A cost-effectiveness analysis of hormone replacement therapy in the menopause

Abstract

Objective: To evaluate the cost-effectiveness of hormone replacement therapy in the menopause with particular reference to osteoporotic fracture and myocardial infarction.

Design: The multiple-decrement form of the life table was the mathematical model used to follow women of age 50 through their lifetime under the "no hormone replacement" and "hormone replacement" assumptions. Standard demographic and health economic techniques were used to calculate the corresponding lifetime differences in direct health care costs (net costs in dollars) and health effects ("net effectiveness" in terms of life expectancy and quality, in "quality-adjusted life-years"). This was then expressed as a cost-effectiveness ratio or the cost ($) per quality-adjusted life-year (QALY) for each of the chosen hormone replacement regimens.

Setting and patients: All women of age 50 in New South Wales, Australia (n = 27,021).

Results: The analysis showed that the lifetime net increments in direct medical care costs were largely contributed by hormone drug and consultation costs. Hormone replacement was associated with increased quality-adjusted life expectancy, a large percentage of which was attributed to a relief of menopausal symptoms. Cost-effectiveness ratios ranged from under 10,000 to over a million dollars per QALY. Factors associated with improved cost-effectiveness were prolonged treatment duration, the presence of menopausal symptoms, minimum progestogen side effects (in the case of oestrogen with progestogen regimens), oestrogen use after hysterectomy and the inclusion of cardiac benefits in addition to fracture prevention.

Conclusions: Hormone replacement therapy for symptomatic women is cost-effective when factors that enhance its efficiency are considered. Short-term treatment of asymptomatic women for prevention of osteoporotic fractures and myocardial infarction is an inefficient use of health resources. Cost-effectiveness of hormone replacement in asymptomatic women is dependent on the magnitude of cardiac benefits associated with hormone use and the treatment duration.