Кардиоваскулярная терапия и профилактика (Apr 2019)

Optimization of management of patients with critical lower limb ischemia, allowing for risk of gangrene

  • A. V. Bykov,
  • N. A. Korenevsky,
  • S. A. Parkhomenko,
  • I. I. Khripina

DOI
https://doi.org/10.15829/1728-8800-2019-2-38-44
Journal volume & issue
Vol. 18, no. 2
pp. 38 – 44

Abstract

Read online

Aim. To develop mathematical models for predicting the risk of gangrene of the lower limb and the algorithm for selecting of treatment strategies for vascular surgeons and angiologists.Material and methods. The results presented in this work are based on a six-year-old (since 2011) observation of 400 patients with chronic obliterating diseases of the lower limb arteries, some of which had combinations of ischemic damage to the central hemodynamic system, heart and brain. Patients had different stages of the disease, up to critical ischemia and gangrene, which requires amputation of the lower limbs. During the exploratory analysis, we selected 8 informative features, which were used to develop mathematical models that divide patients into classes: low confidence in the development of gangrene; average confidence in the development of gangrene; high confidence in the development of gangrene and very high confidence in the development of gangrene. For each identified class, confidence in the development of gangrene was determined by an individual treatment strategy, implemented in accordance with the intelligent decision support system by a decision-making algorithm.Results. It was shown that, compared with traditional treatment tactics, using of presented algorithm can increase the speed of positive results achieving by 3,4 times (68,3%), reduce the risk of lower limb gangrene development by 2,8 times (61,6%) and reduce the risk of limb amputation by 4,1 times (68,1%).Conclusion. The obtained mathematical models should be used in the medical practice of vascular surgeons and angiologists, both in the form of software for smartphones and tablet computers, and as part of decision support systems, including telemedicine systems.

Keywords