—from Epidemiological Bulletin, Vol. 22 No. 2, June 2001—

The Global Youth Tobacco Survey:
Status of its implementation in Latin America and the Caribbean

Introduction
The Global Youth Tobacco Survey (GYTS) is aimed at assessing tobacco use, self-reported exposure to environmental tobacco smoke, knowledge and attitudes, and the factors that make youth susceptible to tobacco use. Its results will enhance the capacity of countries to design, implement and evaluate comprehensive tobacco control programs. The GYTS was developed by the World Health Organization (WHO) through its Tobacco Free Initiative (TFI) in collaboration with the Office on Smoking and Health in the US Centers for Disease Control and Prevention (OSH/CDC) and in consultation with countries in the six WHO regions. It has been implemented in Latin American and English-speaking Caribbean countries with the assistance of PAHO.

Methodological Aspects
Survey Design
The GYTS is a school-based survey of children in grades corresponding to ages 13 through 15 years. It uses a two-stage cluster sampling design within the participating geographical area in each country. The sample is selected from either the country, a state or a city, depending on the size of the country and the financial resources available for the survey. The first stage requires the sampling of schools with the probability of selection proportional to their size. The second stage consists of randomly selecting classes within the eligible grades/years of study, from every selected school.

Sample Selection
All countries follow the same procedures. Within each school, entire classes of students are selected rather than selecting students randomly throughout the school because it is easier, saves time and is less disruptive to the school. The classes are randomly selected from a sequentially numbered list of all the classes. The classes must be sections in which eligible students in the school are enrolled once and only once to avoid selecting the same student repeatedly.

All students in the selected classes are eligible for participation in the survey regardless of their age. If a student is under or above the expected age range for their grade (13-15 years old), he or she is also eligible.

Sample size
The number of schools and classes selected in each country is based on expected school and student response rates of at least 80% each. Therefore, the sample size was inflated by 20% to take into account the potential non-responses. The number of classes to be surveyed may vary from school to school, therefore some schools will have more classes selected than other schools. The sample size calculation for each country takes into account potential design effects. Its size is estimated on a country per country basis.

Data collection
Data are collected by means of a questionnaire, self-administered during the mid-morning classes to avoid eliminating students arriving late. Collection of data during lunchtime is always avoided. The collection of data is conducted under the supervision of a Research Coordinator whose responsibilities are to:

a) Obtain permission from the selected school to participate in the survey, identifying a person to serve as the School Contact for the logistics,
b) Provide the materials (For each student: a GYTS questionnaire, answer sheet, and pencil. For each class: an envelope, a header sheet, and a GYTS classroom - level form. For each school: a GYTS school - level form),
c) Administer the survey,
d) Complete the header sheets and GYTS classroom-level form for each class and the GYTS school level form for the school, e) Collect all the materials and send them to CDC.

Table 1 lists the names and institutions of research coordinators in each country and territory.

Data analysis
Data will be entered into EPI Info. Analysis will be performed by the principal investigator in each country and will be checked against an independent analysis done by the Centers for Disease Control and Prevention in Atlanta, USA. For each country, indicators will be estimated, weighted for non-responses and the varying selection probabilities, and presented with their 95% confidence intervals. The main indicators used are presented in box 1.

Status of implementation of the GYTS in the Americas
A total of 111 countries and territories of the world have conducted the survey, are doing so, or are committed to doing so shortly. Among these are 13 Latin American and 18 English-speaking Caribbean countries. Since 1998, PAHO has facilitated implementation of the GYTS by identifying research coordinators and organizing their training and by providing technical and financial assistance to countries.

Training
In December 1998, experts from Venezuela and five other countries of the world as well as staff of TFI and CDC met to develop the GYTS core questionnaire and methodology. Since March 1999, nine GYTS workshops have been carried out worldwide by TFI/WHO and OSH/CDC to train country research coordinators in design, methodology, field work logistics and data analysis. Barbados, Costa Rica and Venezuela attended workshops conducted in Thailand, Geneva and Singapore and were the first countries in the region to conduct the survey, in 1999. In November 1999 a GYTS workshop was held on Margarita Island, Venezuela, to train research coordinators from Brazil, Chile, Colombia, Dominican Republic and Mexico. Two training workshops were held in Barbados in 2000 (April and September) for research coordinators from Latin America and the Caribbean. The participating countries and territories were Antigua & Barbuda Argentina, Bahamas, Bolivia, British Virgin Islands, Dominica, Ecuador, Grenada, Guyana, Jamaica, Montserrat, Panama, Peru, Saint Lucia, Saint Vincent & the Grenadines, Suriname and Trinidad & Tobago.

Completion of the field work and data analysis
Eight Latin American and 12 Caribbean countries and territories have completed the data collection. In addition, six Latin American and three Caribbean countries are currently conducting their fieldwork and one additional country in each of these two regions is preparing to conduct the survey. Table 2 indicates the status of implementation of the GYTS in each country.

Reporting of results The first three countries to conduct the GYTS — Barbados, Costa Rica and Venezuela — have reported and published their results.(1) However, a summary of results from Buenos Aires (Argentina), Cochabamba, La Paz, Santa Cruz (Bolivia), Coquimbo, Santiago, Valparaíso-Viña del Mar (Chile), Costa Rica, Monterrey (Mexico), Huancayo, Lima, Tarapoto, Trujillo (Peru), Venezuela, Antigua & Barbuda, Bahamas, Dominica, Grenada, Montserrat, Guyana, Suriname, and Trinidad & Tobago are in final revision and will soon be published. Conclusions The GYTS is an excellent model of collaborative effort that pulls together many institutions and countries, utilizing the various strengths of each to reduce cost and increase efficiency. It will produce for the first time in Latin America and the Caribbean comparable data on youth attitudes, knowledge and behavior regarding tobacco use. It is also the first time that globally comparable data relevant to policy decisions will be available for the tobacco industry’s key market: new consumers, almost exclusively young, necessary to replace the smokers that die or quit in order to maintain or even increase tobacco company profits. The effort realized thus far has produced extremely valuable information for a single point in time. The challenge now is to ensure the evolution of the survey into a sustainable surveillance system by repeating it at regular intervals in as many countries as possible.

References: (1) Warren, W.; Riley, L.; Asma, S.; Eriksen, M.; Green, L.; Blanton, C.; Loo, C.; Batchelor, S.; Yach, D. (2000) Tobacco use by youth: a surveillance report from the global youth tobacco survey. Geneva: World Health Organization.

Source: Prepared by Dr. Maritza Rojas of PAHO’s Mental Health Program, Division of Health Promotion and Protection (HPP/HPM), Dr. Beverley Barnett from PAHO’s Caribbean Program Coordination (CPC), Dr. Armando Peruga y Heather Selin from HPP/HPM.

Return to Index
Epidemiological Bulletin, Vol. 22 No. 2, June 2001