Introduction: Intraventricular hemorrhage (IVH) occurs in about 40% of patients with intracerebral hemorrhage (ICH) and is associated with higher mortality and worse outcomes than ICH patients without IVH. Infections are common in ICH patients but data in IVH patients are limited.
Methods: Prospective analysis of adjudicated adverse event infection reporting during first 180 days in 500 patients enrolled in the CLEAR III trial, a multicenter, double-blind, randomized study comparing external ventricular drain (EVD) + intraventricular recombinant tissue plasminogen activator (rtPA) vs. EVD + placebo for treatment of obstructive IVH and intracerebral hemorrhage (ICH) volume <30cc. Primary outcome measures were 90-day and 180-day mortality. Secondary outcome measures were hospital length of stay (LOS). We constructed binary logistic and linear regression models for multivariable analysis.
Results: Infection was reported in 269 patients (53.8%). Pneumonia was the most common infection (33%), followed by UTI (16%), and bacterial ventriculitis (4.4%). Overall 180-day mortality was 20%. Patients with infection were more likely to have older age (p=0.012), lower admission GCS (p=0.007), higher ICH volume (8.8 vs 6.7ml, p=0.001), and higher ICH+IVH volume (37.7 vs 31.7 ml, p=0.002). In the regression model, IVH volume was associated with higher odds of 90-day or 180-day mortality, but presence of any infection was not a significant predictor of mortality. Infection was however associated with longer length of stay (26 vs 22 days, p<0.001). Subgroup analysis of individual infections, showed only bacterial ventriculitis to be associated with 90-day (OR: 3.84, CI: 1.36-10.82), and 180-day mortality (OR: 2.9, CI: 1.05-8.06), while pneumonia and UTI were not.
Conclusion: Patients with IVH have a high incidence of infections, which is associated with longer hospitalization but does not appear to influence mortality. Of the infections, bacterial ventriculitis is a significant predictor of mortality in our 7-factor model. IVH volume did not predict infections but predicted mortality.These results form a basis for future correlation of infectious complications with treatment rendered (thrombolysis versus placebo), with upcoming unblinding of the trial.