A life-time Markov decision-analysis model for a hypothetical cohort of 1000 patients with NVAF was developed.  We compared 2 doses of dabigatran (110mg BD &150mg BD) to warfarin titrated to target prothrombin time. We considered ischemic stroke, intracranial bleed, other major bleeds and death as important outcomes of NVAF. Model assumptions and transition probabilities were based on literature and costs on market prices. Data on out-of-pocket expenses and income lost was taken from a nationally representative household survey. The main model outcomes were life-years and quality-adjusted-life-year (QALY) gained. We adopted a societal perspective; and discounted outcomes and costs at 3%. Incremental cost-effectiveness ratios (ICER) were estimated.  We used a cost-effectiveness threshold of USD 1889 corresponding to GDP/capita. Sensitivity and threshold analyses were conducted.

Dabigatran was associated with gain in life-years  (0.1 for 110mg vs 0.2 years for 150mg) and QALYs  (0.12 for 110mg; 0.17 for 150mg). The discounted ICER/QALY for dabigatran 110mg (USD 7,519) and 150mg (USD 6,634) were both above the cost-effectiveness threshold and on probabilistic sensitivity analysis the probability of being cost-effective was low (13% & 8%  for 110mg & 150mg respectively).  Threshold analysis showed that price reduction of dabigatran 150mg by 60% will make it cost-effective, and to 80%  will make it cost-neutral.