Cancer Control Research5R01CA093967-04
RANDOMIZED STUDY TO DECREASE SMOKING IN COLLEGE STUDENTS
Smoking among college students has increased dramatically in the past decade and has become a cause for concern. Smoking onset during the college years now accounts for 11 percent of college smokers, and 29 percent of college smokers increase their use of tobacco during the college years. Little is known about smoking cessation among college students and few intervention studies have been conducted to assist this population in achieving cessation. In this project, we will recruit 30 four-year, residential colleges in Oregon and Washington state to participate in a group-randomized project to test a comprehensive tobacco control program among college students. We will use innovative methods to reach smokers and non-smokers on campus and work with campus officials and students to create an environment supportive of non-smoking. We will use three major components of intervention: 1) collaboration with state and local Health Department funded coalitions to assist us in implementing tobacco control on college campuses; 2) development and implementation of restrictive policies to develop a non-smoking norm, and 3) provision of a telephone Quit Line and a quit smoking web-site to assist students in achieving cessation. After the 30 colleges are recruited, baseline data will be gathered from a random sample of students at each of the participating colleges. Freshmen will be over-sampled and will form a cohort of approximately 500 students per college to be followed over the project years. Key informants in administrative positions in the colleges and among local tobacco control coalitions will be interviewed. Colleges will be matched on size, survey response rates, and smoking prevalence, and randomized to intervention or control conditions. Primary outcomes of the trial will be change in campus smoking policies, smoking onset rates and cessation rates among students. Secondary outcomes include movement from occasional to regular smoking, change in perception of prevalence of student smoking, and incorporation of program activities. The sample size of 30 colleges has adequate power to assess the primary outcomes. Analysis plans include a permutation test and generalized linear mixed model, with covariates used to control confounders. Durability and dissemination of the program will be structured into the project. Departments of Health in Oregon and Washington are supporting this project and contributing in-kind resources.