Clinical characteristics, microbiology and outcomes of external ventricular drainage-associated infections: The importance of active treatment

Ventriculostomy Associated with Reduced Mortality in Severe Traumatic Brain Injury Compared to Parenchymal ICP Monitoring: A Propensity Score-Adjusted Analysis

BACKGROUND

There is a lack of data on whether intracranial pressure (ICP)-guided therapy with an intraparenchymal fiberoptic monitor (IPM) or an external ventricular drain (EVD) leads to superior outcomes. Our goal is to determine the relationship between ICP-guided therapy with an EVD or IPM and mortality.

METHODS

Retrospective analysis of severe traumatic brain injury cases that required IPM or EVD placement for ICP-guided therapy from January 1, 2010 to December 31, 2020. The data were obtained from the Pennsylvania Trauma Systems Foundation registry.

RESULTS

A total of 2305 patients met the inclusion criteria, with 1048 (45.5%) IPM and 1257 (54.5%) EVD placed. Inpatient mortality occurred in 337 (32.2%) and 334 (26.6%) patients in the IPM and EVD cohorts, respectively (P = 0.003). Even among those treated medically only, inpatient mortality occurred in 171 (30.8%) of those with an IPM and in 100 (23.4%) of those with an EVD (P = 0.010). Multivariable logistic regression analysis showed that older age (odds ratio [OR] 1.03, P < 0.001), lower Glasgow Coma Scale (GCS) score (OR 1.16, P < 0.001), requiring surgery (OR 1.22, P = 0.049), and an IPM (OR 1.40, P = 0.001) were significant predictors of mortality. Propensity score-adjusted analysis using inverse probability of treatment weighted method revealed a 28% decrease in mortality and a 14% decrease in length of hospital stay with EVD use when adjusting for age, sex, GCS, Injury Severity Score, surgery, and Hispanic ethnicity.

CONCLUSIONS

A significant mortality benefit was associated with the use of EVD compared to IPM. This mortality benefit was observed regardless of whether patients required surgery or not.

Intracranial Pressure Monitoring for Acute Brain Injured Patients: When, How, What Should We Monitor

While there is no level I recommendation for intracranial pressure (ICP) monitoring, it is typically indicated for patients with severe traumatic brain injury (TBI) with a Glasgow Coma Scale (GCS) score of 3-8 (class II). Even for moderate TBI patients with GCS 9-12, ICP monitoring should be considered for risk of increased ICP. The impact of ICP monitoring on patient outcomes is still not well-established, but recent studies reported a reduction of early mortality (class III) in TBI patients. There is no standard protocol for the application of ICP monitoring. In cases where cerebrospinal fluid drainage is required, an external ventricular drain is commonly used. In other cases, parenchymal ICP monitoring devices are generally employed. Subdural or non-invasive forms are not suitable for ICP monitoring. The mean value of ICP is the parameter recommended for observation in many guidelines. In TBI, values above 22 mmHg are associated with increased mortality. However, recent studies proposed various parameters including cumulative time with ICP above 20 mmHg (pressure-time dose), pressure reactivity index, ICP waveform characteristics (pulse amplitude of ICP, mean ICP wave amplitude), and the compensatory reserve of the brain (reserve-amplitude-pressure), which are useful in predicting patient outcomes and guiding treatment. Further research is required for validation of these parameters compared to simple ICP monitoring.

The Use of Intraventricular Instillation of Vancomycin to Prevent External Ventricular Drainage Related Infection: A Clinical Prospective Study

BACKGROUND

External ventricular drainage (EVD)-related infection (ERI) represents an important condition with potential high morbidity with significant impact on patient outcomes. Prophylactic systemic antibiotics are routinely administered to patients with EVD, but they do not significantly lower the incidence of ERIs. Intraventricular treatment with vancomycin appeared to be safe and effective, but most reports are case-reports/-series and retrospective studies.

METHODS

A prospective non-randomized case-control study was conducted in a consecutive series of 116 patients treated with EVD insertion. The study includes the group of patients treated with intrathecal vancomycin (Group A, 62 patients) compared with the control group treated with daily intravenous cefazolin (Group B, 54 patients).

RESULTS

No statistically significant differences were found between the 2 groups with regard to the duration of catheterization and occurrence of ERI during hospitalization. EVD was replaced in 16 cases (25.8%) in group A and in 12 cases (22.2%) in the control group B (P 0.67). Three cases (4.8%) of ERI have been found in group A and 5 (9.3%) in the control group (P = 0.34). All reported cases of infection in group A were caused by gram-negative agents; on the opposite, cases of infections in the control group B were caused above all by gram-positive bacteria with a statistical difference (P = 0.03).

CONCLUSIONS

In this first prospective study on this topic, we found that intrathecal Vancomycin administration in EVDs does not reduce the occurrence of ERI compared with intravenous cefazolin prophylaxis, but induces selection of gram-negative bacteria.

Cerebrospinal Fluid System Infection in Children with Cancer: A Retrospective Analysis over 14 Years in a Major European Pediatric Cancer Center

Infection of a cerebrospinal fluid system is a serious medical complication. We performed a retrospective monocentric analysis on temporary and permanent cerebrospinal fluid devices in children with and without cancer, covering a period of over 14 years. Between 2004 and 2017, 275 children with a cerebrospinal fluid system were seen at our institution. Thirty-eight children suffered from 51 microbiologically proven infectious episodes of the cerebrospinal fluid system (12 children with cancer and 26 children without cancer). Independently of the cerebrospinal fluid system used, the incidence of infection did not significantly differ between children with and without cancer and was the highest in children younger than one year. Infection occurred earlier in external ventricular drain (EVD) than ventriculoperitoneal (VP) shunt, and in EVD significantly earlier in children with cancer compared with patients without cancer. The pathogens isolated were mainly Gram-positive bacteria, in particular Staphylococcus spp., which should be taken into account for empirical antimicrobial therapy.

Is External Ventricular Drain Infection Predictable in Neurosurgery Patients? A Following-up Study in Iran

Background: External ventricular drains (EVDs) infection is a life-threatening complication. Objectives: To investigate the rate of EVDs infection and its predictive factors in neurosurgery patients hospitalized at a tertiary teaching hospital in Iran. Methods: In this survey, all patients referred to a subspecialty hospital in Tehran (Iran) with an external ventricular drain (EVDs) from Jun 23, 2018, to Jun 23, 2019, are monitored within 30 days of EVDs placement for infection. Data on demographic information, underlying diseases, number of EVD replacements, length of hospital stay, type consumed antibiotic before neurosurgery, length of tunneling, type of airway, duration of mechanical ventilation, duration of surgery(hours), surgeon name, APACHE II score, length of intensive care unit stay are collected. The diagnosis was based on the Center for Diseases Control and Prevention criteria for meningitis/ventriculitis and clinical vision of infectious or critical care specialists. A logistic regression model was developed to identify factors that can predict the infection. Results: Of 81 patients with EVDs, 39 (48.1%) were infected. The mean age of patients was 44.33 ± 19.5 years, and 55.6% of them were male. According to the multiple logistic regression analysis, mechanical ventilation for more than 6 days (OR: 2.5, 95% CI: 1.6 - 3.56, P = 0.04) and length of tunneling > 5 cm (OR: 1.98, 95%CI: 1.87 - 4.76, P = 0.02) were identified as factors that could predict EVD infection. Also, consuming ceftazidime + vancomycin, as a prophylaxis agent, had a lower odds ratio for EVD infection (OR:0.4, 95% CI: 0.08 - 0.84, P = 0.04). Conclusions: Regarding the predictive factors of EVDs infection, either in the present study or other studies, there should be strategies to manage this life-threatening infection in neurosurgery patients.