Table 2 Selected characteristics of nationally representative studies* of cessation attempts among adolescent smokers according to type of cessation attempt estimate
CharacteristicNumber of estimates
Any cessation attemptMultiple cessation attempts
6-month prevalence12-month prevalence†Lifetime prevalence12-month prevalence‡Lifetime prevalenceDuration of cessation attempt
Geographic location
North America330 37 70 74 75102 33 35 36 50 51 626762 37 70 74 755626627 74337 70 74
Central America and the Caribbean0933 41 47–49 71–730000
South America0233 420000
Europe0444–46 53132000
Africa0739 40 52 57600000
Middle East0533 54560000
Asia0533 38 43 610000
Oceania023316801680
Year of data collection
Prior to 19800012000
1980–1989130 74 75032 74 750174174
1990–1999137102 33 35 39 42 63 72232 37163137137
2000–20051703433 36 38 40 41 43–62 64–67 71 73268 70462 6466168170
Sampling frame
Schools03833 35 36 38–61 71–7352 32 6801680
Households130 74 7562 6267174 7556266174174
Other§237 700237 700137237 70
Sample size||
Not stated022 63016300
<5001702333 40–42 47–49 52 54–57 59–61 71 73268 700168170
500–<10000833 46 62 64–67 720462 646600
⩾1000230 37 74 751135 36 38 39 43–45 50 51 53 5862 32 37 74 750237 74237 74
  • *We defined a study as a sample of individuals in a particular place during a particular time period.

  • †Although in some instances these were available from other reports,31 71 study characteristics were abstracted from The Global Youth Tobacco Survey Collaborative Group33 for the following study populations: Barbados (1999), Fiji (1999), Gaza Strip (2001), Jordan (1999), Montserrat (2000), Northern Mariana Islands (2000), Singapore (2000), Sri Lanka (1999), and St. Vincent and the Grenadines (2001). Study characteristics for Bahamas (2000) were abstracted from Perez Martin and Peruga (2002).71

  • ‡Study characteristics for Canada (2000) were abstracted from Health Canada.66

  • §Other includes studies with a sampling frame that included both schools and households.37 70

  • ||Reflects the sample size of current smokers. The sample size for age and smoking history specific subgroups as well as for the duration of cessation attempts was usually smaller. If the sample size of current smokers was not provided, wherever possible, it was estimated from (a) the total study sample size multiplied by the proportion of current smokers or (b) the large-sample 95% confidence interval for a population proportion (p±1.96*SEp) where p  =  prevalence and SEp  =  standard error of p.