Descriptive review of the literature on breast cancer outcomes: 1990 through 2000

Abstract

Background: There is increasing interest in the incorporation of nonbiomedical outcomes into cancer research.

Objective: Our goal was to review the use of nonbiomedical outcomes in research on breast cancer care.

Data sources: We conducted a MEDLINE search of all studies on breast cancer quality of life, preferences, satisfaction, and economics that were published during the period from January 1, 1990, through December 31, 2000. We also searched bibliographies of published articles.

Study selection: We included original primary research and excluded reviews, methods papers, studies conducted outside the United States and Western Europe, or studies with fewer than 100 subjects.

Data abstraction: Data were abstracted by using a structured tool.

Data synthesis: There were 1089 articles identified; 230 were included. The greatest proportion of research focused on survivorship followed by screening. The most frequently reported outcomes were health-related quality of life (54%) followed by economic analyses (38%) and patient satisfaction (14%); only 9% measured patient preferences. Few studies included more than 10% nonwhite populations or focused on the elderly. No single instrument was used in more than 10% of the studies, and many authors developed de novo tools for use in their study. Methodologic flaws were prevalent and included use of nonstandard economic methods, insufficient power, low or differential response rates, and lack of control for baseline status.

Conclusion: At present, outcomes measurement is variable in its approach and application to breast cancer care. More research is needed to develop practical approaches that are reliable, valid, and feasible in routine practice, that are applicable to diverse populations, and that can inform interventions to improve the quality of care across the full spectrum of services.