Analytical approach:
A model that combined the 12 month cessation rates, relapse rates, lifetime QALYs gained by lifetime cessation, and costs was developed and analysed using Bayesian synthesis methods in WinBUGs. The time horizon was lifetime and the authors reported that a National Health Service (NHS) perspective was used.
Effectiveness data:
The effectiveness parameters came from a non-systematic literature review, and relied heavily on a previous economic model (Woolacott, et al. 2002, see 'Other Publications of Related Interest' below for bibliographic details), which included a systematic review of clinical effectiveness and three studies that investigated the effect of genotype on cessation rates. The main effectiveness estimates included the 12 month cessation rates of the different strategies, as well as their modifications according to different genetic sub-populations, and relapse rates.
Monetary benefit and utility valuations:
Quality-adjusted life-years (QALYs) gained from lifetime smoking cessation were extracted from Woolacott, et al. 2002.
Measure of benefit:
The incremental net benefits were estimated, for different willingness-to-pay per QALY thresholds, and were not discounted.
Cost data:
The cost categories included those of the different treatment strategies, as well as the genetic testing. The costs of treatment were taken from Woolacott, et al. 2002, but the sources for the costs of the genetic testing were not cited. The currency was UK pounds sterling (£) and the price year was 2001. All costs were short term, so discounting was not necessary.
Analysis of uncertainty:
Probabilistic sensitivity analysis was undertaken to take account of multi-parameter uncertainty. The authors stated that the cost of treatment was not influential and so the average costs were used and they were not analysed for uncertainty. The test costs were varied in a scenario analysis.