Assessment of Medical Intervention in the Intention to Quit Tobacco in Uruguay, Argentina y Brazil: Tobacco Control Policy Effects

Special Article – Tobacco and Smoking Cessation

J Fam Med. 2016; 3(3): 1060.

Assessment of Medical Intervention in the Intention to Quit Tobacco in Uruguay, Argentina y Brazil: Tobacco Control Policy Effects

Curti D* and Bianco E

Centro de Investigación Para la Epidemia del Tabaquismo (CIET), 25 de Mayo 469, Montevideo, Uruguay

*Corresponding author: Curti D, Centro de Investigación Para la Epidemia el Tabaquismo, 25 de Mayo 469, Montevideo, Uruguay

Received: April 25, 2016; Accepted: May 30, 2016; Published: June 02, 2016


Purpose: This work is aimed at evaluating the effectiveness of medical intervention in encouraging tobacco quitting, controlling for the characteristics of those who try to quit.

Methods: Using data from the Global Adult Tobacco Survey (GATS) in Argentina (2012), Brazil (2008) and Uruguay (2009), logit models are estimated taking into account the characteristics of smokers and medical intervention in consultation as explanatory variables of motivation to quit tobacco.

Results: In Uruguay and Brazil controlling for smoker’s characteristics, like age, gender, education, type of employment, and smoking frequency, those that have consulted a doctor and within this group who have received brief medical advice are more likely to try quitting. In Uruguay and Argentina we find that people under 45 years have a higher rate of intention to quit than average, although they receive a smaller proportion of brief medical advice than other age groups.

Conclusions: Most intention to quit in Uruguay and Argentina in groups of young people cannot relate to medical advice, so it is postulated that other tobacco control policies such as health warnings on cigarettes packs, smoke free areas, to which in the case of Uruguay tobacco tax increases and the ban on advertising, promotion and sponsorship of tobacco products are added. The cost-effectiveness of tobacco cessation programs tobacco could improve if the proportion of people receiving brief medical advice in the younger age tranches is increased and are designed comprehensive programs that include the management of other risks.

Keywords: Tobacco cessation; Brief medical advice; Cost-effectiveness; Tobacco use disparities; Medical intervention effectiveness


GATS: Global Adults Tobacco Survey; ANOVA: Analysis of Variance; OR: Odds ratio


The consumption of tobacco is the most important cause of preventable deaths worldwide [1], and the prospects are that the epidemic is concentrated in developing countries, because the tobacco industry has focused its efforts to increase sales in them because they have a huge population with rising incomes, a high percentage of young people, and fewer regulations than developed countries. The World Health Organization estimates that by 2030 will die 8 million people annually due to consumption of tobacco, and 80% of them will occur in developing countries [2].

The literature reviewed shows that when taking into account the prevalence by education level, income, occupation and place of residence, tobacco consumption and related diseases affect disproportionately to people of low socioeconomic strata [3-5].

In Montevideo there are neighbourhoods, where live people of low socio-economic strata, whose prevalence is more than double, compared to other districts of the city [6].

A factor that may contribute to the tobacco consumption disparities is that quitting rates are different according to socioeconomic strata. Studies that have examined quitting rates in relation to smoker´ssocio-economic strata determined that the lower income or educational level less likely the smoker will succeed in quitting tobacco, and vice versa, that the higher education or income of people, the higher probability of success [7-9]. When a group of people has a higher prevalence of tobacco than average, interventions targeting these groups could reduce prevalence disparities [10-14].

Regarding intention to quit tobacco, there is no conclusive evidence about its relationship with the socio-economic strata. Studies in Australia, Canada and the United Kingdom [15] indicate that the higher the level of education or income more likely people will try to quit, but other works, also conducted in these countries and in the United States, not found such relationship [16,17].

The motivation to quit is an essential aspect for smoking cessation programs to achieve greater impact in reducing prevalence. Simple interventions that increase smoker´s motivation to quit could improve the cost-effectiveness of cessation programs.

This paper aims to evaluate the effectiveness of interventions that encourage smokers to quit tobacco, such as medical consultation and brief advice, controlling for socioeconomic and personal characteristics of those who intent to quit tobacco. Also it examines whether the intention of quitting is different depending on age tranches and if this is related to brief medical advice.

Materials and Methods

In Uruguay, Argentina y Brazil GATS is a national survey with urban and rural strata coverage. GATS was conducted as a household survey of people aged 15 years and older by the National Institute of Statistics of each country. In Uruguay an initial sample of 6558 households, completing 5591 individual interviews; in Argentina a sample of 9790 households carried out 6645 individual interviews and in Brazil the sample of 51011 households includes 39425 interviews [18-20].

A multi-stage cluster stratified sampling, designed to produce nationally representative data was used in all three cases. An individual was identified in each household randomly selected to participate in the survey. The response rate in households was 79.2%, 97.0% and 95.0%, the individual response rate was 93.8%, 98.5% and 98.9%, and the average response rate was 74.3%, 95.6 % and 94.0% in Argentina, Uruguay and Brazil respectively.

The dependent variable explained by the logit models is:

Intention to quit: Intent to quit smoking in the last 12 months, dichotomous categorical variable (0 do not try to quit, 1 try to quit).

The independent variables of logit models to estimate are [21-30]:

Age: Five dichotomous categorical variables are used, one for each tranche age, is 1 in each age tranche and 0 otherwise; the age tranches are up to 25 years old, between 26 y 35 years old, 36 and 45 years old, 46 and 59 years old, and 60 or more years old respectively.

Education: Three dichotomous categorical variables are used (Argentina and Uruguay), one for each education level, is 1 in the educational level and 0 otherwise; the categories are up to primary school, up to high school/technical school, up to university/high school teacher/technical school teacher respectively. In the case of Brazil four categorical variables are used, an additional variable is used for those without formal education.

Visit to the doctor: Visit to the doctor in the last 12 months. Dichotomous categorical variable, (0 do not visit to the doctor, 1 visit to the doctor).

Receive brief medical advice: Receive medical advice to quit, dichotomous categorical variable (0 do not receive brief medical advice, 1 receive brief medical advice).

Smoking frequency: Daily smoker or occasional smoker, dichotomous categorical variable (0 daily smokers, 1 occasional smoker).

Gender: Man or woman, dichotomous categorical variable (0 man, 1 woman).

Employment: Four categorical variables are used, one for each type of employment; the four types of employment are: government employee, private employee/self-employed, student (this category just to Uruguay and Argentina), housewife/retired/unemployed.

On the one hand, logit models are estimated to determine which personal characteristics and socio-economic have those who have tried quitting, and on the other hand, if medical interventions (medical consulting and brief medical advice) have an effect on those who have tried to quit tobacco. An analysis of variance (not reported) was performed to know if there are statistically significant differences among explanatory variables. After the ANOVA, binary logistic regression and odds ratios (relative risks) are estimated.

Results and Discussion

For each country the first logit is estimated using as an independent variable visit to the doctor in the last 12 months; the results are presented in Table 1.