Objectives: The purpose of this project is to develop a decision model of the detection and treatment of breast cancer. The model will use patient- assessed utilities (when they are available) concerning benefits and side effects or morbidity of the various therapeutic and information-gathering procedures. When utilities are not available the model will focus on duration of survival and default utilities. The model will also incorporate medical costs. Therapeutic and other decisions occurring early in the course of the disease assume that subsequent decisions are optimal for that patient's circumstances and with respect to her utilities. The overall objectives of this project are to (i) Improve patient care and quality of advice given to patients by providing oncologists with a tool for helping them choose interventions, and (ii) Provide a mechanism for evaluating the usefulness of current and proposed interventions, and also of research programs in breast cancer.
Specific Aims : l. Develop a decision model and interactive computer software tailored to individual patient characteristics and utilities (when available). The model will use probability distributions of outcomes based on literature reviews and metaanalyses. Default utilities will be based on the assessments of a panel of breast cancer physicians and nurses. 2. Develop instruments for assessing patient utilities at various time points in the disease. These will be available for use by clinicians as part of the interactive software and will help tailor therapy to an individual patient's characteristics and preferences. 3. Adapt the model to produce cost estimates of the various options. 4. Adapt the model to allow for its use in evaluating the utility of a new intervention or research program. Methods: Choice from among preventive or therapeutic strategies depends on options available in the future. The model developed will consider future options and their possible outcomes explicitly. Namely, the optimization algorithm will proceed backwards in time (that is, using dynamic programming). The approach is to consider the end of the disease cycle first. Using patient utilities of outcomes, the model will evaluate the expected utilities of the various metastatic settings. Each such setting has a probability of occurring depending on the interventions taken in the immediately preceding decision opportunity. Finding these probabilities allows for finding the expected utility of each possible intervention in the previous period by averaging outcome utilities in the current period with respect to the probabilities. Proceeding backwards in time-accruing costs and benefits in the process-leads to the early clinical and preclinical stages of the disease. In the process, each intervention and prevention strategy can be evaluated depending on an individual's characteristics and utilities.
Showing the most recent 10 out of 103 publications