Temporary intraventricular catheters for managing acute obstructive hydrocephalus caused by intraventricular haemorrhage carry a high risk of developing ventriculostomy-related ventriculitis (VRV). The aim of this prospective study was to validate a new parameter for the early detection of an intraventricular infection.
Methods: Patients with external ventricular drainage due to intraventricular haemorrhage were enrolled in this prospective study. Leucocytes and erythrocytes in cerebrospinal fluid (CSF) and peripheral blood as well as bacteriological and chemical analysis of both were examined daily. The ratio of leucocytes to erythrocytes in CSF and leucocytes to erythrocytes in peripheral blood was calculated (so called cell index (CI)) and these values were compared with the "conventionally diagnosed" drain-associated ventriculitis. Furthermore, the CI values of the non-ventriculitis and ventriculitis group were compared using the t-test with adjustment for unequal variances (Welch test).
Results: Thirteen patients with an external ventricular drainage (EVD) expected to be in place for more than seven days were enrolled. Seven patients developed a bacteriologically proven VRV (time 0) within 12 days (mean 8.57). Diagnosis of VRV by CI was possible up to 3 days (mean 2.28) prior to conventional diagnosis. P values (Welch test) showed a significant difference on days -3 (P = 0.03), -2 (P = 0.03) and -1 (P = 0.012) - i.e. 3, 2 or 1 day, respectively, prior to the time point when the CSF culture grew staphylococci -, when compared with the mean cell indices of the controls, and a highly significant difference on time 0 (P < 0.001).
Conclusion: The calculated CI allows the diagnosis of nosocomial VRV in patients with intraventricular haemorrhage at a very early point of time.
Copyright 2004 Springer-Verlag