DATA CLASSIFICATION
In the map and the bar chart, the data is presented in five discrete classes based on the quantile classification method, where each class contains 20% of the total countries. The quintiles are labeled sequentially from Q1 as the first quintile including the lowest fifth (1% to 20%) of the countries to Q5, the fifth quintile representing the class with the highest fifth (81% to 100%) of the countries.
INDICATOR DEFINITION
Indicator name: Alcohol, total (recorded and unrecorded) alcohol per capita (15 years old or older) consumption in liters of pure alcohol in a given year
Short name: Alcohol per capita consumption (APC)
Data type: Rate (liters per person)
Topic: Risk factors
Rationale: The total alcohol per capita consumption (total APC) comprises both the recorded and the unrecorded APC, which together provide a more accurate estimate of the level of alcohol consumption in a country, and as a result, portray trends of alcohol consumption in a more precise way.
Definition: Total alcohol per capita consumption (APC) is defined as the total (sum of recorded and unrecorded alcohol) amount of alcohol consumed per person (15 years of age or older) over a calendar year, in liters of pure alcohol, adjusted for tourist consumption.
The estimates for the total alcohol consumption are produced by summing the per capita (15+ years of age and older) recorded alcohol consumption and an estimate of per capita (15+) unrecorded alcohol consumption for a calendar year. Tourist consumption takes into account tourists visiting the country and inhabitants visiting other countries.
Method of measurement:
Recorded alcohol consumption refers to alcohol consumed according to the official statistics at the country level based on production, import, export, and sales or taxation data. When government national statistics are not available, country-specific alcohol industry statistics in the public domain based on interviews or fieldwork are used; otherwise, data from the Food and Agriculture Organization of the United Nations' statistical database (FAOSTAT) are used, or data from alcohol industry statistics in the public domain based on desk review.
Unrecorded alcohol consumption refers to alcohol that is not taxed and is outside the usual system of governmental control, such as home or informally produced alcohol (legal or illegal), smuggled alcohol, surrogate alcohol (which is alcohol not intended for human consumption), or alcohol obtained through cross-border shopping (which is recorded in a different jurisdiction). When nationally representative empirical data (which are often general population surveys in countries where alcohol is legal) are not available, specific other empirical investigations, or expert opinion supported by periodic surveys of experts at the country level using modified Delphi-technique, are used.
The liters of alcohol consumed by tourists (15 years of age or older) in a country is based on the number of tourists who visited a country, the average amount of time they spent in the country, and how much these people drink on average in their countries of origin (estimated based on per capita consumption of recorded and unrecorded alcohol). Furthermore, tourist alcohol consumption also accounts for the inhabitants of a country consuming alcohol while visiting other countries (based on the average time spent outside of their country (for all people 15 years or older) and the amount of alcohol consumed in their country of origin). These estimations assume the following: (1) that people drink the same amounts of alcohol when they are tourists as they do in their home countries, and (2) that global tourist consumption is equal to 0 (and thus tourist consumption can be either net negative or positive). Tourist consumption is based on UN statistics, and data are provided by IHME.
Method of estimation:
The total APC in 2018 was calculated from a three-year average of recorded (for 2016, 2017, and 2018) per capita consumption and applying unrecorded proportion (for 2017) and tourist consumption (for 2017). For recorded alcohol consumption, if data did not already exist for 2016, 2017, and/or 2018, the relevant years were projected using a linear regression model employing recorded alcohol per capita consumption data since 2012.
Unrecorded alcohol consumption was estimated as a percentage of total alcohol consumption. Country-level proportions of unrecorded alcohol consumption were estimated using regression analysis. Fractional response random intercepts regression models which accounted for clustering of data points within countries were used to estimate what percentage of total alcohol consumption was due to unrecorded alcohol consumption. Univariate models were fitted for alcohol consumption statistics (the prevalence of drinking, recorded liters of alcohol consumed per capita per year, patterns of drinking scores, value-added and excise taxation of alcoholic beverages, presence of a written national alcohol policy, presence of national legislation to prevent illegal production and/or the sale of the informally produced alcoholic beverages, and alcohol prohibition measures) and other predictors (urbanization, migration rates, malnutrition, sanitation, education levels, and per capita gross domestic product adjusted for purchasing power parity). Backward and forward selection (using a significance cut-off level of 0.2) was used in combination with out-of-sample predictions (multiple random 10% sub-samples) and plausibility checks to assess model fit. Covariate data were obtained from the Institute for Health Metrics and Evaluation and the World Bank.
Tourist consumption of alcohol refers to liters of pure alcohol purchased and consumed by tourists in a given country. This figure is adjusted for the alcohol purchased and consumed when people are visiting countries other than their home country. Positive figures denote the total alcohol consumption of outbound tourists being greater than total alcohol consumption by inbound tourists, negative numbers the opposite.
Unit of Measure: Litres of pure alcohol per person (15 years or older) per year
Indicator name: Alcohol, heavy episodic drinking (population) past 30 days
Short name: Heavy episodic drinking (HED)
Data type: Percent
Topic: Risk factors
Rationale: Heavy episodic drinking is part of the key indicators which provide information regarding the patterns of alcohol consumption in a given country. More specifically, it identifies the proportion of the population which consumes high levels of alcohol on single occasions, and consequently highlights
the population which particularly has a higher risk of experiencing alcohol-related acute harm but also developing chronic health complications.
Definition: Heavy episodic drinking is defined as the proportion of adults (15+ years) who have had at least 60 grams or more of pure alcohol on at least one
occasion in the past 30 days. Consumption of 60 grams of pure alcohol is approximately equivalent to 6 standard alcoholic drinks.
Numerator: The (appropriately weighted) number of respondents (15+ years) who reported drinking 60 grams or more of pure alcohol on at least one occasion in the past 30 days.
Denominator: The total number of participants (15+ years) responding to the corresponding question(s) in the survey plus abstainers.
Disaggregation: Sex, Age
Method of measurement: A representative sample of the adult population (15+ years) of the country is asked to answer questions in a survey. The first priority in the decision tree is given to internationally comparative, nationally representative surveys (in this order of preference: WHS, STEPS, GENACIS, and ECAS).
Method of estimation: Weighted percentages of survey respondents, where abstainers were coded as having 0 occasions.
Method of estimation of global and regional aggregates: Survey estimates weighted by population size of countries.
Preferred data sources: Population-based surveys
Expected frequency of data dissemination: Every 3-5 years
Expected frequency of data collection: Every 3-5 years
Preferred data sources: Internationally comparative, nationally representative population-based surveys.
