Chapter 26 ICU Infections

560

SECTION II • Medical and Surgical Crises

been exhausted. Guidewire changes might make sense for patients in whom alternate sites for cath- eter insertion are unavailable or for those at unusu- ally high risk for insertion of a catheter at a fresh site (e.g., coagulopathy, tenuous respiratory status, bilateral femoral vein thrombosis). When receiving a patient from another health care facility, as a gen- eral rule, it is reasonable to treat indwelling cath- eters as contaminated, regardless of their duration of insertion. Diagnosis Although redness, pain, and swelling around the insertion site strongly suggest infection, these signs are often absent in patients with catheter-related infection (Fig. 26-6). Local (soft tissue) catheter infections may be confirmed by Gram staining and culturing the catheter and by “milking the entry wound” to provide material for examination. Because intravascular infections usually produce recurrent and sometimes continuous low-level bac- teremia, collecting several sets of cultures obtained over hours to days is sensible. A positive blood culture withdrawn through a potentially contami- nated intravenous line does not necessarily estab- lish a diagnosis of catheter sepsis; it is possible that the patient has systemic bacteremia from another source. That supposition is bolstered if the same organism is simultaneously recovered from a clean distant venipuncture site. However, if cultures from the catheter are positive but cultures from a peripheral stick negative, the catheter is suspect, especially if the catheter-obtained cultures grow rapidly. In patients with suspected “line” sepsis, the catheter, tubing, and fluids should be replaced with fresh components. Before catheter removal, the

skin should be cleansed with chlorhexidine. The distal centimeter of the catheter tip should then be sent in a sterile container for culture and Gram stain. Semiquantitative culturing is performed by rolling the tip of the catheter across a culture plate. If more than 15 colonies of a single organism are isolated, infection is more likely than colonization. The catheter tip should not be placed into any solu- tion for transport—doing so renders quantitative culturing impossible. Routine catheter changes over a guidewire are not rational in patients with sepsis or inflamed entry sites and are not necessary for asymptomatic patients. Common Organisms Although S. aureus, S. epidermidis, and Candida cause most catheter-related infections, enteric Gram- negative rods are recovered occasionally. Although rare, blood cultures growing Enterobacter agglomer- ans, Pseudomonas cepacia, E. cloacae, Serratia marces- cens, Citrobacter freundii, or Corynebacterium species should suggest a contaminated intravenous solution. Treatment In almost all cases, contaminated catheters should be removed and cultured as outlined earlier (this includes PICC lines, temporary dialysis catheters, Portacath and Hickman devices). Blood cultures should be obtained from a site separate from the catheter insertion site. Considering the high inci- dence of MRSA in many ICUs, initial empiric antibiotic therapy for the patient with sepsis from a suspected intravenous line source should include vancomycin or other effective agent in doses adjusted for renal function. In units in which MRSA is rare, an antistaphylococcal penicillin is a reasonable initial

FIGURE 26-6. Visibly infected catheter sites.