Guest Post by The NeuroEMCrit Team (Casey Albin, MD @caseyalbin & Neha Dangayach, MD @drdangayach)
Before diving into the use, maintenance, and complications of external ventricular drains (EVDs), realize there are several ways to monitor intracranial pressure, both non-invasively and invasively.
Non-invasive strategies tend to provide a gestalt of the intracranial pressure – normal vs high with proxies for the absolute value of the intracranial pressure. These are explored in separate posts:
💡Transcranial Dopplers (coming soon!)
💡Pupillometry (coming soon!)
Invasive ICP monitoring is most often achieved through either a:
💡Fiberoptic parenchymal probe
The EVD has two major advantages: the ability to treat elevated ICP via CSF diversion and the ability to sample CSF.
Note, though, EVDs are not the only way to accomplish CSF diversion. Lumbar drains can also be used for CSF diversion and can be used to give ICP data (PMID: 20953846) although EVDs are still considered the gold standard. If used, ICPs from LD need to be recorded with the patient in the supine position.
What Every Clinician Needs to Know about EVDs
(scroll to the bottom for highest yield take-aways)
Who should get an EVD can be thought of in four categories:
The Set Up:
How are They Placed?
EVDs are commonly inserted at Kocher’s Point, which is approximately 10 cm posterior to the nasion and 3cm to the right (or left) of midline, approximately at the midpupillary line. However, there are alternative locations, such as the Frazier's point, which is used for a parietal-occipital approach.
A sterile catheter is advanced through the skull, dura, and brain parenchyma into the ipsilateral lateral ventricle ideally to the Foramen of Monro (intraventricular foramen).
Note there are several holes in the EVD catheter which can be visualized in the CT by windowing the CT scan to the bone window:
An opening pressure will be obtained on placement. Anything >20 mmHg (~27 cmH2O) should trigger management of intracranial hypertension
Which Side is Best?
- If there is no contraindication, a right frontal approach is preferred. As long as there is no compression of the 3rd ventricle, drainage of one of the lateral ventricles results in CSF diversion from the entire CSF system (including the contralateral side).
- However, if there is compression of the 3rd ventricle, drainage may be unilateral to the side that has the EVD. Drainage of an entrapped ventricle may inadverently lead to a worsening shift / herniation
- If the purpose of the EVD is to promote clearance of blood by administering IT-TPA, EVD placement in the “bloodier” ventricle may result in a more rapid clearance of the IVH (PMID 22067423)
- If there is thick intraventricular blood, you may need to discuss placement of bilateral EVDs with NSGY to relieve hydrocephalus
How Does an EVD Accomplish Drainage?
- The EVD catheter is connected via a sterile catheter to a collecting system
- The chamber is zero’ed to the level of external auditory canal (see “Set Up” diagram above)
- The chamber is then set at a height relative to the zero point.
For example: Here is an EVD set at 0 cm H2O and then raised to 20 cm H2O
Most institutions will set the EVD in cmH2O (conventionally ventriculoperitoneal shunts (VPS) are set in cmH2O). However, you can also set the threshold in mmHg.
- The amount of CSF drainage depends on the pressure difference between the ICP and the height of the chamber
- For example:
- EVD is set to 5 cmH2O: When ICP is greater than 5 cmH2O, CSF will flow into the lower pressure system (the EVD chamber). As such, a low setting like this facilitates drainage (as ICPs are often >5cmH2O).
- EVD is set to 20 cmH2O (~15mmHg): Although ICP may be above this number without being consider pathologic, there should be less drainage as the downstream resistance to drainage is higher. This placement can be considered a “popoff valve” – allowing CSF to escape at a lower threshold than dangerous, but this setting does not encourage drainage.
- A very important practical consideration: YOU SHOULD NEVER ADJUST THE HEAD OF THE BED OR HEAD OF BED WITHOUT CLAMPING THE EVD! Why? Imagine the EVD chamber is set to 5 cmH2O. You come in and want to do an arterial line and raise the bed up… You are essentially dropping the EVD – which means there will be a much lower threshold to drainage, which means the patient could dump CSF, which is dangerous!
What Determines The EVD's Initial Setting?
This is often a conversation with NSGY and per institutional preference. Some guidance:
- EVD at a low setting (~5cmH2O) (reported as “Open at 5”) when you want to encourage drainage, as in the treatment of hydrocephalus, clearance of blood/purulence material
- EVD at a moderate setting (~10cmH2O) when it is primarily placed as an ICP monitor, there is some advantage in draining a moderate amount of CSF
- EVD at a high setting (15-20cmH2O) before an aneurysm is secured, as a weaning setting, or when there is a posterior fossa mass (to prevent upward herniation)
What Data Does and EVD Give You?
- ICP data is displayed on the monitor in mmHg as those are the units which the Guidelines like the NCS and BTF use when discussing the threshold to trigger management of intracranial hypertension
- For most EVDs, the displayed ICP is only accurate when the EVD is clamped! This can be confusing because there may be a number on the monitor even when the EVD is open. But, unless you have a EVD that is built to do continuous ICP monitoring (eg the Hummingbird EVD), the number and waveform displayed with an open EVD are inaccurate.
- If you want to check the ICP or look at the waveform, you need to clamp the EVD (and then remember to unclamp it when you are done!)
Example of EVD Open (⬇️) EVD Clamped (⬇️):
- The CSF waveform is another window into the compliance of the intracranial vault.
- This can be obtained from the EVD when it is clamped.
- When the brain’s compliance is decreased, the P2 wave will be higher than P1. At this point, very small fluctuations in volume can lead to dramatic increases in pressure. A nice little way to remember this is that the “angry” compliance waveform appears to be giving you the “middle finger” 😂
- Often our aim in measuring ICP is to understand the Cerebral Perfusion Pressure (CPP) so as to avoid secondary ischemic insults
- Reminder that CPP = MAP – ICP
- The ICP value is collected from the clamped EVD and the MAP from the A-line. However, the question becomes — does an A-line value leveled at the phlebostatic axis (standard practice) reflect the arterial pressure in the brain?
- When the head of the bed is elevated above the heart, the arterial pressure seen by the brain is lower because of hydrostatic forces (the magnitude of difference depends on the angle of elevation, ie this is a bigger deal in the patient sitting upright than in the HOB at 30 degrees patient)
- Several papers have investigated the magnitude and importance of this difference (PMID: 30906677) and found there can be a >10 mmHg pressure difference even with HOB 30, and a greater difference in patients in chair position (PMID: 24262017) – a potentially meaningful in patients elevated ICPs or borderline MAPs
- To most accurately reflect the blood pressure seen by the cerebral arteries, the A-line should be leveled to the tragus when using the arterial line to calculate the CPP (and then releveled to the fourth intercostal space when you are interested in guiding resuscitation of the rest of the body… or to have two separately leveled reading as suggested by some authors (PMID: 27440648))
- Note though this difference is likely more important in patients sitting bolt upright and has not been widely enforced in terms of guideline for ICP management leading to clinical variability in this practice (PMID: 24262017)
- An EVD also gives you the ability to monitor and test the CSF, whether for infection, cancer (CSF cytology), or inflammatory markers.
Types of EVDS:
What are EVD Complications?
Ventriculostomy-associated infections (VAI): complicate ~5-10% (0-22%) of EVDs (PMID: 29088962), length of time catheter is in place is a risk factor. Note that diagnosis can be difficult as quite often patients with EVDs have inflammatory CSF with elevated protein and pleocytosis; rarely is this ventriculitis. But it is something you need to keep in mind for all patients with an EVD. Here's a table to help you think through the diagnosis and management:
- Tract hemorrhage may complicate up to ~30% of EVDs placement, but the clinically significant rate is low. (PMID:24410156)
- However, even if the bleeding itself is asymptomatic, it may put the patient at risk for GNR ventriculitis (PMID: 31294451)
- Most bleeding complications occur during placement; less commonly, removal.
- Data that has not been reliably reproduced suggests that older age, pre-placement antithrombotic use and INR >1.4 may be risk factors.
- Other probable contributing factors include anti-platelet use within 96 hours of placement (PMID:29514640) and number of attempts at the time of placement (PMID:29514640)
- Although not routinely recommended, some hospitals will check anti-Xa level to confirm that DVT prophylaxis is appropriately dosed (i.e. the Anti-Xa level isn't creeping up into a treatment range)
To think about this … first, think, how much CSF is produce in a day?
About a Coke can amount! That’s somewhere between 300-600 mL daily, about 20ml/hour. Thus, if you drain more than that (consistently draining more than 20mL an hour), you risk over-draining which can lead to subdurals/hygromas and potentially to intracranial hypotension or upward herniation (the later is especially true when dealing with posterior fossa masses).
Given how EVDs work, the CSF should stop draining when the ICP is less than the chamber height, but if the head or height of the bed is adjusted without clamping the EVD…. 🙁 .
Again, ALWAYS CLAMP THE EVD WHEN MOVING THE PATIENT OR BED.
What if the catheter stops draining?
Abnormal physiology that causes this includes:
✨The blood the EVD is draining is thick/coagulated
✨Tissue debris in the system
✨The ventricular wall is collapsed around the catheter
The latter three tend to dampen the ICP waveform.
***DO NOT PROCEED WITH EVD TROUBLESHOOTING WITHOUT PERMISSION OF THE SERVICE THAT PLACED IT ***
STEP 1️⃣: Make sure there are no kinks in the catheters and inspect the insertion site to make sure that the sutures are not obstructing drainage
STEP 2️⃣: Check the catheter’s patency. Lower the drain carefully to 0 cmH2O (or lower). If it starts to drip a few drops of CSF, then it’s not a catheter problem, it’s a low ICP problem. This may result from over-drainage, and counterintuitively, raising the EVD may lead to re-plumping of the ventricles and, later, further drainage.
STEP 3️⃣: Call NSGY or the provider that is managing the drain to the bedside. The next steps involve flushing the catheter first AWAY from the patient and then (with a very small volume of sterile saline) TO the patient. These steps should ONLY be done with whatever service has placed the drain or under the supervision of someone that has experience with EVDs as these steps require aseptic technique and have the potential to worsening ICPs or cause a tract hemorrhage if not done correctly.
When to take the EVD out?
- Weaning or clamp trials should not be started until there is a significant improvement in the pathology that led to placement of the EVD (ICPs are controlled, IVH clearance is improved, meningitis is improving, the dura has healed, etc).
- The major question in SAH management has been whether an EVD should be “weaned” or if readiness for removal should be judged by “intermittent clamping.” In aSAH early clamping and intermittent drainage was associated with lower rates of EVD malfunction (PMID: 23957382) and reduced length of stay (PMID: 15086228).
- A recent (published Sept 2021) multicenter prospective trial confirmed that a rapid weaning protocol was associated with fewer days of EVD use and lower rate of VPS placement (PMID: 34498207). Such studies have led to a call for earlier intermittent clamp trials to determine readiness to d/c the EVD (PMID: 28929378).
- Regardless of this data, most institutions still practice a gradual weaning where the EVD is moved from 5 to 20 cmH2O over the course of several days before being clamped.
- Although not always needed, a head CT before “weaning” may be helpful to have as a comparison for the ventricle’s caliber before clamp and 24 hours after clamping. If ventricular caliber is unchanged, ICPs have not been sustained >20 mmHg, and the patient symptomatically has no evidence of intracranial hypertension (no worsening headache, AMS, vomiting etc), EVD can likely be removed.
- There continues to be a lot of variabilities in practice in how to maintain and wean EVDs. So, be aware of your institutional practices (PMID: 28000129), (PMID: 31773310)
Management of EVDs – The Summary
- NeuroEMCrit – Demystifying the EEG Report - November 11, 2021
- NeuroEMCrit – What Every Clinician Should Know about External Ventricular Drains (EVDs) - September 16, 2021