Economic evaluation of a brief counselling for smoking cessation in dentistry: a case study comparing two health economic models

OBJECTIVES: This study aimed to compare the cost-effectiveness estimates of a brief counselling of smoking cessation in dentistry by using two different health economic models. DESIGN AND OUTCOME MEASURES: Intervention effectiveness was estimated in a cluster randomised controlled trial. The number of quitters was estimated based on 7-day abstinence and on smoking reduction at follow-up. Health economic evaluation was performed using two models: (1) a population-based model employing potential impact fractions and (2) a Markov model estimating the cost-effectiveness of the intervention for the actual participants. The evaluation was performed from healthcare and societal perspectives, and health gains were expressed in quality-adjusted life-years (QALYs). SETTING: Dental clinics in Sweden. PARTICIPANTS: 205 Swedish smokers aged 20–75 years. INTERVENTIONS: A brief, structured behavioural intervention was compared with ‘usual care’. RESULTS: The cost per quitter was US$552 in the intervention and US$522 in the ‘usual care’ condition. The net saving estimated with the population-based model was US$17.3 million for intervention and US$49.9 million for ‘usual care’, with health gains of 1428 QALYs and 2369 QALYs, respectively, for the whole Swedish population during 10 years. The intervention was thus dominated by ‘usual care’. The reverse was true when using the Markov model, showing net societal savings of US$71 000 for the intervention and US$57000 for ‘usual care’, with gains of 5.42 QALYs and 4.74 QALYs, respectively, for lifelong quitters. CONCLUSION: The comparison of intervention and ‘usual care’ derived from small-scale studies may be highly sensitive to the choice of the model used to calculate cost-effectiveness. TRIAL REGISTRATION: The cluster randomised trial is registered in the ISRCTN register of controlled trials with identification number ISRCTN50627997.