Citation: Deleu T, Duerinckx J, Noppeney R, Menovsky T. A comparison between proximal and distal cerebrospinal fluid sampling sites in patients with external ventricular drains. Neurosurgery. 2025;97(4):786-792. doi:10.1227/neu.0000000000003482

Background:

Sampling of cerebrospinal fluid (CSF) from external ventricular drains (EVDs) can be performed either proximally (directly from the intraventricular space via a port near the skull) or distally (from the CSF collection bag). Proximal sampling is considered the gold standard due to concerns about the accuracy of distal sampling, but may carry a higher risk of iatrogenic infection, obstruction, or hemorrhage.

Methods:

This prospective, single-center study compared CSF laboratory findings from proximal versus distal collection methods in patients with an EVD expected to remain in place ≥48 hours. Samples were collected twice weekly under routine conditions and more frequently when clinically indicated (e.g., suspected infection). Paired proximal and distal samples were collected simultaneously using the same EVD system (Codman 82-1731, Integra Lifesciences). To address sedimentation in distal samples noted in prior studies, investigators collected all drip chamber contents, mixed samples by rotating the aspiration syringe ten times, and replaced the drip chamber at least every two hours per institutional protocol.

Primary outcomes were intraclass correlation coefficient (ICC), within-subject standard deviation (SD), and mean absolute difference (MAD) in CSF white blood cell (WBC) count between proximal and distal samples. Secondary outcomes included agreement in lactate, glucose, total protein, and microbial cultures, as well as comparisons by EVD duration (<2 weeks vs >2 weeks).

Results:

Sixty-seven patients provided 394 paired samples; 36 patients had more than two sampling time points, 12 had two, six had one, and seven had none (due to death or withdrawal of consent). The mean age was 47 years, including nine patients under the age of 18. Diagnoses of those who were sampled included nontraumatic hemorrhage (28), trauma (16), infection (12), tumor (4), and ischemic stroke (1). Three types of antibiotic-impregnated catheters were used, and no sampling-related adverse events occurred.

Proximal and distal WBC counts showed strong agreement (ICC 0.731; 95% CI: 0.650–0.796), consistent across all time intervals (days 1–3, 4–7, 8–14, and >14) and secondary measures. Microbial cultures demonstrated perfect concordance, though only 10 of 394 samples were positive. Excluding patients with very high WBC values improved agreement, as large MAD values were concentrated among these samples. The authors note that in clinical practice, detecting infection may be more meaningful than precise leukocyte quantification.

Commentary:

The authors conclude that distal CSF sampling from EVD collection systems is a viable alternative to proximal sampling, offering convenience and a potential reduction in iatrogenic risk. Distal sampling can also be performed by non-neurosurgical providers, including nursing staff.

Strengths of the study include broad inclusion criteria (all ages and pathologies) and consistent agreement across time points. Limitations include the single-center design, small number of positive cultures, and use of a single EVD system, which may limit generalizability. A sub-analysis by catheter type would have strengthened the findings.

The authors addressed prior studies that deemed distal sampling unreliable by correcting methodological issues, particularly sedimentation, through detailed sampling and mixing protocols. Their results suggest that distal sampling is reliable when these techniques are applied and can be implemented successfully by multidisciplinary staff. 

While no sampling-related complications were observed, the study’s rationale is to further reduce procedural risk; although antibiotic-impregnated catheters already lower infection rates, distal sampling may offer an additional safeguard.

Impact on Clinical Practice:

Distal CSF sampling from EVD systems appears to be a safe and accurate alternative to proximal sampling. Meticulous sampling techniques, including collection from the drip chamber and thorough mixing, may be necessary to ensure reliability and avoid errors related to sedimentation.