Decision analysis for transporting critically ill patients with cardiovascular diseases

Abstract

The transportation of critically ill patients with cardiovascular diseases is a critical aspect of emergency medical care, particularly in countries with vast territories, dispersed populations, and centralized healthcare services. In Kazakhstan, where specialized cardiac centers are concentrated in major urban areas, the need to transfer patients across long distances presents a complex decision-making problem that intertwines logistical, clinical, and economic challenges. Selecting the most appropriate mode of transport whether by airplane, helicopter, ambulance, train, or private clinical vehicle requires a careful evaluation of trade-offs between cost efficiency, transportation time, and patient safety during transit. This dissertation addresses this multifaceted problem by applying a rigorous decision-analytic framework grounded in Multiattribute Utility Theory (MAUT). The methodological contribution of this research lies in the development and application of a multiattribute utility function, U(X1, X2, X3), which incorporates three key attributes: transportation cost savings (X1), time savings (X2), and the medical impact of transportation on the patient’s health status (X3). The widely used APACHE II scoring system is used to measure how sick a patient is, which gives a more complete and accurate picture of their health than the color-based triage system that is currently used in Kazakhstan. We got our data from official sources, like the National Coordination Centre for Emergency Medicine (NCCEM), and we also got structured expert input from experienced medical professionals, such as an anesthesiologist who is familiar with transporting high-risk patients. The results show that the decision maker's preferences are not utility independent across the three attributes. This insight led to the construction of a utility function under partial utility independence (PUI) condition, using a functional form derived from the Utility Dependence Matrix (UDM) and elicited through certainty equivalent methods. The resulting analysis produced a ranking of the transportation alternatives under the PUI framework: airplane (1st), helicopter (2nd), ambulance (3rd), private clinical cars (4th), and train (5th). Further comparison was conducted using a multilinear utility function assuming complete utility independence and a series of single-attribute evaluations. These comparative analyses demonstrated the sensitivity of rankings to assumptions about attribute independence and underscored the value of employing a nuanced, realistic representation of decision maker’s preferences in high-stakes healthcare contexts. This study contributes to both the theoretical and applied domains of decision analysis. It offers a novel decision-support model tailored to the healthcare landscape of Kazakhstan, while also serving as a replicable framework for other nations facing similar geographical and infrastructural constraints. Ultimately, the proposed model enhances the objectivity and transparency of transport decision-making, supporting policymakers and emergency coordinators in their efforts to deliver timely, cost-effective, and clinically appropriate care to the most vulnerable patient populations.