Resumption of linezolid therapy after myelotoxicity.

TIPS procedures performed at our institution , our case definition was modeled after that for infective endocarditis. We evaluated patients with TIPS devices in place who developed sustained bactere-mia; the criteria for sustained bacteremia were the same as those used for sustained bacteremia associated with infective en-docarditis [4, 5]. Certainly, other potential sources of sustained bacteremia, such as infective endocarditis, must be definitively excluded before TIPS infection is suspected. The 3 patients with presumptive TIPS infection described by Armstrong and MacLeod [1] developed sustained and prolonged bacteremia due to Escherichia coli, methicillin-resistant Staphylococcus aureus (MRSA), and Pseudomonas aeru-ginosa. In an attempt to create a more specific standardized definition of TIPS infection, and on the basis of the very prolonged bacteremia that these 3 patients experienced , the authors proposed an alternative definition of sustained bacteremia. They proposed that sustained bacteremia should be defined as у2 blood cultures positive for the same organism, the first and last being separated by у7 days. I disagree with this proposed definition. The definition of sustained bacteremia due to endovascular infection should not differ according to the site of infection. Infection of the endovascular system is defined by continuous bacteremia over time. Once continuous bacteremia is documented, there is no reason to apply an arbitrary duration of bacteremia, such as 7 days. It may be noteworthy that the 3 patients described in this report were bacteremic for several days, but this could be explained by the virulence of these 3 organisms or the antimicrobial therapy used. MRSA, P. aeruginosa, and E. coli may be more difficult to eradicate than some of the other organisms that have been implicated in TIPS infection. For example, resolution of TIPS-associated bacteremia due to Strep-tococcus sanguis or Lactobacillus acidophi-lus may occur more quickly than that due to MRSA. Furthermore, it has been suggested that resolution of bacteremia due to MRSA may be slower with vancomycin therapy than with b-lactam antibiotic therapy [6]. Another factor that may contribute to the duration of bacteremia associated with TIPS infection is the underlying immune status of the patient. Bacteremia is a known complication of cirrhosis, and underlying cirrhosis may affect response to antimicrobial therapy [7]. In our analysis, we classified each patient by underlying etiology of cirrhosis and by Child-Pugh class to identify whether these factors affected the duration of TIPS infection. Armstrong and MacLeod [1] do not provide the Child-Pugh classification for the …