Point: Should Coagulopathy Be Repaired Prior to Central Venous Line Insertion? YesRequire Coagulotherapy for Central Venous Line: Why Take Chances? FREE TO VIEW

Although the number of studies in the literature that address the question of coagulopathy reversal prior to central venous catheter (CVC) insertion is small, and the question has been a difficult one to address with large trials, the issue is an extremely important one. Rather than operating based on experiential data and “gut feelings” (as we often do), a review of the existing literature in the context of my personal experience practicing in an ICU that cares for many patients with coagulopathy leads me to conclude that reversal of coagulopathy should be pursued prior to central line insertion (when feasible to do so). Given the increasing availability of peripherally inserted central catheters for more stable patients, the insertion of traditional central venous lines is increasingly used more exclusively in critically ill patients for monitoring and support in the ICU. Thus, the risk of any mechanical complication is heightened in critically ill patients with coagulopathy who have little reserve to tolerate additional hemodynamic derangements. I will first review the small pool of existing recent literature describing the risks of mechanical complications following central line insertion with ultrasound guidance; then, for comparison, I will briefly review the risks of life-threatening complications for patients after receipt of blood products for reversal of coagulopathy. Finally, I will conclude that the data support my experience and gut feelings that the risk-to-benefit ratio for our critically ill patients favors reversing a coagulopathy prior to central line insertion.

To frame the argument in favor of correction of coagulopathy prior to central line insertion, one must first assess the risks of CVC insertion and, in particular, the risk of mechanical complications that include arterial puncture, bleeding, and hematoma formation. Notably, mechanical complications occur in 5% to 19% of line insertions and vary with the site chosen for line insertion, with many reports suggesting that the femoral vein site has the highest overall rates of mechanical complication (especially arterial puncture and hematoma), followed by the subclavian vein site (especially pneumothorax and hemorrhage requiring transfusion), followed by the internal jugular (IJ) vein site as the lowest-risk site, with arterial puncture as the most frequent mechanical complication.¹ Importantly, the risk of mechanical complications with CVC insertion has declined in recent years with the introduction of ultrasound-guided CVC insertion (US-CVC), and a number of professional societies have recommended US-CVC preferentially using the IJ location for both the relative safety of insertion at this site and the amenability of this site for ultrasound guidance.^2-4 Thus, I will focus the rest of the debate in the context of US-CVC at the IJ site (even though this gives my opponent a substantial advantage, as this approach is now believed to be the safest overall approach in terms of minimizing mechanical complications). Even with use of US-CVC, studies still report significant mechanical complications, including bleeding, with a study from an academic medical center⁵ reporting a mechanical complication rate of close to 20% (Table 1).^6-8

Although studies have identified different risk factors for bleeding with CVC, including severe coagulopathy, end-stage renal disease, very high or low BMI, dehydration, and number of needle passes required for successful cannulation, many studies identified inexperience of the operator as a critical risk factor for the development of mechanical complications. Interestingly, one recent study called into question the definition of an experienced operator, in that the highest risk of complications was observed in operators with intermediate levels of experience, whereas novice and senior operators exhibited less risk of complications.⁵ The authors speculate that novice operators were under the supervision of more senior operators, whereas intermediate operators might have been on their own and less experienced as the sole or supervising proceduralists. Notably, this study was conducted in an academic ED and, thus, included analysis of US-CVCs from attending physicians as well as house officers with varying degrees of experience.

Interestingly, even with experienced operators and the use of ultrasound, some studies have suggested that mechanical complications may be associated with severe thrombocytopenia (eg, platelet counts <25,000-50,000/μL) or an international normalized ratio (INR) ≥1.5, or with multiple simultaneous coagulopathic derangements (eg, combinations of thrombocytopenia and factor deficiencies with INR ≥1.5 and/or partial thromboplastin time [PTT] >50 s).^3,7 Some authors have argued that even if an arterial puncture is not a frequent event with IJ line insertion (reported as 6% in some studies), there exists the potential for significant risk of these punctures regardless of catheter size. Hematomas are reported to develop following as many as 40% of carotid punctures and can produce life-threatening bleeding if rapid expansion occurs, causing airway compromise. Additionally, carotid punctures pose the risk of development of a life-threatening cerebrovascular event from the need to compress the bleeding site adjacent to the carotid artery (especially in elderly patients with underlying vascular disease).⁹ Two studies that did not report an increase in bleeding risk with US-CVC were either retrospective studies⁶ or included only a small number of patients with coagulopathy,⁷ making definitive conclusions difficult (Table 1).

In general, there has been a growing reluctance to transfuse blood products, unless absolutely indicated, especially within the ICU population, given increasing appreciation of possible immune-mediated events in addition to other traditional risks of blood product transfusion.¹⁰ However, the blood supply has never been safer in terms of risk of blood product-acquired infections, and we must assess the relative risk of a blood product infusion vs the real risk of a mechanical complication related to US-CVC as described previously.

Life-threatening risks from blood product administration are generally quite infrequent and can be divided into infectious and noninfectious risks (Table 2). Risks of HIV and hepatitis C are on the order of 1:2 million units transfused or less.^11,12 Human error in administering a mismatched product occurs in 1:6,000 to 1:20,000 units, with massive hemolysis as a result of mismatch at a rate of 1:100,000 units.¹³ Transfusion-associated lung injury (TRALI) is defined as new onset of acute lung injury within 6 h of transfusion not explained by an alternate cause and arises from donor antibodies reacting against the recipient endothelium, producing vascular leak. The risk of TRALI ranges from 1:2,000 to 1:4,000 units (which may be an underestimate due to underreporting), and though TRALI most often subsides spontaneously, it can produce serious consequences.¹⁴ Although risks of nonanaphylactic allergic transfusion reactions and volume overload are fairly common as a result of blood product transfusion, these are rarely life threatening.

CVC insertion, even with the use of ultrasound at the IJ site, carries a significant risk of mechanical complications. Although the absolute rate of complications is not easily discernible given the small number of studies in the literature and multiple complex factors that contribute to these risks, it is fairly well agreed upon that the employment of experienced operators leads to fewer mechanical complications. However, even in an optimal setting with a low rate of mechanical complications, a serious IJ site bleed can be life threatening given its location near the airway and carotid artery. Although administration of blood products is certainly not without risks and has become increasingly limited to the most severe situations in the ICU setting, the rates of life-threatening complications of transfusions are quite low, and reversal of severe thrombocytopenia or factor deficiencies may decrease the risk of CVC insertion-related bleeding in patients with severe coagulopathies.

Thus, in my experience of working in an academic ICU with operators of varying experience (and many who think they are more experienced than they are) and in caring for a large number of oncology patients with bleeding dyscrasias, I maintain that the published data support the risk-to-benefit ratio favoring reversal of severe coagulopathies (eg, platelets <50,000/μL and/or INR ≥1.5, and/or PTT ≥50 s) if time allows. Perhaps most importantly, inexperienced operators should be supervised by experienced proceduralists, and our hospital has now formalized the training protocol for line insertion and for supervision of less-experienced operators in US-CVC (as have many other centers). Given that there is little that we can control in determining the outcomes of our sickest patients, why tempt fate when sticking needles into the necks of patients with coagulopathy without taking preventive measures to correct what might be a reversible risk? In other words, why take chances?