Analytical approach:
The analysis was based on a Markov model, with a hypothetical cohort of women with a previous diagnosis of idiopathic heavy menstrual bleeding. Two scenarios were considered: a contraception scenario, where patients do not wish to have children and contraception was included, and a dysfunctional uterine bleeding scenario, which focused on the control of bleeding only. A five-year time horizon was considered. The authors stated that the analysis was conducted from the perspective of the Spanish National Health System.
Effectiveness data:
The clinical sources of evidence were identified by a review of the literature in commonly used electronic databases. These inputs were validated by Spanish clinical experts. The key model input was the treatment success, which was defined as the proportion of patients responding to treatment and not becoming pregnant.
Monetary benefit and utility valuations:
The utility values were from published studies that used the European Quality of life (EQ-5D) scale. UK tariffs were used, as there were no Spanish values.
Measure of benefit:
Three benefit measures were considered: symptom-free months, surgery-free months, and quality-adjusted life-months (QALMs). A 3% annual discount rate was applied.
Cost data:
The economic analysis included the costs of the LNG-IUS (device and gynaecology visit), the combined oral contraceptive (only gynaecology visit), progestogens (gynaecology visit and medications), resection, hysterectomy, abortion, and pregnancy and birth. The cost of the combined oral contraceptive medication was zero to the health system because it was paid by patients. The resource use was from a panel of Spanish experts with experience in the management of dysfunctional uterine bleeding. The costs were estimated using the eSalud database, which collected costs from the literature, and using tariffs from national and regional health services. They were in Euros (EUR) and were discounted at an annual rate of 3%. The price year was 2008.
Analysis of uncertainty:
A probabilistic sensitivity analysis was undertaken, using a second-order Monte Carlo simulation with 10,000 iterations. Log-normal distributions were selected for costs, normal distributions for resource use, and beta-distributions for probabilities. A one-way sensitivity analysis was carried out, by varying the model inputs one at a time. The ranges of values were based on authors’ opinions or published sources.