[Application of the cigarette purchase tasks in achieving China' s tobacco control goals]

To assess cigarette demand among Chinese smokers through a cigarette purchase task (CPT) and to evaluate cigarette prices under different hypothetical scenarios in order to meet the goals of smoking prevalence reduction in China.

In the study, 447 participants completed a hypothetical CPT at baseline assessments of a trial, thus, cigarette demand curves were individually fitted for each participant using an exponentiated version of the exponential demand model. Typically, five demand indices were derived, intensity (consumption when free), breakpoint (first price at which consumption is suppressed to 0), maximum output (Omax), maximum price (Pmax, price at which Omax occurred), and elasticity (the ratio of the change in quantity demanded to the change in price). A one-way analysis of variance was used to explore the correlations between the cigarette purchase task indices and socio-demographic and smoking characteristics. The one-way decay model was employed to simulate the smoking cessation rates and determine optimal cigarette prices in a series of scenarios for achieving 20% smoking prevalence.

The price elasticity drawn from CPT was 0.54, indicating that a 10% price increase could reduce smoking by 5.4% in the participated smokers. Smokers with higher income were less sensitive to cigarette prices (elasticity=-2.31, P=0.028). Cigarette purchase task indices varied significantly among the smokers with different prices of commonly used cigarettes, tobacco dependence, and smoking volume. The smokers who consumed cigarettes of higher prices reported higher breakpoint, Omax and Pmax, but lower intensity (P=0.001). The smokers who were moderately or highly nicotine dependent reported higher intensity, breakpoint, Omax and Pmax, and they had lower intensity (P=0.001). The smokers who had a higher volume of cigarettes reported higher intensity and Omax, and lower intensity (P < 0.001). To achieve the goal of reducing smoking prevalence to 20% in mainland China, we estimated the desired increase on smoking cessation rate and prices accordingly in a series of scenarios, considering the gender variance and reduced smoking initiation. In scenario (a), to achieve a smoking prevalence goal of 20%, it would be necessary for 24.81% of the current smokers to quit smoking when there were no new smokers. Our fitting model yielded a corresponding value of 59.64 yuan (95%CI 53.13-67.24). Given the assumption in scenario (b) that only males quitted smoking, the desired cessation rates would be 25.82%, with a higher corresponding price of 62.15 yuan (95%CI 55.40-70.06) to induce desired cessation rates. In the proposed scenario (c) where 40 percent of the reduction in smoking prevalence came from reduced smoking initiation, and females and males equally quitted smoking due to increased cigarette prices, the price of a pack of cigarettes would be at least 37.36 yuan (95%CI 32.32-42.69) (equals to $ 5.20) per pack to achieve the cessation rate of 14.89 percent. In scenario (d) where only males quitted smoking due to increased cigarette prices considering the reduced smoking initiation, the respective smoking cessation rates should be 15.49% with the desired prices of 38.60 yuan (95%CI 33.53-44.02). After adjusting for education levels and income levels in scenario (c), the price of cigarettes would be at least 37.37 yuan/pack (equals to $ 5.20) (95%CI 30.73-44.94) and 37.84 yuan/pack (equals to $ 5.26) (95%CI 31.94-44.53), respectively.

Cigarette purchase task indices are significantly associated with income levels and prices of commonly used cigarettes, levels of tobacco dependence, and smoking volume, which is inspiring in studying price factors that influence smoking behavior. It is suggested that higher cigarette prices, surpassing the current actual market level, is imperative in mainland China. Stronger policy stra-tegies should be taken to increase tobacco taxes and retail cigarette prices to achieve the Healthy China 2030 goal of reducing smoking prevalence to 20%.