A cross-sectional survey was conducted using GYTS of secondary school students (16- to 18-year-old boys and girls) in Riyadh city, Saudi Arabia from the 24th of April, 2010, to the 16th of June, 2010.
Riyadh education department provided a list of all secondary schools in Riyadh city, the number of students in each school, and the grade level of the students. The city hosts 160 secondary schools for boys (72 private schools) and 245 schools for girls (113 private schools). The overall number of Saudi secondary school students in Riyadh city during 2009-2010 academic years was 161223 students, or 83056 (51.5%) boys and 78167 (48.5%) girls.
Schools and students were selected using a two-stage cluster sample. In the first sampling stage, a computer-generated random selection method was used to choose 46 secondary schools in Riyadh city from a list of 405 schools for a total of 19834 Saudi students (9771 boys and 10063 girls). To conduct the sampling, we divided Riyadh city into four districts based on their proximity to two major highways (North, East, South, and west). Fifteen percent of boys’ schools and 10% of girls’ schools from each district were selected based on proportional probability. Fourteen schools were selected from the east district (7 boy's schools and 8 girls` schools), 12 schools from north district (6 boy's schools and 6 girl's schools), 10 schools from west district (5 boy's schools and 4 girl's schools), and 10 schools from south district (5 boy's schools and 5 girls’ schools).
During the second sampling stage, 3 classes, or one for each grade level, were selected from each school. Thus, a total of 138 classes were identified in the sample. Each class was considered to be a cluster and all students in the selected classes constituted the target group of the present study. Thirty boys and 30 girls were selected from each school (10 students from grade 1, 10 students from grade 2, 10 students from grade 3).
Students were interviewed by trained medical students and they completed questionnaires by themselves in the classroom under the supervision of the medical students. The medical students received half a day training session that covered an overview of the study and its methodology and they were taught to avoid explanations that could interfere with the participants’ answers. The GYTS is a school-based anonymous self-administered questionnaire designed to gather data on the following domains: Knowledge and attitudes of young people toward cigarette smoking, the prevalence of cigarette smoking and other tobacco use among young people, the role of the media and advertising in young people's use of cigarettes, access to cigarettes, and environmental tobacco smoke. The Arabic-language version of the GYTS was available from the WHO, and to address the objectives of this study, certain questions related to cigarette smoking were selected from the survey.
Smoking prevalence is defined as the percentage of students who are current smokers. The Center for Disease Control (CDC) defines current smokers as persons who have smoked cigarettes on one or more days during the 30-day period before the survey was administered.[15
] A person who had “ever smoked” was defined as a student who had tried or experimented with cigarette smoking, even one or two puffs, at any stage of his or her life. Any student who did not fulfill the definition of “ever” or “current” was considered to be “never” smoked.
The survey was conducted with permission from the Ministry of Education. We received an approval from the Research and Ethics Committee of the Saudi Thoracic Society. Verbal consent was given by the students after explaining the purpose of the study and highlighting that surveys would be anonymous and data would be kept confidential during handling and storage.
Data were analyzed using STATA V9.1 (SAA Institute, NC). All variables were summarized and reported across the study sample using descriptive statistics. Continuous variables such as age at time of interview and age at first cigarette smoking were collected and entered as categorical variables and then summarized and reported as frequency distributions in the same manner as all other categorical variables such as gender and grade level. A logistic regression analysis was used to analyze the relationship between the dependent variable “ever smoking” and independent variables including gender, friends who smoke, parents who smoke, and type of school. All results were summarized using an odds ratio with a 95% confidence interval. A P-value of 0.05 or less was considered as significant.