Applications of stochastic modeling and data analytics techniques in healthcare decision making

Healthcare;Optimization;Stochastic modeling;data analytics;medical decision making;Influenza;Amyotrophic lateral sclerosis

Dalgic, Ozden Onuradvisor:Erenay, Fatih Safa ; affiliation1:Faculty of Engineering ; advisor:Ozaltin, Osman ; Erenay, Fatih Safa ; Ozaltin, Osman ;

University of Waterloo

关键词: data analytics;    Influenza;    Healthcare;    medical decision making;    Amyotrophic lateral sclerosis;    Stochastic modeling;    Doctoral Thesis;    Optimization;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/12754/3/Dalgic_Ozden.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

We present approaches utilizing aspects of data analytics and stochastic modeling techniquesand applied to various areas in healthcare. In general, the thesis has composed ofthree major components.Firtsly, we propose a comparison analysis between two of the very well-known infectiousdisease modeling techniques to derive effective vaccine allocation strategies. This study,has emerged from the fact that individuals are prioritized based on their risk profileswhen allocating limited vaccine stocks during an influenza pandemic. Computationallyexpensive but realistic agent-based simulations and fast but stylized compartmental modelsare typically used to derive effective vaccine allocation strategies. A detailed comparison ofthese two approaches, however, is often omitted. We derive age-specific vaccine allocationstrategies to mitigate a pandemic influenza outbreak in Seattle by applying derivative-freeoptimization to an agent-based simulation and also to a compartmental model. We comparethe strategies derived by these two approaches under various infection aggressiveness andvaccine coverage scenarios. We observe that both approaches primarily vaccinate schoolchildren, however they may allocate the remaining vaccines in different ways. The vaccineallocation strategies derived by using the agent-based simulation are associated with up to70% decrease in total cost and 34% reduction in the number of infections compared to thestrategies derived by the compartmental model. Nevertheless, the latter approach may stillbe competitive for very low and/or very high infection aggressiveness. Our results provideinsights about the possible differences between the vaccine allocation strategies derived byusing agent-based simulations and those derived by using compartmental models.Secondly, we introduce a novel and holistic scheme to capture the gradual amyotrophiclateral sclerosis progression based on the critical events referred as tollgates. Amyotrophiclateral sclerosis is neuro-degenerative and terminal disease. Patients with amyotrophic lateralsclerosis lose control of voluntary movements over time due to continuous degenerationof motor neurons. Using a comprehensive longitudinal dataset from Mayo Clinic’s ALSClinic in Rochester, MN, we characterize the progression through tollgates at the bodysegment (e.g., arm, leg, speech, swallowing, breathing) and patient levels over time. Wedescribe how the progression based on the followed tollgate pathways varies among patients and ultimately, how this type of progression characterization may be utilized for furtherstudies. Kaplan-Meier analysis are conducted to derive the probability of passing eachtollgate over time. We observe that, in each body segment, the majority of the patientshave their abilities affected or worse (Level1) at the first visit. Especially, the proportionof patients at higher tollgate levels is larger for arm and leg segments compared to others.For each segment, we derive the over-time progression pathways of patients in terms of thereached tollgates. Tollgates towards later visits show a great diversity among patients whowere at the same tollgate level at the first clinic visit. The proposed tollgate mechanismwell captures the variability among patients and the history plays a role on when patientsreach tollgates. We suggest that further and comprehensive studies should be conductedto observe the whole effect of the history in the future progression.Thirdly, based on the fact that many available databases may not have detailed medicalrecords to derive the necessary data, we propose a classification-based approach to estimatethe tollgate data using ALSFRS-R scores which are available in most databases. Weobserved that tollgates are significantly associated with the ALSFRS-R scores. Multiclassclassification techniques are commonly used in such problem; however, traditionalclassification techniques are not applicable to the problem of finding the tollgates dueto the constraint of that a patients’ tollgates under a specific segment for multiple visitshould be non-decreasing over time. Therefore, we propose two approaches to achieve amulti-class estimation in a non-decreasing manner given a classification method. Whilethe first approach fixes the class estimates of observation in a sequential manner, thesecond approach utilizes a mixed integer programming model to estimate all the classes ofa patients’ observations. We used five different multi-class classification techniques to beemployed by both of the above implementations. Thus, we investigate the performance ofclassification model employed under both approaches for each body segment.