Diana B. Petitti

PURPOSE: To critically review the risks and benefits of hormone therapy for asymptomatic postmenopausal women who are considering long-term hormone therapy to prevent disease or to prolong life. DATA SOURCES: Review of the English-language literature since 1970 on the effect of estrogen therapy and estrogen plus progestin therapy on endometrial cancer, breast cancer, coronary heart disease, osteoporosis, and stroke. We used standard meta-analytic statistical methods to pool estimates from studies to determine summary relative risks for these diseases in hormone users and modified lifetable methods to estimate changes in lifetime probability and life expectancy due to use of hormone regimens. RESULTS: There is evidence that estrogen therapy decreases risk for coronary heart disease and for hip fracture, but long-term estrogen therapy increases risk for endometrial cancer and may be associated with a small increase in risk for breast cancer. The increase in endometrial cancer risk can probably be avoided by adding a progestin to the estrogen regimen for women who have a uterus, but the effects of combination hormones on risk for other diseases has not been adequately studied. We present estimates for changes in lifetime probabilities of disease and life expectancy due to hormone therapy in women who have had a hysterectomy; with coronary heart disease; and at increased risk for coronary heart disease, hip fracture, and breast cancer. CONCLUSIONS: Hormone therapy should probably be recommended for women who have had a hysterectomy and for those with coronary heart disease or at high risk for coronary heart disease. For other women, the best course of action is unclear.

OBJECTIVE: To assess the association of unopposed estrogen or estrogen plus progestin and the risk of developing endometrial cancer or dying of that disease. DATA SOURCES: A literature search of English-language studies was performed using MEDLINE, a review of bibliographies, and consultations with experts. METHODS OF STUDY SELECTION: We identified 30 studies with adequate controls and risk estimates. DATA EXTRACTION AND SYNTHESIS: Risk estimates were extracted by two authors and summarized using meta-analytic methods. The summary relative risk (RR) was 2.3 for estrogen users compared to nonusers (95% confidence interval [CI] 2.1-2.5), with a much higher RR associated with prolonged duration of use (RR 9.5 for 10 or more years). The summary RR of endometrial cancer remained elevated 5 or more years after discontinuation of unopposed estrogen therapy (RR 2.3). Interrupting estrogen for 5-7 days per month was not associated with lower risk than daily use. Users of unopposed conjugated estrogen had a greater increase in RR of developing endometrial cancer than users of synthetic estrogens. The risk for endometrial cancer death was elevated among unopposed estrogen users (RR 2.7, 95% CI 0.9-8.0). Among estrogen plus progestin users, cohort studies showed a decreased risk of endometrial cancer (RR 0.4), whereas case-control studies showed a small increase (RR 1.8). CONCLUSIONS: Endometrial cancer risk increases substantially with long duration of unopposed estrogen use, and this increased risk persists for several years after discontinuation of estrogen. Although not statistically significant, the risk of death from endometrial cancer among unopposed estrogen users is increased, similar to the increased risk of developing the disease. Data regarding risk for endometrial cancer among estrogen plus progestin users are limited and conflicting.

This book is an introduction to three methods of quantitative synthesis—meta-analysis, decision analysis, and cost-effectiveness analysis. These methods are used widely to summarize information in order to guide the formulation of clinical recommendations and guidelines, and in clinical decision-making and health policy. The book gives step-by-step instructions on how to conduct studies that use each of the three methods, emphasizing the need for rigor. Important controversies about the statistical and mathematical theories that underlie the methods are highlighted, and key assumptions are identified. The methods are critically appraised and practices that should be avoided are identified. Despite the time that has elapsed between the last revision in 2000, the book remains a relevant and highly accessible source of information on how to conduct studies that use the three methods.

1. Introduction 2. Overview of the Methods 3. Planning the Study 4. Information Retrieval 5. Data Collection 6. Advanced Issues in Meta-Analysis 7. Statistical Methods in Meta-Analysis 8. Other Statistical Issues in Meta-Analysis 9. Complex Decision Problems 10. Estimating Probabilities 11. Utility Analysis 12. Advanced Cost Effectiveness Analysis 13. Utility and Cost-Utility Analysis 14. Exploring Heterogeneity 15. Sensitivity Analysis 16. Reporting Results 17. Limitations