Financial Costs, Benefits, and Patient Risk Preferences in Node-Negative Breast Cancer: Insights from a Decision Analysis Model

Abstract

Since the 1988 National Cancer Institute Clinical Alert [1], treating nodenegative breast cancer patients outside the context of a clinical trial has been actively debated. To address this controversial question, we developed a decision analysis model to explore the effectiveness of adjuvant therapy in node-negative women at different levels of recurrence risk [2]. Decision analysis is one of a variety of techniques used to gather, analyze, and synthesize information. Other tools include randomized clinical trials, metaanalysis, database analysis, expert reviews of the literature, and consensus panels. Using these methods, a physician or policy maker ideally reaches a conclusion that will be translated into a clinical care or health care policy recommendation. Where does decision analysis fit into this hierarchy? It is an analytic tool that uses a decision tree which incorporates data from each of these sources to concurrently consider “apples and oranges,” for example, the trade-offs that must be made in the risks and benefits of any health care intervention [3]. Ideally, decision analysis “quantifies the qualifiers” and provides an integrated view of a specific question. Breast cancer is a fertile area for decision analysis since the key ingredients are identifiable and quantifiable: clear definitions of the disease and its natural history, definable benefits and risks with different modes of therapy, compet- ing alternative approaches, and controversy.