Smoking Prevalence Evaluation by Three Different Methods among First-Year Health Sciences Students at theUniversity of Leon, Spain, 2006

Abstract

Background: Intervention in the prevention and control of the smoking habit among health sciences students in order to improve the beliefs and attitudes of future professionals concerning this serious public health problem. The objective of this study is to compare the validity of three different methods for calculating the prevalence of smokers. Methods: First-year students at a Health Sciences School. Smoking was evaluated by means of a self-administered questionnaire, the carbon monoxide (CO) in expiratory air and saliva cotinine. The prevalence found using each method, the agreement with the Kappa index and the internal validity of the questionnaire and the cooximetry were evaluated using the cotinine test as the gold standard. Results: A total of 29.3% of the 181 students studied were classified as smokers by some measurement; 26.5% (cotinine), 22.1% (self-stated) and 9.4% (cooximetry). Two thirds of the smokers had low nicotine dependence. Good agreement was found only between the questionnaire and the cotinine test (Kappa Index=73.1%). Compared to the cotinine test, the questionnaire and the cooximetry respectively showed 72.9% and 35.4% sensitivity, and a 96.2% and 100% specificity. According to the questionnaire, 81.5% of the low-dependence smokers tested positive for the cotinine test, 14.5% having tested positive in the cooximetry. Conclusions: Cooximetry does not seem to be a good method for detecting smokers in this population. The cotinine test and the questionnaire are found to agree well, but the internal validity of the questionnaire in comparison to the cotinine test shows a lesser degree of sensitivity than expected. The gold standard criterion used must be revised for a low-dependence population with possible exposure to second-hand smoke.