| To Register SMDM Homepage | |
|
|
|
||||
Purpose: Determining when to screen for breast cancer with mammography is a complex problem involving multiple decision makers with competing objectives. The goal of this research is to develop a model for balancing patient and payer preferences to determine policies for mammography screening as a function of patient age and perceived condition.
Method: We use the theory of partially observable Markov processes (POMDPs) to develop efficient frontiers for balancing patient and payer preferences for breast cancer screening and treatment. A POMDP is a generalization of a Markov decision process that allows for incomplete state information. The condition of a patient is a function of many unobservable physical characteristics. There is uncertainty about any underlying disease and uncertainty associated with the response of a patient to a given treatment. In addition, different diagnostic and treatment procedures entail varying costs. In this case, the disease is observed only indirectly via a collection of incomplete or imperfect observations. We use a POMDP to incorporate uncertainty associated with the partial observability of the disease by the decision maker, and the uncertainty associated with the treatment outcome in determining the effectiveness of screening.
Results: We present a POMDP that can be used to determine: 1) when to recommend a mammogram and, 2) given the information provided by the mammogram, what treatment to provide. Further, we use the POMDP structure to develop a medical decision making tool for determining a “cost-effective” plan for mammography screening and breast cancer treatment. We develop efficient frontiers in order to explore the relationship between patient and payer preferences and to determine conditions for mammography screening to be cost-effective.
Conclusions: We have developed a model for determining cost-effective policies for breast cancer screening and treatment under conditions of uncertainty. This model incorporates uncertainty associated with the partial observability of the disease by the decision maker, the uncertainty associated with the treatment outcome, and the conflicting preferences of the patient and payer decision makers in determining the effectiveness of screening. The results show great promise as an alternative means for determining cost-effective monitoring and treatment policies for breast cancer.
See more of Oral Concurrent Session A - Simulation
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)