The Authors Respond

We have received multiple thoughtful comments via letters to the editor regarding our article, "Interprofessional Approach to Reducing Central–Line Associated Bloodstream Infections in a Cardiac Surgical Intensive Care Unit."1 The comments highlight that some practices remain controversial and that further evidence from randomized controlled trials is needed to better guide patient care. We would like to respond to several of the comments that were made.First, several letter authors questioned the safety of using midline catheters for vasopressor infusions. We acknowledge that some guidelines recommend against this practice. However, based on our anecdotal experience over several years, it can be done safely. As we described in our article, our institution has limitations on which drugs and dosages can be infused through a midline catheter, which adds a layer of safety to this practice. Multiple cohort studies support the safety of using midline catheters for low-dose vasopressor infusions.2-4 The extravasation rate appears to be low in these studies, although there is a risk for venous thrombosis.Second, as noted by the letter authors, peripheral intravenous catheter–related infections can occur; however, when high-quality studies are considered, the risk of bacteremia appears to be lower compared with central venous catheters (CVCs).5 Large epidemiologic studies would be helpful in determining the true risk for bacteremia associated with midline catheters. Given this uncertainty, we agree with the letter authors that patients and families should be involved in vascular access decisions whenever possible.Third, our statement that the insertion of midline catheters is a "straightforward" skill was imprecise, and we did not intend to minimize the training required to perform the procedure properly. Insertion can be technically challenging, particularly among patients with obesity or small veins. A more accurate statement is that, with proper training, midline catheter insertion has a low complication rate and can be performed reliably at the bedside by vascular access nurses, advanced practice providers, and physicians.Fourth, we do not believe our "3-attempt policy" for CVC insertion has increased the risk of complications for patients. All CVC insertions at our hospital are supervised by an attending physician or advanced practice provider, who may choose to perform the procedure themselves before a third attempt is made by a learner. At our institution, 3 attempts are rare, and the third attempt serves as a strict upper limit or "hard stop." In teaching institutions, a tradeoff must be considered when deciding whether to allow a learner multiple attempts or to stop after one. In our opinion, learners should be allowed several supervised attempts if the supervisor believes it is safe. This process is an essential part of a trainee's learning, as not all CVC placements will be uncomplicated in future practice.Finally, regarding the trialysis catheters discussed in our article, our institution has discontinued the use of non–antibiotic-coated catheters and believes this change has helped reduce our central catheter–associated bloodstream infection rate. We now use only antibiotic-coated temporary dialysis catheters or CVCs, which are associated with lower infection rates.6,7 This practice requires 2 separate insertion sites in some patients, but we believe the benefits outweigh the risks.