Antibiotics in intensive care: too little or too much?

T he article by Fragou et al (1), in this month’s Critical Care Medicine, provides compelling evidence that the routine use of ultrasound guidance increases the safety of central venous access by the subclavian route. In this cohort of critical care patients, cannulation was achieved in 100% of patients in the ultrasound group as compared with 87.5% in the landmark arm. Access time and the number of attempts were reduced in the ultrasound group as were arterial puncture, hematoma, hemothorax, pneumothorax, and brachial plexus injury. Catheter misplacements did not differ between groups. This has important implications for clinicians performing such procedures, including those from critical care, anesthesiology, cardiology, and other acute specialities. Landmark guided subclavian vein access has been used for many years, but its success has been marred by a small but persistent risk of arterial damage and pneumothorax (2). Such complications, even if recognized and managed correctly, are very likely to have adverse effects on an already critically ill ventilated patient. Ultrasound guidance has the potential to significantly reduce the risk of such early complications. This study follows similar findings from the same group in relation to internal jugular access (3) and I provided editorial comment in 2006 asking the question “Can you justify not using ultrasound guidance for central venous access?” (4). Similar reasoning would apply to the subclavian route following this study. It is also likely that later infective and thrombotic complications would be also reduced by a reduction in needle passes. Subclavian access (as opposed to jugular or femoral) has been recommended to reduce catheterrelated infections in many local and national guidelines, e.g., the Matching Michigan initiative. Until this study, there has only been evidence from prospective randomized studies to support the use of ultrasound guidance by the internal jugular route. Some clinicians have used this distinction to justify not using it at other sites. I would argue that it is intuitive that if ultrasound guidance reduces complications at one vascular site, then providing adequate imaging can be achieved, similar benefits should be achievable at all other commonly used sites, e.g., femoral, subclavian, and peripheral sites. Some elements of the study deserve further comment. The entry criteria, in terms of choice of patient, and any exclusion criteria, for this or other routes of access were not clearly identified. Only planned nonurgent cases were recruited, which may limit the applicability in more urgent cases. The use of preprocedure ultrasound screening of the subclavian vein in the landmark arm is likely to have introduced bias. Operators then excluded patients with evidence of vein thrombosis and may have modified their landmark technique if the vein was seen to be very deep/superficial or empty. Ultrasound guidance was used to rescue failed landmark procedures and in those requiring procedures in the head-up tilt position. However, the study findings were significant even with any such bias, and if this had been avoided, it is likely that the advantages of ultrasound guidance would have been further increased. Three broad categories of technique exist in relation to accessing the axillary or subclavian veins with the first two of most relevance to critical care: 1) landmark based techniques (5, 6); 2) ultrasound guidance (1, 7); and 3) radiograph screening of peripherally injected contrast flowing centrally (http://emedicine. medscape.com/ article/1348912-media, Accessed July 3, 2011) (8). Many clinicians have not yet used ultrasound guidance for the infraclavicular or supraclavicular routes of access to the axillary and subclavian vein in the belief that the presence of the clavicle blocks the image. This is true in adults, in whom the relevant section of the vein is under the clavicle, but the vein can be visualized more laterally as the axillary vein (7) (Fig. 1) and above the clavicle over the