Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

In 2006, the smoking prevalence in Switzerland was 29%, i.e. one of the highest among all western European countries, far behind Ireland or Sweden whose prevalence rates lie under 20%. More than half of Swiss smokers had the intention to quit in 2006, but only 11% within a one month period. A large number of smokers try to quit each year, however the success rate is low. The estimated 'cold-turkey' quit rate at one year is 5%. With the best treatment, it is around 30%.

Only a small number of smokers seek help despite the fact that many smoking cessation interventions exist. These can be pharmaceutical or non-pharmaceutical (medical counselling, group therapies, help line, exercises, acupuncture, etc.). The pharmaceutical interventions are divided into two main groups: the nicotine replacement therapies (NRTs) that include patches, gums, nasal sprays; and the medications like bupropion (brand name: Zyban) and, since 2006, varenicline (brand name: Chantix). These treatments are proven to be effective vs. placebo but the former is not reimbursed by Swiss health insurances and the latter is not even available on the Swiss market. Economic studies about this kind of treatments have largely focused on their cost-effectiveness but rarely on patients' valuation of their characteristics.

In this study we used Discrete Choice Experiments (DCE) to assess smokers' preferences for hypothetical cessation medications. The treatments were described trough six attributes - price, efficacy, benign side-effects, availability, duration of the treatment and effect on weight gain - derived after a systematic review of the literature and experts consultation. Each choice set presented to respondents contained two alternatives and an opt-out option. A fractional factorial design, which was divided into two blocks of 8 choice tasks, was used. In each block, constructed scenarios were added to assess respondent's rationality. The following information was also gathered about individuals' characteristics: smoking history, level of dependence, cessation history, physical and psychological health status, weight concern, income and demographics.

Respondents were recruited in the French-speaking part of Switzerland. Data about 250 individuals was collected and used to assess the relative importance of the different attributes and the trade-offs respondents were willing to make between them. The inclusion of a price attribute allowed us to estimate the value of changes in individual attributes (marginal willingness-to-pay) and of combinations of attributes (WTP for the entire medication). The influence of respondents' characteristics on preferences was assessed in conducting subgroup analysis and in estimating interaction effects with attributes levels.