Diagnosestellung und Risikofaktoren des Delirs bei neurochirurgischen Patientinnen und Patienten

Delirium (DL) has a negative impact on the clinical outcome of patients. The incidence is 37-46% in postoperative patients and particularly high in elderly patients. Studies on patients with cerebral diseases report a lower incidence of 12-26%. We therefore assumed that DL is underappreciated in neurosurgical patients. A previous study, using the CAM-ICU-Score (CIS) showed that oral sedative drugs correlated significantly with DL. As the study was possibly underpowered, we collected more patients and looked for additional risk faktors.

We conducted a prospective study. Patients >50 years of age on a normal ward who underwent neurosurgery under anesthesia were included. We divided our patients into two groups. In one group, the CIS was collected daily to determine whether delirium was present. In the other group, no scoring method was used and the interpretation whether delirium was present was made by the medical staff. Differences between the two groups were analyzed using Fishers test. Furthermore, we correlated risk factors with the occurrence of DL using Spearman´s correlation and conducted a multivariate regression.

213 patients were prospectively included. The overall incidence of DL was 9.4% (20 patients). In the group in which CIS was used, it was higher at 12.7% (14 patients) than without the use of CIS (5.8%, 6 patients). The difference was not statistically significant (p=0.102). Risk factors for the occurrence of DL were: cerebral disease (vs. spine; [p=0.003]), higher age (p=0.008), oncological disease (p=0.001), postoperative ICU stay (p=0.012), elevated leukocytes (p=0.029), neurodegenerative diseases (p=0, 016), alcohol abuse (p=0.006), hypacusis (p= 0.03), liver disease (p=0.012), and the use of sedatives (p=0.010), antiarrhythmics (p= 0.010) and steroids (p=0.008). A multivariate analysis only revealed a significant correlation with the occurrence of DL for age (p=0.047), cerebral disease (vs. spine) (p=0.014) and alcohol abuse (p=0.032).

The application of CIS seems to lead to a more than two-fold incidence of DL, however, statistical analysis failed to show a significance. The current data indicate that typical risk factors do play an important role. The fact that oncological diseases, cerebral operations, and the use of steroids are significantly associated with the occurrence of delirium, suggests, that patients with malignant brain tumors might be particularly at risk even if this could not be confirmed in the multivariate analysis.