A Bayesian approach to analysing the cost-effectiveness of two primary care interventions aimed at improving...

To assess the cost-effectiveness of two primary care interventions, a letter and a flag, aimed at improving attendance for breast screening among (i) all women invited for breast screening and (ii) non-attenders.

A probabilistic decision analytic model was developed using Markov chain Monte Carlo simulation implemented in WinBUGS. The model was populated using economic and effectiveness data collected alongside two randomised controlled trials.

For all women invited, the incremental cost-effectiveness ratio (ICER) for the letter compared with no intervention is 27 pounds per additional attendance, and the ICER for the combined letter and flag intervention compared to the letter alone is 171 pounds. The corresponding ICERs for non-attenders are 41 pounds and 90 pounds. The flag intervention is an inefficient option in both settings. A large proportion of the costs fall on the practices (25-67%), depending on the intervention and target population. The total costs incurred do not, however, seem prohibitive. Expected value of perfect information suggests that there is greater value in carrying out further research on the intervention implemented among all women invited for breast screening rather than on non-attenders.

The flag intervention alone does not appear to be an efficient option. The choice between the letter and both interventions combined is subjective, depending on the willingness to pay for an additional screening attendance.