Article Text


Failure of regular external ventricular drain exchange to reduce CSF infection
  1. I E Sandalcioglu,
  2. D Stolke
  1. 1Department of Neurosurgery, University of Essen, Hufelandstr 55, 45122 Essen, Germany
  1. Correspondence to:
 Dr I Erol Sandalcioglu;
  1. W S Poon2,
  2. G K C Wong2
  1. 2Division of Neurosurgery, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong;

    Statistics from

    Dr Wong and colleagues undertook quite a careful prospective randomised trial aiming to determine whether routine changing of external ventricular drainage catheters reduces the risk of CSF infection.1 Patients were randomised into two groups: group 1 (n  =  51) had routine changes of the external ventricular drain at five day intervals; in group 2 (n  =  52) the ventricular drain was not changed. There was no difference with respect to the basic demographic data and the incidence of CSF infection. The authors observed four CSF infections in group 1 (7.8%) and two in group 2 (3.8%). Despite the higher CSF infection rate in group 1, this difference was not statistically significant. Based on their results, the authors concluded that “routinely changing external ventricular drainage catheters at five day intervals did not reduce the risk of CSF infection”.

    The topic of ventricular catheters and the risk of CSF infection has been dealt with in numerous reports. The continuing interest for neurosurgeons is largely based on the fact that quite controversial recommendations have been published regarding the use of external ventricular catheters.

    In general, our experience with CSF infections is similar to that of Wong. We investigated which factors increase the incidence of CSF infections in a prospective study including 133 patients who underwent 152 surgical procedures for external CSF drainage.2 Assessed variables included basic demographic data, with special reference to the duration of surgery, diameter of the catheter used (5 F v 10 F), distance of the subcutaneous tunnel between the burr hole and the cutaneous exit point, additional surgical procedures, and duration of CSF drainage.

    In our study group we had a CSF infection rate of 4.5% per patient and 3.9% per surgical procedure. Whereas most of the variables assessed showed no statistically significant correlation with the incidence of CSF infection, interestingly we observed a close correlation between the length of the subcutaneous tunnel and the incidence of infection. In 83% of the patients with CSF infections the catheter was tunnelled subcutaneously for less than 5 cm, whereas in only 17% was the catheter tunnelled for more than 5 cm. This observation was associated with the fact that there was a higher incidence of CSF leakage through the cutaneous exit point with shorter tunnels despite correct operative management.

    Taking into consideration that in the study by Wong et al, “all the bacteria are common in the skin flora of patients in the intensive care unit” and “all infections occurred after day 10” (mean 13 days), these findings strongly support our observation of increased CSF infections caused by secondary contamination rather than as by contamination during the catheter placement procedure.

    In agreement with Dr Wong, we do not recommend routine replacement of the ventricular catheter, but based on our data we strongly recommend a sufficient length of subcutaneous tunnelling (5 cm or more) to reduce the risk of CSF infection, because despite efficient antibiotic treatment a CSF infection is still a serious complication and must be avoided.


    Authors’ reply

    We were pleased to see the above letter about the importance of tunnelling. Subgaleal or subcutaneous tunnelling of ventricular drains has been accepted since the late 1970s as a way of reducing ventriculostomy related CSF infections. In accordance with this concept, our protocol is to use a tunnel of 4 cm or more as necessary. It is gratifying to see recent confirmation of this in the correspondents’ own series. Our own low CSF infection rate in the “no change” group (3.8%) in such a high risk group of patients further supports this concept.

    There is still much debate on what constitutes the most favourable tunnel length. Some would advocate a short tunnel of 4.5 cm, whereas others prefer the tunnel to reach the lower chest or upper abdomen.1 All of these documented series, including our own, had a low CSF infection rate of 3–4%, giving a long average duration of catheter placement of 11 to 18 days. In Khanna’s series1 the change to a long tunnel appeared to contain the infection rate, giving an average of 18.3 days for an indwelling catheter. In cases where a long duration of catheter placement is likely, conversion to a long tunnel may be advisable, both to reduce the infection rate and for convenience in mobilisation.

    The concept used in our paper of relating the number of ventricular catheter insertions to the CSF infection rate concurs with earlier series2 as well as that of the correspondent. It is important to investigate the possible pathogenesis and to consider viable means of achieving improved results. Results from our own data indicate that the source of infection is bacteria found in the patients’ own skin flora. Regular changing of the catheter (which in theory should reduce the opportunity for colonisation leading to infection) has not only failed to reduce infection but may even have increased it. Tunnelling may be helpful in preventing colonisation from progressing to infection. Most infections appear to be caused by resistant skin flora introduced at the time of the procedure, despite the use of standard aseptic technique and prophylactic antibiotic cover. Regular audits to ascertain the MRSA status of both the intensive care unit and operating environment are therefore of great importance.


    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.