The analysis was based on one clinical study for each intervention to determine the intermediate clinical effect. The intermediate clinical effect was linked to the incidence of several diseases through risk equations. The DisMod II model was used to estimate outcomes, over a lifetime, using differential equations, without accounting for the time dependence of variables. The authors stated that the perspective was that of the health service.
The main clinical data were the treatment effect on body mass index (BMI) for both interventions and on physical activity and fruit and vegetable consumption for the Lighten Up intervention. Weight loss, physical activity, plus fruit and vegetable consumption were used as variables to modify the relative risk of ischaemic heart disease, ischaemic stroke, type 2 diabetes, breast cancer, and colon cancer. For the Lighten Up intervention, the clinical data came from the Lighten Up programme, which was implemented in 20 Health Districts across Queensland, Australia. There was no control group, so a background trend in BMI during the intervention and follow-up period was estimated from national surveys. A survey of study participants was conducted at two months and 12 months. At two months, 238 participants were surveyed; at 12 months, 23 participants were surveyed. The effectiveness of the Weight Watchers intervention came from the only independently sponsored randomised controlled trial (RCT) of Weight Watchers. This was conducted in the UK and was a six-month programme. The six-month data were assumed to apply to 12 months. A published meta-regression study of 46 dietary and dietary and exercise counselling trials indicated that weight loss was regained after five and a half years, so the intervention effect was modelled to have an exponential pattern of decay so that it was virtually absent at five and a half years. The risk equations for diseases based on the intermediate outcomes were based on Australia's burden of disease and injury, and various published studies.
Monetary benefit and utility valuations:
Disability weights for the calculation of disability-adjusted life-years (DALYs) were obtained for the different diseases included in the model.
Measure of benefit:
The measure of benefit was the number of DALYs averted. Future benefits were discounted at 3% annually.
The costs included those for providing the weight loss programmes and treating the different diseases included in the model. The costs of Lighten Up included project coordination, delivering the intervention, and the cost to each patient for time and travel in attendance. The cost of Weight Watchers included the cost of primary care referral and attendance vouchers. The costs for diseases averted were evaluated using data from the Australian Institute of Health and Welfare Disease Costs and Impacts Study 2001. Costs were adjusted to year 2003 using the Australian Health Price Index. Future costs were discounted at 3% annually. The currency was Australian dollars (AUD).
Analysis of uncertainty:
A probabilistic sensitivity analysis was conducted where each parameter was assigned probability distributions and values sampled simultaneously thousands of times. Sensitivity analyses were conducted modifying decay rates (a potential loss of effect up to 12 months), intervention delivery costs, and increasing Lighten Up recruitment. Probabilistic sensitivity analyses were presented on cost-effectiveness plane scatter plots, and with cost-effectiveness acceptability curves.