Analytical approach:
The analysis was based on a decision-analytic model, for the short-term (screening, assessment, and treatment), followed by a Markov model, with six alcohol-related health states, for a lifetime horizon. The authors stated that they adapted a societal perspective.
Effectiveness data:
The effectiveness data were derived from published studies identified through MEDLINE search, expert opinion, and relevant documents. Only bibliographical details were given for the sources of effectiveness data. The key clinical endpoints were the transitions between alcohol-related states (non-drinker, safe drinker, at-risk drinker, alcohol abuser, alcohol dependent, and alcohol dependent in recovery), with each scenario and age. The transition rate from the non-drinker state and the transition rate into the dependent state were estimated by adjusting the other transition rates until the proportions entering those states were consistent with published data.
Monetary benefit and utility valuations:
The utility values for alcohol-related health states were derived from a published study, which used the standard gamble method to elicit the preferences. The utility values for non-drinkers and safe drinkers were based on published estimates.
Measure of benefit:
Quality-adjusted life-years (QALYs) and life-years gained were the summary measures of benefit. Future QALYs and life-years were discounted at an annual rate of 3%.
Cost data:
The cost categories were physician time, %CDT laboratory testing, venipuncture, travel paid by patients, and hourly wages lost for patients. The sources of resource use were published evidence and authors' assumptions. The sources of cost data were Medicare reimbursement, the Medical Expenditure Panel Survey, and the Bureau of Labor Statistics. All costs were expressed in US dollars ($). The price year was 2006 and a 3% annual discount rate was applied.
Analysis of uncertainty:
One-way, two-way, and probabilistic sensitivity analyses were performed.