Full Record

New Search | Similar Records

Author
Title Optimal Timing of Statin Initiation for Patients with Type 2 Diabetes
URL
Publication Date
Date Accessioned
Degree MS
Discipline/Department Operations Research
Degree Level thesis
University/Publisher North Carolina State University
Abstract HMG Co-A reductase inhibitors (statins) are an important part of the treatment plan for patients with type 2 diabetes. However, the optimal time to initiate treatment is influenced by many factors. We investigate two such factors in this thesis: (1) the patient's long-term adherence to treatment and (2) the decision maker's criteria for optimal treatment initiation. Many patients who are prescribed statins stop taking the drug altogether or take less than the prescribed amount within the first year. This imperfect adherence can lessen the drug's benefit. We propose a Markov decision process model to optimize the treatment decision for hypercholesterolemia for patients with type 2 diabetes while considering issues of adherence to statins. Our model incorporates a discrete time Markov process for adherence states of the patient. We found that in the long run approximately 25% of patients remain highly adherent, taking 80 to 100% of their medication. We also find that patients with imperfect adherence should start statins 5 to 7 years later than their perfectly adherent counterparts. Although adherence levels greatly affect the optimal start time for statins. We found that starting statins later in life did not significantly increase the expected quality adjusted life years for patients with imperfect adherence. We conclude that it is more important for patients to improve their adherence than to adjust the timing of initiation to help compensate for imperfect adherence. We also consider three different decision making criteria with our model: society, patient, and third-party payer. Decision makers with these different perspectives have different objectives in mind. The patient is concerned with his or her quality of life, the third-party payer is concerned with minimizing costs, and society is concerned with maximizing rewards minus costs. These decision maker objectives are reflected in different reward functions in our MDP model. We find that it is optimal for patients to initiate statins early in the decision horizon under the patient perspective while the earliest optimal start times under the society and third-party payer perspectives are generally 4 and 15 years later, respectively. Finally, we formulate an inverse optimization model to estimate the implied societal willingness to pay. We use our MDP model and U.S. guidelines for initiating statins to estimate the implied reward for a year of quality life. Our estimates indicate a societal willingness to pay of between $120,000 and $160,000 per quality adjusted life year.
Subjects/Keywords statins; Markov decision process; inverse optimization; perspectives; diabetes
Contributors Salah E. Elmaghraby, Committee Member (advisor); Julie S. Ivy, Committee Member (advisor); Brian T. Denton, Committee Chair (advisor)
Rights I hereby certify that, if appropriate, I have obtained and attached hereto a written permission statement from the owner(s) of each third party copyrighted matter to be included in my thesis, dis sertation, or project report, allowing distribution as specified below. I certify that the version I submitted is the same as that approved by my advisory committee. I hereby grant to NC State University or its agents the non-exclusive license to archive and make accessible, under the conditions specified below, my thesis, dissertation, or project report in whole or in part in all forms of media, now or hereafter known. I retain all other ownership rights to the copyright of the thesis, dissertation or project report. I also retain the right to use in future works (such as articles or books) all or part of this thesis, dissertation, or project report.
Country of Publication us
Record ID oai:repository.lib.ncsu.edu:1840.16/1728
Repository ncsu
Date Retrieved
Date Indexed 2018-12-06
Issued Date 2009-08-04 00:00:00

Sample Search Hits | Sample Images

…Markov Decision Process NCEP: National Cholesterol Education Program NHI: National Health Informatics PDHR: Post Decision Horizon Reward QALY: Quality Adjusted Life Year RCT: Randomized Control Trial TC: Total Cholesterol UKPDS: United Kingdom Prospective…

…x5B;8] [9] [10], methods for using this information to determine a patient’s treatment plan are needed. We propose a Markov decision process (MDP) model to determine the optimal time for patients to start statins…

…of the parameters in our model, the monetary reward for a patient’s year of life, is highly debated. We formulate and solve an inverse optimization problem to estimate the implied societal willingness to pay based on U.S. guidelines for statin…

…Literature Review 2.1 Introduction Richard Bellman coined the term Dynamic Programming to describe the process of solving multi-stage decision problems. Bellman’s classic paper [11] describes the initial theory of dynamic programming. MDPs extend…

…decision making. In 1993, Sonnenberg and Beck [13] highlighted that many clinical problems could be modeled using Markov processes. Recent years have seen an increase in research in this area, as well as extensions to treatment optimization based…

…balance preferences. We combine the patient and third-party payer perspectives into the society perspective. 8 2.5 Inverse Optimization Lee et al. [25] use a computer simulation to model the treatment costs and effects on QALYs of a…

…patient’s quality year of life, we formulate a linear program to determine this value. Our inverse optimization problem determines the implied societal willingness to pay based on U.S. guidelines for statin initiation. 9 Chapter 3 Patient Adherence to…

…probabilistically over time according to a Markov process as patients age. HDL is known as the “good” cholesterol and LDL is known as the “bad” cholesterol. These two, together with triglycerides, make up TC. These continuous attributes are represented as a discrete…

.