To the Editor: In most serious infections associated with intravascular devices, the device is a central venous catheter (CVC).1,2 Good clinical practice dictates that these devices should be removed when no longer needed.3

Our intensive care unit maintains a clinical practice of prompt removal of CVCs once they are no longer required. In addition, CVCs are to be removed before patients are discharged from the unit (for “general” patients) or within 24 hours (for cardiothoracic surgical patients). We conducted an audit to determine how often this practice was followed and whether it had unintended adverse clinical consequences (eg, need to reinsert a CVC).

The audit was conducted over 8 weeks in 2001 and included 126 CVCs in 103 patients. Fifty-eight CVCs (46%) were removed by the predetermined time, and 68 (53%) were retained past this time (Box).

The data demonstrated:

• Low removal rates, with 31% (23/74) of CVCs removed in general patients, but higher rates in cardiothoracic surgical patients (67%; 35/52).

• APACHE II scores, ventilation times and lengths of stay in the intensive care unit were similar in the group who had the CVC removed and the group who retained the CVC, implying that severity of illness was not a factor biasing retention.

• Among patients who had a CVC removed, none had another CVC reinserted; cannulation rates with short peripheral catheters were low and acceptable.

• Retention of the CVC past the predetermined time resulted in significant prolongation of CVC in-situ time (eg, general patients 7.0 v 3.2 days).

The reason given for CVC retention in general patients was antibiotic administration for 37/40 (93%) (vancomycin, 6; β-lactams, 20; aminoglycosides, 9; and quinolones, 2), while total parenteral nutrition and poor peripheral access were factors in only seven (18%). In the 13 cardiothoracic surgical patients, the reason given was inotrope infusion in four (30%) and amiodarone infusion in four; no reason could be ascertained in the other five.

Two issues emerge from this audit. Firstly, there did not appear to be good reasons for retaining many of these CVCs. Drug therapy was most often quoted, but many of these drugs (antibiotics and amiodarone) could have been safely administered via a short peripheral intravenous cannula, with markedly lower risk of infection.2-5 We believe that when a patient left the intensive care unit with a CVC, it most likely remained in place as a “convenience” factor for busy nursing and junior medical staff on the wards. However, this was at the cost of a significant increase in CVC in-situ times, increasing the risk of both mechanical and infectious complications.

Secondly, implementation of predetermined CVC removal appears safe in our hospital. Although removal rates were lower than expected, nearly half of all CVCs inserted over the 8-week study period were removed as per “clinical practice”, without any need for CVC reinsertions.

We have now further refined our clinical practice and introduced a formal written policy that:

• All CVCs should be removed and replaced with a suitable alternative before patient discharge from the intensive care unit, unless there is a specific indication for retention.

• Retention of a CVC should be based on simple guidelines, such as need for total parenteral nutrition, poor peripheral venous access, or use of drugs that require central access.

We recommend implementation of this simple policy in other intensive care units.