the structure of existing pathogenic factors. The additional estimation the risk of cardiopulmonary bypass emergent use can reduce the likelihood of such situations occurrence.
Keywords: the beating heart coronary aortic bypass grafting, circulatory support with cardiopulmonary bypass, predicting algorithms.
1 Mykheev A.A., Klyuzhev V.M., Karpun N.A. Surgery on coronary arteries on a working heart without artificial circulation in IHD patients. M.: Medicine, 2001. 43 p. (in Russian).
2 Allen B. S., Rosenkranz E.R., Buckberg G.D. Studies of controlled reperfusion after ischemia. VII. High oxygen requirements of dyskinetic cardiac muscle. Journal of Thoracic and Cardiovascular Surgery. 1986. No92. P. 543–552.
3 Mo A., Lin H., Wen Z.Efficacy and safety of on-pump beating heart surgery. Ann Thorac Surg. 2008. No 86. P. 1914–1918.
4 Puskas J. Presidential Address, 2009: ISMICS Means Innovation. Innovations: Technology & Techniques in Cardiothoracic & Vascular Surgery. 2009. No 4. P. 240–247.
5 Shabalkyn B.V., Zhbanov Y.V. Minimally invasive myocardial revascularization or aortocoronary bypass without artificial circulation? Bulletin of the Center Bakulev RAMS. V All-Russian Congress of Cardiovascular Surgeons. Novosibirsk, 1999. 152 p. (in Russian).
6 Beauford R.B., Goldstein D.J., Sardari F.F. Multivessel off-pump revascularization in octogenarians: early and midterm outcomes. Ann. Thorac. Surg. 2003. Vol. 76. P. 12–17.
7 Stamou S., Bail A., Boyce S.Coronary revascularization of the circumflex. Ann. Thorac. Surg. 2000. Vol. 70. P. 1371–1377.
8 Witten Ian H., Frank Eibe, Hall Mark A. Data Mining: Practical Machine Learning Tools and Techniques. [3rd Ed.]. Morgan Kaufmann, 2011. P. 664.
9 McHugh M. L. The odds ratio: calculation, usage, and interpretation. Biochemia Medica. 2009. No19 (2). P. 120–126.
10 Sperandei S. Understanding logistic regression analysis. Biochemia Medica. 2014. 24(1). P. 12–18.