Data-driven management of post-transplant medications: An ambiguous partially observable markov decision process approach

Alireza Boloori, Soroush Saghafian, Harini A. Chakkera, Curtiss B Cook

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Problem definition: Organ-transplanted patients typically receive high amounts of immunosuppressive drugs (e.g., tacrolimus) as a mechanism to reduce their risk of organ rejection. However, because of the diabetogenic effect of these drugs, this practice exposes them to a greater risk of new-onset diabetes after transplantation (NODAT), and hence, becoming insulin dependent. We study and develop effective medication management strategies to address the common conundrum of balancing the risk of organ rejection versus that of NODAT. Academic/practical relevance: Our research contributes to the healthcare operations management literature by developing a robust stochastic decision-making framework that allows for incorporating (1) false-positive and false-negative errors of medical tests, (2) inevitable estimation errors when data sets are used, (3) variability among physician’ attitudes toward ambiguous outcomes, and (4) dynamic and patient risk-profile-dependent progression of health conditions. Methodology: We apply an ambiguous partially observable Markov decision process (APOMDP) approach where dynamic optimization with respect to a “cloud” of possible models allows us to make decisions that are robust to potential misspecifications of risks. Results: We first provide various structural results that facilitate characterizing the optimal medication policies. Utilizing a clinical data set, we then compare the performance of the optimal medication policies obtained from our APOMDP model with the policies currently used in the medical practice. We observe that, in one year after transplant, our proposed policies can improve the life expectancy of each patient up to 4.58%, while reducing the medical expenditures up to 11.57%. Managerial implications: Balancing the risks of organ rejection and diabetes complications and considering factors such as physicians’ attitudes toward ambiguous outcomes, partial observability of medical tests, and patient-specific risk factors are shown to result in more cost-effective strategies for management of post-transplant medications compared with the current medical practice. Finally, simultaneous management of medications can facilitate the care coordination process between transplantation/nephrology and endocrinology departments of a hospital that are typically in charge of administering such medications.

Original languageEnglish (US)
Pages (from-to)1066-1087
Number of pages22
JournalManufacturing and Service Operations Management
Volume22
Issue number5
DOIs
StatePublished - Sep 2020

Keywords

  • Ambiguous POMDP
  • Cloud of models
  • Conservatism level
  • Diabetes
  • Immunosuppressive drug
  • Kidney transplant

ASJC Scopus subject areas

  • Strategy and Management
  • Management Science and Operations Research

Fingerprint

Dive into the research topics of 'Data-driven management of post-transplant medications: An ambiguous partially observable markov decision process approach'. Together they form a unique fingerprint.

Cite this