Background: While a general consensus exists that postmenopausal women with osteoporosis (bone mineral density [BMD] T-score < -2.5) should receive treatment to reduce the risk of fracture, the appropriate management of the larger number of women with osteopenia (BMD T-score between -1.0 and -2.5) remains unclear. While a recent study of the bisphosphonate alendronate found that this agent was not cost-effective for fracture prevention in osteopenic postmenopausal women, the cost-effectiveness of raloxifene in this population has not yet been examined.

Methods: We developed two state-transition network simulation models, each with a 40-year time-horizon, to evaluate three treatment approaches to the patient with postmenopausal osteopenia: expectant management, alendronate and raloxifene. The first model included only primary clinical events - i.e., four types of fractures (hip, vertebral, wrist, and other) and death while the second further incorporated additional outcomes (i.e., breast cancer and venous thromboembolism [VTE]) whose risks may be influenced by raloxifene. Patients were assumed to remain on therapy for 5 years and to derive ongoing but diminishing benefit during the 5 years following treatment discontinuation. Transition probabilities and estimated direct medical costs were obtained from the published literature. For each model, Monte Carlo simulations (100,000 iterations) were conducted to produce distributions for expected costs, effectiveness (in QALYs), and incremental cost-effectiveness ratio (ICER) for each strategy. The base case consists of women 55-64 with osteopenia. Benefits and cost were discounted at 3%.

Results: For the model including only primary fracture events, alendronate was the preferred strategy with an ICER of ~$127,000/QALY compared to expectant management. However, raloxifene strongly dominated alendronate in the model that included other treatment-related events, with an ICER of ~$62,500/QALY. While alendronate and raloxifene yielded similar reductions in fracture incidence (~10/10,000 person-years), raloxifene produced the greatest overall benefit given its 17.7% reduction in breast cancer incidence (~8/10,000 person-years), even with an increase in incidence of VTE (~2/10,000 person-years). Results were most sensitive to the population risks of fracture and breast cancer.

Conclusion: When taking into account all relevant disease and treatment-related outcomes, raloxifene may be a cost-effective treatment option for postmenopausal women with osteopenia, particularly if they are at increased risk of breast cancer.