Effect and offset of effect of treatments for hip fracture on health outcomes.

We investigated the cost-effectiveness of treatments that reduce the risk of hip fracture using a computer simulation model. Cost-effectiveness was measured as cost per quality-adjusted life-year (QALY) gained using a threshold value for cost-effectiveness of $30,000/QALY gained. The baseline simulations assumed a 5-year intervention that reduced the risk of hip fracture by 50% during the intervention period, and an effect which reversed to the pretreatment risk during the next 5 years. Sensitivity analyses included the effects of age, different fracture risks, and different treatment costs and duration of therapeutic effect once treatment was stopped. Cost-effectiveness was critically dependent upon absolute risk determined by the age and the relative risk of hip fracture at any given age. Reasonable cost-effectiveness was shown even with relatively high intervention costs for women with a risk about twice the average at the age of 70 or more years. Cost-effectiveness was critically dependent upon the assumptions made concerning offset of effect of intervention after the end of treatment. Where no residual effect was assumed, it was difficult to show cost-effectiveness from any intervention except for the most effective and least expensive. Conversely, cost-effectiveness improved considerably where effectiveness persisted for a longer time. These studies support the view that intervention in the elderly with agents affecting skeletal metabolism alone may be preferred to such interventions at the time of the menopause, and that offset time, hitherto poorly characterized, is a critical component of cost-effectiveness, particularly in younger women.