Analytical approach:
The study was based on a decision tree, followed by a Markov model, in which patients were stratified by genotype. A lifetime horizon was considered and the authors stated that the perspective of the third-party payer was adopted.
Effectiveness data:
The clinical data were from a selection of relevant studies, including large cohort studies, clinical trials, and national reports. The relationship between the INR and the incidence of bleeds and thromboembolisms was the key clinical input and the data were from a large longitudinal study. The effect of genetic testing on the INR was derived from the COUMAGEN trial, which included 200 patients who were about to be initiated on warfarin. A one-month follow-up was considered.
Monetary benefit and utility valuations:
The utility values were from US population-based European Quality of life (EQ-5D) questionnaire scores.
Measure of benefit:
Quality-adjusted life-years (QALYs) were the summary benefit measure and a 3% annual discount rate was applied.
Cost data:
The economic analysis included the costs of warfarin, genetic test, and events such as transient ischaemic attack, ischaemic stroke, myocardial infarction, extracranial bleed, intracranial haemorrhage, and sequelae. The unit costs were from various sources including the Intermountain Healthcare database, publicly available tests, and the Healthcare Cost and Utilization Project database. All costs were in US dollars ($) and were discounted at an annual rate of 3%. The price year was 2007.
Analysis of uncertainty:
A series of one-way sensitivity analyses was carried out for all the model inputs, using plausible ranges of values, based on published sources or authors’ opinions. A probabilistic analysis, based on a Monte Carlo simulation, was also performed to provide confidence intervals around the model outcomes. Alternative scenarios were considered, with various assumptions on the risk of bleeding and the effect of genetic testing.
