Related Experiment Videos
[Critical evaluation of school-based anti-smoking education in Japan]
Masahiro Shima1, Itsuro Ogimoto, Akira Shibata
1Kurume University School of Medicine, Department of Public Health.
Objective:
A critical evaluation was performed on school-based anti-smoking education reported over the past 25 years in Japan.
Methods:
Relevant papers were retrieved by the key words of "smoking prevention" or "anti-smoking education" and those on anti-smoking education actually performed for pupils or students in Japan were collected. According to the criteria for whether they might be regarded as quasi-experimental studies regarding use of a control group, and performance of pretest and posttest assessment. Response rate, individual behavioral change, target group for evaluation, comparison methods e.g. before-after comparison, and points of evaluation were reviewed for each paper.
Results:
Of 27 relevant papers, approximately 40% utilzed a control group and pretests were not conducted in 30%. Two kinds of posttest were performed; one was an immediate posttest 2 weeks after the anti-smoking education, and the other featured a longer interval. An immediate posttest only, a later posttest only and both posttests were performed in 10, 5, and 10 studies, respectively Eight, or 30%, were regarded as quasi-experimental for which the ranked of the quality of evidence might be II-1. Individual behavioral change was examined in only one paper.
Conclusions:
Three fourths of the 27 papers reviewed had some drawbacks in terms of study design. One half of the control groups were set in the same school as the education group and the remainders set in separate schools. Preferably, control groups should be set both in and out of the school of the education group. Apretest is an essential step and a posttest at the age of 20 is preferable where possible. Methodological innovation is required for individual identification and follow-up. In addition, objective indices of behavioral change should be analyzed and biases such as arising from selection bias should carefully be watched. Knowledge of epidemiological study designs is of essential importance to improve the quality of evaluations of health education programs.