Objective: To assess the prevalence and outcome of external cerebral ventricular drainage-associated ventriculitis in neurocritical patients before and after the implementation of a bundle of external cerebral ventricular drainage-associated ventriculitis control measures.
Design: Clinical prospective case series.
Setting: University Hospital of Larissa, Greece.
Patients: Consecutive patients were recruited from the ICU of the hospital. Patient inclusion criteria included presence of external ventricular drainage and ICU stay more than 48 hours.
Intervention: The bundle of external cerebral ventricular drainage-associated ventriculitis control measures included 1) reeducation of ICU personnel on issues of infection control related to external cerebral ventricular drainage, 2) meticulous intraventricular catheter handling, 3) cerebrospinal fluid sampling only when clinically necessary, and 4) routine replacement of the drainage catheter on the seventh drainage day if the catheter was still necessary. The bundle was applied after an initial period (preintervention) where standard policy for external cerebral ventricular drainage-associated ventriculitis was established.
Measurements: External cerebral ventricular drainage-associated ventriculitis prevalence, external cerebral ventricular drainage-associated ventriculitis events per 1,000 drainage days (drain-associated infection rate), length of ICU stay, Glasgow Outcome Scale at 6 months, and risk factors for external cerebral ventricular drainage-associated ventriculitis.
Main results: Eighty-two patients entered the study in the preintervention period and 57 patients during the intervention period. During the preintervention and intervention period, external cerebral ventricular drainage-associated ventriculitis prevalence was 28% and 10.5% (p = 0.02) and drain-associated infection rate was 18 and 7.1, respectively (p = 0.0001); mean (95% CI) length of ICU stay in patients who presented external cerebral ventricular drainage-associated ventriculitis was 44.4 days (36.4-52.4 d), whereas mean (95% CI) length of ICU stay in patients who did not was 20 days (16.9-23.2 d) (p < 0.001). Furthermore, the length of ICU stay was associated with length of drainage (p = 0.0001). Therefore, the presence of external cerebral ventricular drainage-associated ventriculitis and the length of drainage were the only variables associated with a prolonged ICU stay. Unfavorable outcome in Glasgow Outcome Scale at 6 months was not associated with the presence of external cerebral ventricular drainage-associated ventriculitis (p = 0.5). No significant differences were found when Glasgow Outcome Scale was analyzed according to the two study periods.
Conclusions: The implementation of a bundle of measures for external cerebral ventricular drainage-associated ventriculitis control was associated with significantly decreased postintervention prevalence of the infection.