Estimating the Patient's Price of Privacy in Liver Transplantation

Abstract

In the United States, patients with end-stage liver disease must join a waiting list to be eligible for cadaveric liver transplantation. Due to privacy concerns, the details of the composition of this waiting list are not publicly available. This paper considers the benefits associated with creating a more transparent waiting list. We study these benefits by modeling the organ accept/reject decision faced by these patients as a Markov decision process in which the state of the process is described by patient health, quality of the offered liver, and a measure of the rank of the patient in the waiting list. We prove conditions under which there exist structured optimal solutions, such as monotone value functions and control-limit optimal policies. We define the concept of the patient's price of privacy, namely, the number of expected life days lost due to the lack of complete waiting list information. We conduct extensive numerical studies based on clinical data, which indicate that this price of privacy is typically on the order of 5% of the optimal solution value.Subject classifications: dynamic programming/optimal control: applications, Markov; health care: treatment.

Optimally solving Markov decision processes with total expected discounted reward function: Linear programming revisited

“…al. [18] 2008 LP Bello et al [9] 2006 LP Le Ny et al [34] 2006 LP Sun [12] 2011 PIA Sandikci et al [39] 2008 PIA Alagoz et al [5] 2007 PIA Alagoz et al [6] 2007 PIA Idoumghar et al [24] 2006 PIA Kuppuswamy et al [27] 2005 PIA Chang et al [13] 2005 PIA Mosharaf et al [33] 2005 PIA Flapper et. al.…”

“…Our work provides evideence for answering this unresolved question, performing a careful computational experiment comparing PIA (typically superior to VIA) to LP. We focus on comparing PIA and LP particularly for MDPs with transition probability matrices populated around the diagonal entries, a common structure observed in MDPs developed for treatment optimization problems [4,5,6,28,39,40,42]. In these MDP models, the state space typically represents the health state of the patient and the transition probabilities determine how the patient health changes over time.…”

Alleviating the Patient's Price of Privacy Through a Partially Observable Waiting List

“…The primary motivation of this paper is to analyze the effect of using this imperfect information when making the accept/reject decisions. Sandıkçı et al (2008) introduce the concept of a patient's price of privacy (see §6.2) to measure the value of accessing perfect waiting list information and estimate the average price (reported as the percent increase in the patient's life expectancy) to be around 5% when compared with no information. Utilizing the published partial description of the waiting list should reduce their estimates; however, the question of whether the published partial information is good enough to effectively eliminate a patient's price of privacy is still open.…”

“…Our model builds upon the completely observable model of Sandıkçı et al (2008). However, whereas they assume perfect waiting list information, we relax this assumption and consider imperfect waiting list information, which is consistent with current practice.…”

“…This distinction makes our model a better representation of reality, but it also makes our problem significantly more difficult to analyze and solve. Furthermore, we provide a more accurate probability model than the one available in Sandıkçı et al (2008).…”