Title: Incidence of catheter-related complications in patients with central venous or hemodialysis catheters: a health care claims database analysis

Title: Incidence of catheter-related complications in patients with central venous or hemodialysis catheters: a health care claims database analysis Thank you for providing your assessment on our manuscript, "Incidence of catheter-related complications in patients with central venous or hemodialysis catheters: a health care claims database analysis". We are pleased that you highlight the potential for the manuscript to be considered for publication in BMC Cardiovascular Disorders based on the suggested revisions. Please find below point-by-point responses to the reviewer's comments below, along with our amended manuscript with the changes we have made highlighted for ease of review. Reviewer 1 Major Compulsory Revisions 1. Although not specifically stated in the text, the authors aimed to determine the complications attributable to CVC and HD catheters. The authors acknowledged that because of the use of the claims database, they are unsure whether the complications that they report is the result or the cause of the line placement. This is problematic because the authors are unable to answer the hypothesis they set out to answer. The index date was catheter placement and the first complication event following first catheter insertion (following 180-day complication-free period prior to catheter insertion) was counted. Claims database studies are not always able to determine a causative relationship between exposure and outcome. The following measures were taken to ascertain temporality between exposure (catheter insertion) and outcome (catheter complication): • 180 days prior to index catheter insertion patient had to be free of event outcomes of interest • Record of catheter insertion AND removal and replacement was required to establish outcome accrual during placement (not after removal) The Methods section has been modified to more accurately reflect study procedures. 2. It is unclear why patients were divided into CVC and HD groups. Also, if patients had both HD and CVC, the complications were attributed to HD. This is again a problem because the authors set out to attribute complications to a particular catheter type. In the presence of both catheters, it is unclear whether the complication, particularly infection, was due to which catheter. Hemodialysis (HD) catheterization related to renal dialysis represent patients with potentially different complication rates than patients with central venous catheter (CVC) catheterization (not related to hemodialysis). The authors acknowledge that due to the nature of the claims data source, it is impossible to achieve perfect ascertainment in CVC and HD. The following measures were taken to most appropriately …