Hemorrhagic complications of ventriculostomy placement: a meta-analysis.

INTRODUCTION

The reported intracerebral hemorrhage rate due to ventriculostomy placement varies widely. As studies emerge regarding alternative techniques of ventriculostomy placement, and placement by non-neurosurgeons, further definition of the true intracerebral hemorrhage rate associated with ventriculostomy is warranted. We performed a meta-analysis of the existing literature to further elucidate the incidence of intracerebral hemorrhage due to ventriculostomy.

METHODS

We performed an extensive literature search using Ovid MEDLINE and PubMed for relevant studies published after 1970. Only studies with more than 25 ventriculostomy procedures were included. Data were extracted regarding number of hemorrhages, clinically significant hemorrhages, and the use of routine post-ventriculostomy CT scanning. We performed subgroup analyses based on the use of routine post-ventriculostomy CT scanning. Chi-squared test was used to determine statistical significance.

RESULTS

Overall, 102 hemorrhagic complications from 1,790 ventriculostomies were reported, a hemorrhage rate of 5.7%. Of the 102 hemorrhages, 11 were clinically significant (clinically significant hemorrhage rate = 0.61%). In studies that used routine post-placement CT scans, the hemorrhage rate was 10.06%, compared to a hemorrhage rate of 1.53% in studies in which routine CT scans were not performed (P < 0.001). Eight clinically significant hemorrhages (0.91%) were identified in the studies utilizing routine post-procedural CT scanning, compared to three clinically significant hemorrhages (0.33%) in studies without routine CT scans (P = 0.113).

CONCLUSION

The overall hemorrhage risk associated with ventriculostomy placement based on the existing literature is 5.7%. Clinically significant hemorrhage due to ventriculostomy is less than 1%. Modifications of technique that might reduce hemorrhage risk, and the utility of routine post-procedural CT scanning, merit further investigation.