Analytical approach:
This economic evaluation was based on a published Markov model with a 50-year time horizon (until death or a maximum age of 100 years). The model contained eight disease states associated with HRT, such as fractures (hip, vertebral, and wrist) and cancer (breast and colorectal). Two separate analyses were conducted based on uterus status (intact or removed). The authors stated that the perspective was that of the UK National Health Service (NHS).
Effectiveness data:
The effects of HRT on disease risk during therapy were derived from the Women’s Health Initiative, a randomised controlled trial. The UK disease risk and mortality was based on national and in-patient registries and empirical published studies. The authors made some assumptions when literature-based estimates were not available. The clinical outcomes were death and HRT-related adverse events, such as fracture (hip, vertebral, and wrist), cancer (breast and colorectal), and cardiovascular diseases.
Monetary benefit and utility valuations:
The utilities for each treatment-related adverse event were based on several published studies and all except one of these used the European Quality of life (EQ-5D) questionnaire. The one exception was for stroke, where the utility estimates were from a published meta-analysis of outcomes that were assessed by various methods. The utility gain from menopausal symptom relief was measured using the time trade-off method and derived from a Swedish study (Zethraeus, et al. 1997, see 'Other Publications of Related Interest' below for bibliographic details).
Measure of benefit:
Quality-adjusted life-years (QALYs) were the benefit measure and were discounted at an annual rate of 3%.
Cost data:
The analysis considered the annual intervention costs, which included drugs and physician consultations. The short-term (within one year) and long-term (if the event persisted for more than a year) costs associated with different disease states were also included. The data were from published studies, the details of which were not provided, and authors' assumptions. All costs were in UK pounds sterling (£) and the price year was 2006. Future costs were discounted at an annual rate of 3%.
Analysis of uncertainty:
One-way sensitivity analyses tested for the effects of changes in the key parameters, which included cohort ages, treatment duration, and the discount rate. Threshold analysis was conducted to investigate the minimum increase in quality of life required for HRT to become cost-effective.