Diagnosis of central venous catheter related sepsis--a critical look inside.

The use of central venous catheters for the administration of pharmaceutical agents, including chemotherapy regimens, inotropic support in the intensive care setting, intravenous nutrition, cardiac monitoring, and as a means of maintaining long term venous access, has increased dramatically in the past three decades. Complications associated with central venous catheterisation include those associated with insertion—for example, pneumothorax, haemorrhage, nerve injury, catheter tip misplacement, and cardiac arrhythmias—and those associated primarily with longer term use, including thrombosis and infection. Catheter related infections represent by far the greatest risk associated with the use of central venous catheters, and the rate of catheter related sepsis is variably reported to range from 4% to 14%. The magnitude of this variance reflects true diVerences in the incidence of catheter related sepsis in some patient groups—for example, the rate of catheter related sepsis in burns patients managed in intensive care is approximately 15-fold higher than in those with respiratory disease. Infections associated with central venous catheters therefore represent approximately 30 000 and 400 000 cases in the United Kingdom and the USA, respectively, each year. However, diVerences in the reported incidence of catheter related sepsis also result from a lack of standardisation in diagnostic approach. Catheter related sepsis is associated with significant morbidity and mortality, and with case fatality as high as 10–20%. Major complications of sepsis were reported in 32% of cases in one series. There has been a twoto threefold increase in the cases of primary nosocomial bloodstream infections in the last decade, the large proportion of which have been attributed to catheter infections. The incidence of hospital acquired infection in a recent surveillance study was sevenfold higher in patients with an invasive device. Catheter related sepsis represents a significant burden to the health service, and the excess hospital cost associated with these bloodstream infections has been estimated at $40 000. Another report calculated the cost of a single episode of catheter related sepsis in patients on an intensive care unit (ICU) as up to $28 000. Episodes of catheter related sepsis cause a major proportion of the septicaemias due to coagulase negative staphylococci, Staphylococcus aureus, and Candida spp. This review aims to discuss contentious issues relating to the aetiology and diagnosis of catheter related sepsis, and to challenge some beliefs using recently available data. Definitions Perhaps one of the greatest diYculties in reviewing the subject of central venous catheter infection is the large variation in what is considered to be an “infected catheter.” This problem appears to have arisen as a result of the many methods which have been described to culture catheters. In addition to the multitude of catheter culture techniques much debate still exists as to precisely what is a significant quantity of bacterial growth. Central venous catheters are inserted through and reside in close proximity to skin containing approximately 10–10 bacteria/cm. Furthermore, approximately 10 skin scales/person are shed daily, about 10% of which contain bacteria. Hence, separating infected, colonised, and contaminated central venous catheters can be extremely problematic. Furthermore there is no gold standard method whereby all techniques for the diagnosis of central venous catheter infection can be compared, and thus the vast majority of sensitivities and specificities quoted should be critically assessed and cannot be taken at face value. Despite these shortcomings, there is a generally accepted definition of catheter related sepsis (or catheter related bacteraemia) which requires the following three criteria to be present: (1) A significantly positive catheter culture (although the definition of “significant” is contentious). (2) A positive peripheral blood culture taken before catheter removal. (3) The same microorganism isolated in both (1) and (2). By insisting on the presence of an associated peripheral bacteraemia this allows for more accurate comparisons of methodology and diagnosis of catheter related sepsis. However, the significance of a colonised catheter in the absence of a systemically proven infection can be strongly debated. While it is accepted that a positive catheter culture in the absence of peripheral bacteraemia may occasionally represent either poor peripheral blood sampling or even a transient fall in peripheral bacterial load at the time of sampling, it is the only method by which one can truly compare all of the methods for the detection of catheter related sepsis. Unfortunately, blood culture contamination by skin microorganisms is common, and recent studies have highlighted the low positive predictive value of blood cultures positive for coagulase negative staphylococci. 14 For example, of 89 blood cultures positive for skin flora, 91% involved coagulase negative staphylococci, and the incidence of significant and indeterminate coagulase negative staphylococcal bacteraemia and of contamination was J Clin Pathol 1999;52:165–172 165