Article

Odds ratio or relative risk for cross-sectional data?

International Journal of Epidemiology (Impact Factor: 6.98). 03/1994; 23(1):201-3.
Source: PubMed
3 Bookmarks
 · 
445 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To describe the socio-demographic factors associated with exposure to second-hand smoke (SHS) in different settings (home, leisure, and workplace). We analysed cross-sectional data on self-reported SHS exposure in 1059 non-daily smokers interviewed in the Cornellà Health Interview Survey Follow-up Study in 2002. We calculated age-adjusted prevalence rates and prevalence rate ratios of SHS exposure at home, at the workplace, during leisure time, and in any of these settings. The age-standardized prevalence rate of SHS exposure in any setting was 69.5% in men and 62.9% in women. Among men, 25.9% reported passive smoking at home, 55.1% during leisure time, and 34.0% at the workplace. Among women, prevalence rates in these settings were 34.1%, 44.3% and 30.1%, respectively. Overall exposure to SHS decreased with age in both men and women. In men, SHS exposure was related to marital status, physical activity, smoking, and alcohol intake. In women, SHS exposure was related to educational level, marital status, occupational status, self-perceived health, smoking-related illness, and alcohol intake. The prevalence of SHS exposure in this population was high. The strongest association with exposure were found for age and occupational status in men, and age and educational level in women.
    BMC Public Health 02/2007; 7:194. · 2.08 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To identify the sociodemographic factors associated to self-medication (i.e. use of non-prescription medicines) and undesirable self-medication, a cross-sectional study was carried out using a sample (n = 20,311) representative of the population of adults of 16 years of age and older in Spain. Multivariate Cox's regression was used. The prevalence of self-medication in the sample was 12.7% during the two weeks preceding the interview. Self-medication is more prevalent among women, persons who live alone, and persons who live in large cities. For persons who reported acute disorders, self-medication prevalence was higher among those with higher educational levels. The prevalence of undesirable self-medication in the sample was 2.5% during the two weeks previous to the interview. Undesirable self-medication is twice as common among persons older than 40 years, as compared to persons younger than 27 years. Undesirable self-medication prevalence is 53.0% higher among those who live alone as compared to those who live with their partner (95% confidence intervel (CI): 15.2–103.2) and 36.8% higher among students as compared to full-time workers (95% CI: 1.9–83.5). People over 40 years of age, people living alone, and students should be the priority target populations for public health education programs aimed at improving the quality of self-medication behavior.
    European Journal of Epidemiology 12/1999; 16(1):19-26. · 5.12 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Abstract Background It is usually preferable to model and estimate prevalence ratios instead of odds ratios in cross-sectional studies when diseases or injuries are not rare. Problems with existing methods of modeling prevalence ratios include lack of convergence, overestimated standard errors, and extrapolation of simple univariate formulas to multivariable models. We compare two of the newer methods using simulated data and real data from SAS online examples. Methods The Robust Poisson method, which uses the Poisson distribution and a sandwich variance estimator, is compared to the log-binomial method, which uses the binomial distribution to obtain maximum likelihood estimates, using computer simulations and real data. Results For very high prevalences and moderate sample size, the Robust Poisson method yields less biased estimates of the prevalence ratios than the log-binomial method. However, for moderate prevalences and moderate sample size, the log-binomial method yields slightly less biased estimates than the Robust Poisson method. In nearly all cases, the log-binomial method yielded slightly higher power and smaller standard errors than the Robust Poisson method. Conclusion Although the Robust Poisson often gives reasonable estimates of the prevalence ratio and is very easy to use, the log-binomial method results in less bias in most common situations, and because it fits the correct model and obtains maximum likelihood estimates, it generally results in slightly higher power, smaller standard errors, and, unlike the Robust Poisson, it always yields estimated prevalences between zero and one.
    BMC Medical Research Methodology 01/2008; · 2.21 Impact Factor

Full-text

View
28 Downloads
Available from