Central Venous Line Thromboprophylaxis

Abstract

As health care providers, our directive when caring for patients is to ‘‘do no harm,’’ with hope of providing benefit. In the absence of clinical trials establishing the safety and efficacy of therapeutic interventions, we are forced to turn to personal or overall clinical ‘‘experience’’ to provide patient care. Preventative clinical practices based on ‘‘experience’’ have been described as ‘‘arrogant,’’ with many resulting in catastrophic outcomes. Examples include the use of oxygen in premature infants to maintain blood saturations close to those of healthy full-term infants with resulting retrolental fibroplasia or the recommendation to place infants prone for sleeping, which has subsequently been demonstrated to be associated with sudden infant death syndrome (SIDS). These and other clinical catastrophes have provoked a shift in health care, both in clinical practice and in resource allocation, to provide ‘‘evidence-based’ care to patients with clinical trials providing safety and efficacy determinations of the clinical practice in question. The clinical practice of placing central venous catheters (CVC) in critically ill infants and children is required to administer life-saving therapies, yet results in thrombosis in up to 50% 5 of children. Health providers know that the consequences of CVC-related thrombosis can be devastating, including mortality (3%), morbidity (20% of recurrence and post thrombotic syndrome [PTS]) and loss of vasculature necessary to support a child through organ transplantation or future surgical interventions. Consequently, in an effort to prevent this problem, many health providers, 45% of those who responded to this survey, have hypothesized that the use of heparin is beneficial without causing harm. A meta-analysis of studies in adults have shown some benefit for use of anticoagulants to prevent CVC-related thrombosis. However, the use of unfractionated heparin (UFH) is associated with serious adverse events; most commonly hemorrhage, demonstrated to be 20% in a critical care population, and infrequently, heparininduced thrombocytopenia (HIT). The safety and efficacy of thromboprophylaxis of CVCs in children using radiographically confirmed thrombosis as an outcome has not been studied in properly designed clinical studies. Studies have demonstrated that line patency is preserved with heparin, but these studies used patency as a surrogate for thrombosis. In addition, CVC-related thrombosis has been demonstrated to occur in the presence of line patency. Using thrombosis as the endpoint, the use of a vitamin K antagonist (warfarin) for CVC thromboprophylaxis in a randomized clinical trial in children with acute lymphoblastic leukemia was not demonstrated to be efficacious. The absence of safety and efficacy data for thromboprophylaxis of CVCs is reflected in the American College of Chest Physicians (ACCP) recommendation to not use thromboprophylaxis. Clarke et al have recognized the absence of safety and efficacy data for the use of UFH for CVC thromboprophylaxis and have taken the first step in the process toward completion of well-designed studies. The survey was developed by the authors’ in an attempt to determine current clinical practice in PICUs in the United States. The response rate using a Web-based reply was 26.9%, and of the 96 respondents (1 response allowed per centre), 45% reported using a low-dose heparin infusion (5-10 m/kg/h) for CVC thromboprophylaxis. The most common indications for CVC prophylaxis cited were the child’s underlying health condition, 62%, and the preference of the attending critical care physician, 42%. Health providers who administered thromboprophylaxis acknowledged the potential occurrence of heparin-associated adverse events, but their ‘‘experience’’ indicated that events were rare and that the incidence of CVC-related thrombosis was possibly decreased with heparin thromboprophylaxis. This clinical ‘‘experience’’ prompted these health providers to administer a low-dose heparin infusion. The results of this survey should be interpreted with caution due to the low response rate. The authors point out the