Applications of stochastic modeling and data analytics techniques in healthcare decision making
Year: 2017
Type: dissertation
Abstract: We present approaches utilizing aspects of data analytics and stochastic modeling techniques \nand applied to various areas in healthcare. In general, the thesis has composed of \nthree major components. \nFirtsly, we propose a comparison analysis between two of the very well-known infectious \ndisease modeling techniques to derive effective vaccine allocation strategies. This study, \nhas emerged from the fact that individuals are prioritized based on their risk profiles \nwhen allocating limited vaccine stocks during an influenza pandemic. Computationally \nexpensive but realistic agent-based simulations and fast but stylized compartmental models \nare typically used to derive effective vaccine allocation strategies. A detailed comparison of \nthese two approaches, however, is often omitted. We derive age-specific vaccine allocation \nstrategies to mitigate a pandemic influenza outbreak in Seattle by applying derivative-free \noptimization to an agent-based simulation and also to a compartmental model. We compare \nthe strategies derived by these two approaches under various infection aggressiveness and \nvaccine coverage scenarios. We observe that both approaches primarily vaccinate school \nchildren, however they may allocate the remaining vaccines in different ways. The vaccine \nallocation strategies derived by using the agent-based simulation are associated with up to \n70% decrease in total cost and 34% reduction in the number of infections compared to the \nstrategies derived by the compartmental model. Nevertheless, the latter approach may still \nbe competitive for very low and/or very high infection aggressiveness. Our results provide \ninsights about the possible differences between the vaccine allocation strategies derived by \nusing agent-based simulations and those derived by using compartmental models. \nSecondly, we introduce a novel and holistic scheme to capture the gradual amyotrophic \nlateral sclerosis progression based on the critical events referred as tollgates. Amyotrophic \nlateral sclerosis is neuro-degenerative and terminal disease. Patients with amyotrophic lateral \nsclerosis lose control of voluntary movements over time due to continuous degeneration \nof motor neurons. Using a comprehensive longitudinal dataset from Mayo Clinic’s ALS \nClinic in Rochester, MN, we characterize the progression through tollgates at the body \nsegment (e.g., arm, leg, speech, swallowing, breathing) and patient levels over time. We \ndescribe how the progression based on the followed tollgate pathways varies among patients and ultimately, how this type of progression characterization may be utilized for further \nstudies. Kaplan-Meier analysis are conducted to derive the probability of passing each \ntollgate over time. We observe that, in each body segment, the majority of the patients \nhave their abilities affected or worse (Level1) at the first visit. Especially, the proportion \nof patients at higher tollgate levels is larger for arm and leg segments compared to others. \nFor each segment, we derive the over-time progression pathways of patients in terms of the \nreached tollgates. Tollgates towards later visits show a great diversity among patients who \nwere at the same tollgate level at the first clinic visit. The proposed tollgate mechanism \nwell captures the variability among patients and the history plays a role on when patients \nreach tollgates. We suggest that further and comprehensive studies should be conducted \nto observe the whole effect of the history in the future progression. \nThirdly, based on the fact that many available databases may not have detailed medical \nrecords to derive the necessary data, we propose a classification-based approach to estimate \nthe tollgate data using ALSFRS-R scores which are available in most databases. We \nobserved that tollgates are significantly associated with the ALSFRS-R scores. Multiclass \nclassification techniques are commonly used in such problem; however, traditional \nclassification techniques are not applicable to the problem of finding the tollgates due \nto the constraint of that a patients’ tollgates under a specific segment for multiple visit \nshould be non-decreasing over time. Therefore, we propose two approaches to achieve a \nmulti-class estimation in a non-decreasing manner given a classification method. While \nthe first approach fixes the class estimates of observation in a sequential manner, the \nsecond approach utilizes a mixed integer programming model to estimate all the classes of \na patients’ observations. We used five different multi-class classification techniques to be \nemployed by both of the above implementations. Thus, we investigate the performance of \nclassification model employed under both approaches for each body segment.
Source: UWSpace (University of Waterloo)
Author: Özden O. Dalgıç
Language: English
Subfield: Applied Mathematics
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2777608508