Analytical approach:
A state-transition model was used to synthesise the data for the first two years and then for the subsequent years. A lifetime horizon was used. The authors stated they took both a health care and a societal perspective.
Effectiveness data:
The natural history model parameters were mainly based on the Swiss HIV Cohort Study, which had a cohort of over 14,000 people, with a mean age of 33 years and 80% men (Ledergerber, et al. 1994, see ‘Other Publications of Related Interest’ below for bibliographic details). The relative effectiveness of GART compared with expert opinion was derived from published randomised trials.
Monetary benefit and utility valuations:
The utility values were derived from a study on the quality of life of patients enrolled in the Swiss HIV Cohort Study (Zinkernagel, et al. 1999, see ‘Other Publications of Related Interest’ below for bibliographic details). Visual analogue scale data were transformed to standard gamble utilities, and regression analysis was used to derive the utilities for the different health states. The disutilities of experiencing a disease indicating acquired immune deficiency syndrome were derived from a published meta-analysis. An adjustment was made to exclude any income effects of HIV disease in the QALY estimation, since productivity losses were included in the cost analysis.
Measure of benefit:
Life-years and quality-adjusted life-years (QALYs) were the measures of benefit. These were discounted at different rates according to the health care (4%) or societal (2%) perspective.
Cost data:
For the health care perspective, only those costs due to health care resource consumption were included, using previously reported micro-costing data. These included antiretroviral therapy, drugs to treat or prevent opportunistic diseases and other drugs, ambulatory costs, in-patient costs, and GART costs. For the societal perspective, the number of hours a patient worked in each health state was valued at an average Swiss wage rate. All costs were expressed in 2005 US dollars ($). Benefits were discounted at different rates according to the health care (4%) or the societal (2%) perspective.
Analysis of uncertainty:
Probabilistic sensitivity analysis was used to evaluate the uncertainty around the cost and effect estimates and the methods were adequately reported. The joint distribution of the resulting incremental costs and effects was summarised in graphs of cost-effectiveness 95% credibility intervals and acceptability curves.