- : Ms Word, Ms Word Format
- : 67 Pages
- : ₦5000
- : 1-5 Chapters
- Click to DOWNLOAD Materials
TOBACCO SMOKING AMONG SCHOOL CHILDREN IN COLOMBO DISTRICT, SRI LANKA
Abstract
Tobacco smoking is an important problem among schoolchildren. The authors studied the patterns of tobacco smoking among schoolchildren in Colombo, Sri Lanka, using a self-administered questionnaire. Multistaged stratified random sampling was used to select 6000 students. Response rate was 90.7% (5446), out of which 53.4% were males. Prevalence rates for males and females, respectively, were as follows: having smoked at least 1 complete cigarette: 27.0% and 13.3%, smoked more than 100 cigarettes: 2.3% and 0.3%, daily smoking: 1.8% and 0.2%. Mean age of starting to smoke was 14.16 years. The tobacco products most used were cigarettes (91.5%) and bidis (3.8%). In univariate analysis, male gender, parental smoking, studying non-science subjects, peer smoking, and participating in sports were significantly associated with smoking of at least 1 complete cigarette (P <.05). In multivariate analysis, the most significant correlates were having close friends (odds ratio = 3.29, confidence interval = 2.47–4.37) or parents who smoked (odds ratio = 1.86, confidence interval = 1.28–2.71). Female smoking has increased from previously reported values. These high-risk groups can be targets for preventive programs.
Introduction
Tobacco smoking is a leading cause of premature death worldwide. The current estimate of 4 million deaths due to tobacco is expected to double by the year 2020, and tobacco smoking leads to more deaths than any single cause worldwide.1 Majority of these deaths are expected to occur in the developing countries.2 Tobacco consumption remains to be a significant health problem in South Asia where more than one quarter (25.7%) of the males who were older than 15 years smoked tobacco in the year 2000.3
In Sri Lanka and other South Asian countries, tobacco is smoked mainly as cigarettes and locally produced “bidi.4,5 Betel chewing with tobacco (smokeless tobacco) is another common way of consuming tobacco in this region.6
A higher prevalence of smoking among younger age categories has been observed in India.7 There were several studies carried out among school children between the ages of 11 to 18 years in Sri Lanka. The prevalence of smoking had ranged from 10% to 17% among males and 0.1% to 3% among females. The first study we reviewed was published in 1990 by Mendis et al.8 The prevalence of male smoking has reduced over time in subsequent studies.9–11 However, a definitive pattern in female smoking was not noticed.
Adults who started using tobacco early in life are known to be heavier smokers.12,13 Therefore, it is important to emphasize prevention of smoking among schoolchildren. An understanding of the prevalence and determinants of smoking in this population is essential to develop effective preventive strategies.
Methods
This study was conducted in the district of Colombo, which is the capital of and the most populated district in Sri Lanka. In this district, there are both urban and rural areas, according to the classification of the Sri Lankan government. There are approximately 350 000 students studying in more than 400 schools in this district.
Sampling Strategy
Students from grades 10 and 12 were selected for this study. Grade 10 students were included to ensure inclusion of those with lower academic performance not proceeding to higher levels after the barrier exam in grade 11 (ordinary level). Students of grades 11 and 13 were excluded since they are preparing for their exams (GCE [General Certificate of Education] ordinary levels and GCE advanced levels respectively). Students younger than these ages were not included to avoid unnecessary exposure to information on addictive substances.
A sample of 1581 was estimated to detect a prevalence of 4.3% obtained from a previous study among schoolchildren8 with 90% power and 95% confidence interval with 90% power.14 We failed to obtain intracluster correlation coefficients for Sri Lanka for the calculation of the design effect. Therefore, we decided to use the number 2 as the design effect as a conservatively higher estimate. The final sample size was further inflated to 6000, based on increasing the power for the other illicit substances and accounting for nonresponse. Two-staged stratified random cluster sampling was used for recruitment. A cluster was defined as 30 students, which was the approximate number of students in a classroom in majority of the schools.
The database obtained from the Ministry of Education in Sri Lanka was used for sampling. This database used a stratification of schools based on multiple factors such as resources, number of students, number of teachers, and academic and nonacademic performance of students. Probability proportionate to size technique was used to ensure proper representation from all the strata in this classification.
Principals of selected schools were contacted through the Ministry of Education and the local education authorities to obtain permission. In addition to written information, a script, explaining the objectives and procedures of the study, was read out by one of the data collectors. Written consent of the students was obtained in separate forms.
Data Collection Procedures
This survey was conducted in the classrooms of the students. Data collection was carried by medical doctors or third- and fourth-year medical students from the Colombo Medical Faculty. They were given a brief training in research, survey procedures, and ethics, including confidentiality in research prior to data collection.
The class teachers were asked to leave the classroom while the questionnaires were administered. To minimize underreporting, students were assured of anonymity and confidentiality. The students and school authorities were reassured that the data will not be analyzed or published at the level of individual schools. Students were instructed to avoid mentioning information that could identify them or their schools. Answer sheets were collected into unmarked, sealed boxes containing answer sheets of several schools.
Survey Instruments
A self-administered questionnaire was developed initially in English. A group of content specialists did the face validation, and then the questionnaires were administered to a group of students for cognitive debriefing. Then the questionnaires were translated to Sinhala and Tamil languages, both by professional translators and by a group of final-year medical students proficient in English and the native languages. In the validation process, responses were back-translated to English by separate individuals, and the two English versions were compared and further adjustments were made to the Sinhala and Tamil language translates. The survey questionnaire contained 114 items, mostly in the form of multiple-choice questions and short-answer questions. It was designed to be completed in 40 minutes (1 class period). Information on currently practiced drug usage behaviors were obtained from National Dangerous Drugs Authority and the National Alcohol and Tobacco Agency of Sri Lanka. In addition, the questionnaire included items from the Center for Disease Control’s Behavioral Risk Factor Surveillance Survey,15 Transdisciplinary Tobacco Use Research Center,16 the Global Tobacco Use Survey,17 and other sources. This questionnaire also assessed the students’ socioeconomic status, academic and nonacademic performance, and awareness and perception of smoking behaviors and quitting programs. Because of the large variety and volume of data obtained in this study, this article will focus only on the prevalence and correlates of tobacco smoking.
Definitions
A lifetime smoker was defined as having smoked 1 or more complete cigarette in his or her lifetime. A student who has consumed 1 or more cigarette in the previous month was considered to be a current smoker. A smoker who has been able to abstain for more than 6 months was considered to be successful in quitting. A lifetime smoker who was not a current smoker and had not consumed cigarettes during the past 6 months was defined as an ex-smoker.
Data Analysis
Data were entered in duplicate and cleaned. Analysis was performed using STATA version 11 (StataCorp, 2009, Stata Statistical Software: Release 11, College Station, TX). Sampling weights were adjusted to correct sampling error, which occurred because of nested data structure. Estimates of prevalence were obtained using complex survey data analysis method in STATA after declaring the survey design for the dataset.18 Descriptive and demographic data were obtained using frequency analysis. χ2 test was used to compare the prevalence between different categorical variables. Univariate and multivariate logistic regression analysis were carried out to assess the correlates of smoking.
