Author(s): Barros A, Hirakata V
Background: Cross-sectional studies with binary outcomes analyzed by logistic regression are frequent in the epidemiological literature. However, the odds ratio can importantly overestimate the prevalence ratio, the measure of choice in these studies. Also, controlling for confounding is not equivalent for the two measures. In this paper we explore alternatives for modeling data of such studies with techniques that directly estimate the prevalence ratio.
Methods: We compared Cox regression with constant time at risk, Poisson regression and log-binomial regression against the standard Mantel-Haenszel estimators. Models with robust variance estimators in Cox and Poisson regressions and variance corrected by the scale parameter in Poisson regression were also evaluated.
Results: Three outcomes, from a cross-sectional study carried out in Pelotas, Brazil, with different levels of prevalence were explored: weight-for-age deficit (4%), asthma (31%) and mother in a paid job (52%). Unadjusted Cox/Poisson regression and Poisson regression with scale parameter adjusted by deviance performed worst in terms of interval estimates. Poisson regression with scale parameter adjusted by chi2 showed variable performance depending on the outcome prevalence. Cox/Poisson regression with robust variance, and log-binomial regression performed equally well when the model was correctly specified.
Conclusions: Cox or Poisson regression with robust variance and log-binomial regression provide correct estimates and are a better alternative for the analysis of cross-sectional studies with binary outcomes than logistic regression, since the prevalence ratio is more interpretable and easier to communicate to non-specialists than the odds ratio. However, precautions are needed to avoid estimation problems in specific situations.
Referred From: https://www.ncbi.nlm.nih.gov/pubmed/14567763
Author(s): Brubacher JR, Mabie A, Ngo M, Abu-Laban RB, Buchanan J, et al.
Author(s): Hungerford DW, Williams JM, Furbee PM, Manley WG, Helmkamp JC, et al.
Author(s): Barros RE, Tung TC, Mari JJ
Author(s): Del-Ben CM, Tung TC
Author(s): Foli-Andersen NJ
Author(s): Kohler S, Hofmann A
Author(s): McCullumsmith C, Clark B, Blair C, Cropsey K, Shelton R
Author(s): Hungerford DW, Pollock DA, Todd KH
Author(s): Helmkamp JC, Hungerford DW, Williams JM, Manley WG, Furbee PM, et al.
Author(s): Cherpitel CJ, Ye Y
Author(s): Allen JP, Litten RZ, Fertig JB, Babor T
Author(s): Brown JM, Miller WR
Author(s): Booth RH, Corsi KF, Mikulich-Giberson SK
Author(s): Heffner JL, Tran GQ, Johnson CS, Barrett SW, Blom TJ, et al.
Author(s): Longabaugh R, Woolard RF, Nirenberg TD, Minugh AP, Becker B, et al.
Author(s): Chafetz ME, Blane HT, Abrams HS, Golner JH, Lacy E, et al.
Author(s): McCambridge J, Cunningham JA
Author(s): D’Onofrio G, Degutis LC
Author(s): Stange K, Flocke S, Goodwin M
Author(s): Wallace P
Author(s): Nordqvist C, Johansson K, Bendtsen P
Author(s): Segatto ML, Pinsky I, Laranjeira R, Rezende FF, Vilela TR
Author(s): Jayaram G, Triplett P
Author(s): Havard A, Shakeshaft A, Sanson-Fischer R
Author(s): Carr CJA, Xu J, Redko C, Lane DT, Rapp RC, et al.
Author(s): Xu J, Rapp RC, Carr CA, Lane DT, Wang J, et al.
Author(s): Allen K
Author(s): Allen K
Author(s): Rapp RC, Xu J, Carr CA, Lane DT, Wang J, et al.
Author(s): Breton AR, Taira DA, Burns E, O´Leary J, Chung RS
Author(s): Aseltine RH
Author(s): Babor TF, Higgins-Biddle, JC
Author(s): Brown RL, Saunders LA, Bobula JA, Mundt MP, Koch PE
Author(s): Daeppen JB, Gaume J, Bady P, Yersin B, Calmes JM, et al.
Author(s): Leontieva L, Kimberly HE, Helmkamp J, Furbee M, Jarrett T, et al.
Author(s): Wright S, Moran L, Meyrick M, O’Connor R, Touquet R
Author(s): Rockett IRH, Putnam SL, Jia H, Smith GS
Author(s): Curran GM, Sullivan G, Williams K, Han X, Allee E, et al.
Author(s): Pal HR, Yadav D, Mehta AS, Mohan I
Author(s): Blow FC, Walton MA, Murray R, Cunningham RM, Chermack ST, et al.
Author(s): Stokes ME, Davis CS, Koch GG