Prevalence of current tobacco use among adults aged ≥ 15 years (%)

Percent

The prevalence of current tobacco smoking among adults is an important measure of the health and economic burden of tobacco, and provides a baseline for evaluating the effectiveness of tobacco control programmes over time. While a more general measure of tobacco use (including both smoked and smokeless products) would be ideal, data limitations restrict the present indicator to smoked tobacco. Adjusted and age-standardized prevalence rates are constructed solely for the purpose of comparing tobacco use prevalence estimates across multiple countries or across multiple time periods for the same country. These rates should not be used to estimate the number of smokers in the population.

Current smoking of any tobacco product prevalence estimates, resulting from the latest adult tobacco use survey (or survey which asks tobacco use questions), which have been adjusted according to the WHO regression method for standardising described in the Method of Estimation below.  "Tobacco smoking" includes cigarettes, cigars, pipes or any other smoked tobacco products. “Current smoking” includes both daily and non-daily or occasional smoking.

Sex

Outcome

WHO has developed a regression method that attempts to enable comparisons between countries. If data are partly missing or are incomplete for a country, the regression technique uses data available for the region in which the country is located to generate estimates for that country. The regression models are run at the United Nations sub-regional level 3 separately for males and females in order to obtain age-specific prevalence rates for that region. These estimates are then substituted for the country falling within the sub-region for the missing indicator. Note that the technique cannot be used for countries without any data: these countries are excluded from any analysis. Information from heterogeneous sources that originate from different surveys and do not employ standardized survey instruments render difficult the production of national-level age-standardized rates. The four types of differences between surveys and the relevant adjustment procedures used are listed below. Differences in age groups covered by the survey: In order to estimate smoking prevalence rates for standard age ranges (by five-year groups from age 15 until age 80 and thereafter from 80 to 100 years), the association between age and daily smoking is examined for males and females separately for each country using scatter plots. For this exercise, data from the latest nationally representative survey are chosen; in some cases more than one survey is chosen if male and female prevalence rates stem from different surveys or if the additional survey supplements data for the extreme age intervals. To obtain age-specific prevalence rates for five-year age intervals, regression models using daily smoking prevalence estimates from a first order, second order and third order function of age are graphed against the scatter plot and the best fitting curve is chosen. For the remaining indicators, a combination of methods is applied: regression models are run at the sub-regional level to obtain age-specific rates for current and daily cigarette smoking, and an equivalence relationship is applied between smoking prevalence rates and cigarette smoking where cigarette smoking is dominant to obtain age-specific prevalence rates for current and daily cigarette smoking for the standard age intervals. Differences in the types of indicators of tobacco use measured: If we have data for current tobacco smoking and current cigarette smoking, then definitional adjustments are made to account for the missing daily tobacco smoking and daily cigarette smoking. Likewise, if we have data for current and daily tobacco smoking only, then tobacco type adjustments are made across tobacco types to generate estimates for current and daily cigarette smoking. Differences in geographic coverage of the survey within the country: Adjustments are made to the data by observing the prevalence relationship between urban and rural areas in countries falling within the relevant sub-region. Results from this urban-rural regression exercise are applied to countries to allow a scaling-up of prevalence to the national level. As an example, if a country has prevalence rates for daily smoking of tobacco in urban areas only, the regression results from the rural-urban smoking relationship are used to obtain rural prevalence rates for daily smoking. These are then combined with urban prevalence rates using urban-rural population ratios as weights to generate a national prevalence estimate as well as national age-specific rates. Differences in survey year: For the WHO Report on the Global Tobacco Epidemic, 2009, smoking prevalence estimates were generated for year 2006. Smoking prevalence data are sourced from surveys conducted in countries in different years. In some cases, the latest available prevalence data came from surveys before the year 2006 while in other cases the survey was later than 2006. To obtain smoking prevalence estimates for 2006, trend information is used either to project into the future for countries with data older than 2006 or to backtrack for countries with data later than 2006. This is achieved by incorporating trend information from all available surveys for each country. For countries without historical data, trend information from the respective sub-region in which they fall is used. Age-standardized prevalence: Tobacco use generally varies widely by sex and across age groups. Although the crude prevalence rate is reasonably easy to understand for a country at one point in time, comparing crude rates between two or more countries at one point in time, or of one country at different points in time, can be misleading if the two populations being compared have significantly different age distributions or differences in tobacco use by sex. The method of age-standardization is commonly used to overcome this problem and allows for meaningful comparison of prevalence between countries. The method involves applying the age-specific rates by sex in each population to one standard population. The WHO Standard Population, a fictitious population whose age distribution was artificially created and is largely reflective of the population age structure of low- and middle-income countries, was used. The resulting age-standardized rate, also expressed as a percentage of the total population, refers to the number of smokers per 100 WHO Standard Population. As a result, the rate generated using this process is only a hypothetical number with no inherent meaning in its magnitude. It is only useful when contrasting rates obtained from one country to those obtained in another country, or from the same country at a different points in time. In order to produce an overall smoking prevalence rate for a country, the age-standardized prevalence rates for males and females must be combined to generate total prevalence. Since the WHO Standard Population is the same irrespective of sex, the age-standardized rates for males and females are combined using population weights for males and for females at the global level from the UN population data for 2006. For example, if the age-standardized prevalence rate for tobacco smoking in adults is 60% for males and 30% for females, the combined prevalence rate for tobacco smoking in all adults is calculated as 60 x (0.51) + 30 x (0.49) = 45%, with the figures in brackets representing male and female population weights. Thus, of the total smoking prevalence (45%) the proportion of smoking attributable to males is 66.7% [= (30 ÷ 45) x 100] and to females 33.3% [= (15 ÷ 45) x 100]. Predominant type of statistics: adjusted

Regional and global aggregates are based on population-weighted averages weighted by the total number of population aged ≥15 years. They are presented only if available data cover at least 50% of total population aged ≥15 years in the regional or global groupings.

N/A

Continuous

Developing standard methods for adjusting and reporting the prevalence of tobacco use represents our best effort for developing a baseline with which to compare future prevalence estimates of tobacco use. The ideal would be to have national government agreement on a standard framework for collecting survey data on chronic disease risk factors, including tobacco use, within a common timeframe. As this may take a little time, these estimates are intended to be the baseline for tobacco control efforts worldwide.