Fig 2. The odds ratio estimates of the risk of developing adult myeloid disease in smokers.

(A) A forest plot of the risk of developing adult myeloid disease in current smokers. (B) Duval and Tweedie’s trim-and-fill funnel plots were used to observe and adjust the publication bias in current smokers. (C) A forest plot of the risk of developing adult myeloid disease in ever-smokers. (D) Duval and Tweedie’s trim-and-fill funnel plots were used to observe and adjust the publication bias in ever-smokers. Among Fig B and D, The white circles represent observed studies, and black circles represent possibly missed studies imputed using Duval and Tweedie’s trim-and-fill method. White and black rhombuses represent observed and theoretical combined effect size, respectively.