Analytical approach:
The analysis was based on a decision-tree model, with a lifetime horizon, for a hypothetical cohort of 65-year-old patients presenting with intracranial, gastrointestinal, or retroperitoneal haemorrhage. The authors stated that the analysis was carried out from the perspective of the UK NHS.
Effectiveness data:
A systematic literature search was undertaken to identify the relevant sources of evidence, in MEDLINE and the Cochrane Library, for articles from 1998 to 2008. The search criteria were reported. Clinical trials, retrospective cohort studies, and observational studies were selected. The mortality estimates were the key inputs for the model and they were from 16 studies for intracranial haemorrhage, three studies for gastrointestinal haemorrhage, and two studies for retroperitoneal haemorrhage. Average values were calculated without weighting by sample size, which some studies did not report.
Monetary benefit and utility valuations:
The utility values were from the literature.
Measure of benefit:
Life-years and quality-adjusted life-years (QALYs) were the summary benefit measures.
Cost data:
The economic analysis included the costs of hospital admission, drugs and other therapies, imaging studies and other diagnostic tests, ambulance transport, and stroke rehabilitation. The resource use was estimated by a panel of nine consultant physicians (six haematologists and three anaesthetists), with relevant clinical experience of warfarin reversal, plus the opinions of two of the authors. A structured questionnaire was used. The costs were estimated from official NHS prices. They were expressed in UK pounds sterling (£) at 2007 to 2008 prices.
Analysis of uncertainty:
A probabilistic sensitivity analysis was carried out, using Monte Carlo simulation. Conventional probability distributions were assigned to the model inputs and cost-effectiveness acceptability curves were generated. A deterministic sensitivity analysis was conducted to identify those inputs that affected the cost-effectiveness decision.