Friday, 12 December 2003
Sheraton Boston Hotel Grand Ballroom (1100)
EVAL-176-249

This presentation is part of EVAL-176. Poster Session

Perception of Smoking Risk in a Poor Population

Brion J. Fox, JD, University of Wisconsin Comprehensive Cancer Center, Cancer Control Outreach, bjfox@uwccc.wisc.edu, Xiaodong Kuang, MA, kuang@uwccc.wisc.edu.

Learning Objectives: Identify those traits that are likely to predict who fails to understand the relationship between smoking and lung cancer in a poor population

Abstract:
Problem/Objective: Surveys indicate that most people are aware of the risks of smoking. Nevertheless, a significant proportion of smokers report that a smoker’s risk of contracting lung cancer is at or below that of a non-smoker. Past efforts to characterize smokers who hold this inaccurate belief (risk-unaware smokers) have suggested that there are not significant demographic predictors. We examined whether this general finding holds in a sample of smokers taken from a predominantly poor population, to further test if being risk-unaware is related to demographic characteristics.
Methods: A subset of respondents from the Wisconsin Dependence Survey presented a convenience sample of individuals from inner city Milwaukee (n=192) to test our hypothesis. Pairwise comparisons were conducted to identify relevant variables for a logistic regression model, which also controlled for other demographic characteristics.
Results: Pairwise comparisons found four variables to be statistically significant predictors of being risk-unaware: being African-American, having less than a college education, having a low level of tobacco dependence (measured by the Fagerstrom Test of Nicotine Dependence), and not having a smoking-related disease. When these four variables were placed in a logistic regression model that also controlled for income, gender and age, FTND was no longer statistically significant.
Conclusions: Unlike a recent statewide survey, these results suggest that there may be identifiable groups of individuals who are more susceptible to being risk-unaware. Since studies have shown a connection between intent to quit and being risk-aware, these groups should be further identified and targeted for prevention and treatment efforts.


Back to Poster Session
Back to Evaluation and Surveillance

Back to The 2003 National Conference on Tobacco or Health (December 10-12, 2003)