Total parenteral (intravenous) nutrition in the home setting (Home TPN or HPN) is a life saving strategy in patients who suffer from severe intestinal failure. Unfortunately, this treatment modality remains complicated by the development of frequent infectious complications. This especially relates to the device that is used to establish venous access, mostly a catheter (Hickman-Broviac type) or a port-a-cath.

Taurolodine is an antimicrobial agent without any known side effects or resistance-related problems that holds promise as an effective antibiotic lock solution to prevent catheter infections, as demonstrated recently in a pilot study in HPN patients [Jurewitsch, 2005].

In addition, recently a test has been described which may enable to confirm a suspected diagnosis of central line infection within one hour, the so-called Acridine Orange Leukocyte Cytospin test (AOLC) [Bong, 2003]. For this test, blood that is drawn from the central line is stained with the fluorescent dye acridine orange. Next, microscopic evaluation for the presence of micro-organisms takes place. Implementation of this test might enable to start treatment of the infection and prevent unnecessary removal of non-infected central lines without the necessity to several days wait for culture results.

Two hypotheses will be tested in the AOT trial:

1. The use of taurolidine as an antibiotic lock (5 mL) to flush the central line after infusing the TPN formulation, when compared to the present use of heparin solution (5 mL, 150 E/mL) decreases the rate of catheter-related infections in HPN patients
2. Implementation of the acridine orange staining test allows an earlier diagnosis (or rejection of this diagnosis) compared to culture results. Importantly, in the present trial the results of this test will not be used for treatment decisions.

To this end, patients who are admitted to the clinical ward because of suspected central line infection (mostly because of fever and/or chills) will be treated according to the current protocol for catheter sepsis, after obtaining central line and peripheral blood cultures. This protocol includes the administration of medicinase through the central line and i.v. antibiotics peripherally. After obtaining informed consent, in addition 5 cc of EDTA blood will be drawn from the central line to perform (within 24 hours) the acridine orange cytospin staining test and a Gram stain. These results will than be compared to the culture results. Based on culture results, the antibiotics will be tailored.

If the patient recovers and TPN administration can be restarted, randomisation between heparin lock (5 mL, 150 E/mL) versus taurolidine lock solution 2% (5 mL) will take place. In case the catheter has to be removed because of ongoing or recurrent sepsis, randomisation takes place after a new central line has been placed and the patient continues on TPN.

A new episode of proven catheter sepsis is considered as the primary end-point.

Inclusion Criteria:

• informed consent
• patient on home TPN with central line (Hickman or Porth-a-cath)
• suspected central line infection (acridine orange test)
• proven central line infection (randomisation tauroline vs heparin)

Exclusion Criteria:

• no informed consent
• patient on home TPN with arteriovenous fistula (shunt)