Author's response to reviews Title:Systematic review of cigar smoking and all-cause and smoking-related mortality Authors:

To the Editor-in-Chief: We respectfully resubmit to BMC Public Health our manuscript (#1872779892138581) entitled " Systematic review of cigar smoking and all-cause and smoking-related mortality. " We have revised the manuscript based on the comments from the reviewers. Please see below, the point-by-point response to each of the reviewer comments, including the revisions made to the manuscript. Attached are the highlighted and clean versions of the revised manuscript. Thank you for considering this work for publication. Please contact me if you have any questions. 1. The literature search was incomplete, as many studies may use incidence as the primary outcome and mortality as the secondary outcome. With reference to the existing search strategy, these studies were ignored. I also have a random check from PubMed. I can easily find a potential paper which mentioned cancer mortality (i.e. Shapiro JA, et al. J Natl Cancer Inst 2000;92:333-7. This study presented cancer mortality in table 3) Author response: Our original search strategy included handsearching of references in the selected studies which would have captured additional studies not found in the electronic database searches. In particular, we examined every study referenced in the comprehensive review by Shanks et al. 1998 that included studies of both incidence and mortality. Based on this handsearching, we identified 3 additional studies as described in our methods. The Shapiro et al. 2000 study identified by Reviewer 1 is already included in our review and widely referenced throughout the manuscript. We are slightly puzzled why the reviewer suggested that Shapiro et al. assessed both incidence and mortality when in fact only mortality was assessed in the CPS II cohort. Finally, we expanded the Pubmed search to include any studies of cigar and risk that were published after the Shanks et al. 1998 review. We used the following search terms: (cigar[tw] OR cigars[tw] OR cigarillo[tw] OR cigarillos[tw] OR cheroot[tw] OR cheroots[tw] OR stogies[tw]) AND risk[tw]. No additional studies of incident disease that also included mortality were identified. Thus, we believe that our search strategy adequately captured all of the studies of cigar use and mortality. 2. The current presentation of MRs is in a range rather than a summary estimate from meta-analysis. Methodology should be revised according to the guidelines. Authors should apply the statistical techniques to combine the results across different studies. Author response: While we considered pooling the study results into summary estimates, as we stated in the …