Intracranial pressure (ICP) is monitored using either invasive devices placed inside the skull or non-invasive techniques that estimate pressure indirectly. Normal ICP in adults ranges from 5 to 15 mmHg, and sustained readings above 20 to 22 mmHg typically require aggressive treatment. The method chosen depends on the clinical situation, the need for continuous readings, and whether draining cerebrospinal fluid (CSF) is also part of the treatment plan.
What the Numbers Mean
In healthy adults and children over age 8, ICP sits between 5 and 15 mmHg when lying down. Younger children run lower, around 1 to 7 mmHg, and obese patients can have baseline readings up to 18 mmHg without anything being wrong. Once pressure climbs above 20 mmHg, it can start damaging brain tissue and reducing blood flow to the brain.
The 2016 Brain Trauma Foundation guidelines set the treatment threshold at 22 mmHg, up from the previous target of 20 mmHg, based on research linking that level to increased mortality. That said, more recent studies have questioned whether a single universal cutoff makes sense. One large analysis found that an ICP limit of 18 mmHg better predicted both survival and outcomes in their patient population. In practice, clinicians treat trends and the full clinical picture rather than reacting to a single number.
Pressure in the 20 to 30 mmHg range is considered mildly elevated. Sustained ICP above 40 mmHg is severe and life-threatening.
When ICP Monitoring Is Needed
The most established indication is severe traumatic brain injury, defined as a Glasgow Coma Scale score of 3 to 8 after resuscitation, combined with an abnormal CT scan showing hematomas, contusions, brain swelling, herniation, or compressed fluid spaces at the base of the brain. Even with a normal CT scan, monitoring may be warranted if two or more of these factors are present at admission: age over 40, abnormal motor posturing on one or both sides, or systolic blood pressure below 90 mmHg.
Beyond trauma, ICP monitoring is commonly used in patients with bleeding into the brain’s fluid-filled ventricles, subarachnoid hemorrhage, brain tumors causing fluid buildup, central nervous system infections, and shunt malfunctions in patients who already have implanted drainage systems.
External Ventricular Drain (EVD)
The external ventricular drain has long been considered the gold standard for ICP monitoring. A thin catheter is inserted through a small hole in the skull and threaded into one of the brain’s ventricles, the fluid-filled chambers where CSF is produced. From there, it connects to an external collection system that both displays the pressure reading and allows excess fluid to be drained when needed.
This dual capability is the EVD’s biggest advantage. It doesn’t just measure the problem; it can treat it. When ICP spikes, opening the drain allows CSF to flow out, immediately lowering pressure. The system works on a simple fluid-filled column principle: changes in pressure inside the ventricle travel through saline in the tubing to a sensor on the outside. To stay accurate, the system needs to be re-zeroed regularly, and the drain must be closed to drainage during pressure readings.
The downsides are real. EVDs carry an infection rate of roughly 10%, based on large retrospective reviews. They can also become blocked by blood clots or brain tissue. CSF samples can be drawn from the drain and tested for signs of meningitis, which helps catch infections early. EVD catheters are MRI-compatible, which is a practical advantage when follow-up brain imaging is needed.
Intraparenchymal Monitors
These are small sensors placed directly into the brain tissue itself, rather than into a ventricle. They come in two main types: fiberoptic sensors that detect pressure changes using light, and strain-gauge sensors that measure pressure through tiny changes in electrical resistance across a diaphragm at the catheter tip.
Fiberoptic catheters can also be placed in the subdural space (just beneath the skull’s inner lining) rather than deep in the brain tissue. Both types provide continuous pressure readings and are generally easier to place than EVDs, since hitting a ventricle with a catheter can be technically challenging, especially in patients whose ventricles are compressed by swelling.
The tradeoff is that these monitors can only measure pressure. They cannot drain CSF, so they offer no direct treatment benefit. They also have a known issue called zero-drift, where the baseline reading gradually shifts over days, potentially leading to inaccurate numbers and misguided treatment decisions. Fiberoptic sensors can be damaged by kinking or bending. And unlike EVDs, intraparenchymal monitors are typically not MRI-compatible, which can limit imaging options during a patient’s hospital stay.
Reading the ICP Waveform
Invasive monitors don’t just display a single number. They produce a continuous waveform with three distinct peaks that carry diagnostic information:
- P1 (percussion wave): caused by the pulse of arterial blood, specifically from the choroid plexus inside the ventricles
- P2 (tidal wave): reflects the brain’s compliance, meaning how well it can absorb increases in volume without a large jump in pressure
- P3 (dicrotic wave): corresponds to the closure of the aortic valve in the heart
In a healthy brain, P1 is the tallest peak. As the brain loses compliance and pressure rises, P2 gradually grows taller than P1, and the waveform takes on a rounded appearance. This shape change is often an early warning sign, sometimes appearing before the numerical ICP reading crosses a dangerous threshold. When compliance deteriorates further, large slow waves called Lundberg A waves (or plateau waves) appear, signaling dangerously high pressure that demands immediate intervention.
Non-Invasive Methods
Optic Nerve Sheath Ultrasound
The optic nerve is wrapped in a sheath that communicates directly with the fluid spaces around the brain. When ICP rises, this sheath expands. Using a standard bedside ultrasound probe placed gently over a closed eyelid, clinicians can measure the diameter of the optic nerve sheath about 3 millimeters behind the eyeball. A diameter greater than 5.6 mm suggests elevated ICP, with a sensitivity of about 94% and a specificity of 87% for detecting pressure above 20 mmHg. This makes it one of the more reliable bedside screening tools when invasive monitoring isn’t yet in place or isn’t practical.
Transcranial Doppler Ultrasound
Transcranial Doppler measures the speed of blood flowing through the brain’s major arteries by aiming an ultrasound probe at specific thin spots on the skull. The key measurement is the pulsatility index, which captures the difference between peak and minimum blood flow speeds relative to the average. A strong correlation exists between the pulsatility index and actual ICP: one study of neurocritical care patients found a correlation coefficient of 0.938, with the ability to estimate ICP within about 2.5 mmHg across a pressure range of 5 to 40 mmHg. This technique is most useful as a screening tool to help decide whether a patient needs an invasive monitor placed.
Automated Pupillometry
Handheld infrared pupillometers measure how the pupil reacts to light and generate a Neurological Pupil Index (NPi) score. The idea is that rising ICP compresses the nerve controlling pupil size, producing sluggish or absent reactions. While the concept is sound, a large secondary analysis from the ORANGE study found that NPi values were not significantly associated with ICP values across a broad population of brain injury patients. An abnormal NPi may still indicate brainstem problems, but it does not reliably substitute for direct pressure measurement. Pupillometry is better understood as one piece of the neurological exam rather than a standalone ICP estimate.
Complications of Invasive Monitoring
Any device placed through the skull carries risks. Infection is the most common, occurring in roughly 10% of cases in large trauma series. The risk increases the longer the monitor stays in place, which is why most centers aim to remove devices as soon as they’re no longer clinically necessary. Bleeding during or after insertion is the other major concern, though clinically significant hemorrhages are less common than infections. Both EVDs and bolt-type monitors require careful daily assessment of the insertion site, the waveform quality, and the patient’s neurological status to catch problems early.