16.1 ICP Monitoring (02.C)

Intracranial pressure (ICP) monitoring is a procedural and interpretive skill, not a sticker number to recopy onto an exam answer sheet. Items in this diagnostic-studies domain ask you to choose a device, a Brain Trauma Foundation (BTF) threshold, a waveform pattern, a zeroing landmark, or a complication. Independent OpenExamPrep teaching here covers ICP monitoring as listed under diagnostic studies and procedural skills (02.C). It is not an ABIM, ABPN, or Brain Trauma Foundation product.

Why the number is never just a number

The cranial vault is a nearly rigid box. Blood, cerebrospinal fluid (CSF), and brain tissue share a fixed volume (the Monro–Kellie relationship). A new hematoma, edema, or hydrocephalus must displace venous blood or CSF, or else ICP rises. Cerebral perfusion pressure (CPP) is mean arterial pressure (MAP) minus ICP when the references are comparable:

CPP = MAP − ICP

Adult ICP at rest is usually about 5–15 mmHg. Sustained intracranial hypertension reduces cerebral blood flow, produces ischemia, and can precipitate herniation. Monitoring exists so you can treat pressure and perfusion before the pupil changes. A single snapshot is less useful than a trend plus the waveform, the examination, and the CT.

The BTF 4th edition (2016, still the edition cited in ICU protocols) does not claim that every millimeter is equally dangerous. It recommends treating ICP above 22 mmHg because values above that level are associated with increased mortality (Level IIB). Older pocket cards still say 20 mmHg from the 3rd edition—that is a classic wrong-answer trap. Target CPP 60–70 mmHg (Level IIB). Whether 60 or 70 is the better floor depends on the patient's autoregulatory status, which you often do not know in real time. Avoid aggressive attempts to keep CPP above 70 mmHg with fluids and vasopressors because of the risk of adult respiratory distress syndrome (Level III).

Worked numbers: MAP 85 mmHg and ICP 28 mmHg give CPP 57 mmHg, below the 60–70 mmHg band. Treating only the blood pressure while the EVD is clamped and undrained leaves the numerator of perfusion fighting a high denominator. Treating only the ICP while MAP is 55 mmHg can still leave the brain ischemic.

Device choice: EVD versus parenchymal monitor

An external ventricular drain (EVD) is the gold standard for ICP measurement. A fluid-coupled catheter sits in a lateral ventricle, usually placed at Kocher's point (about 10–11 cm from the nasion and 2.5–3 cm lateral to midline in the mid-pupillary line) and aimed toward the ipsilateral foramen of Monro. Because it is fluid-filled and referenced to a transducer you can re-zero, it does not suffer the same long-term electronic drift as many parenchymal sensors. Because it is in the ventricle, it also drains CSF, treating hydrocephalus and lowering ICP, and it allows CSF sampling when ventriculitis is a concern.

An intraparenchymal monitor (strain-gauge or fiberoptic; clinical families include Codman-type strain gauges and Camino-type fiberoptic cables) is tunneled into brain tissue, often right frontal white matter. It is faster when ventricles are slit-like, it does not require a perfect ventricular hit, and infection risk is generally lower than with a ventriculostomy connected to a drainage bag for days. It cannot drain CSF. Some devices drift after several days and cannot be recalibrated in situ. Local tissue pressure can also differ from ventricular CSF pressure when mass lesions create compartments.

Epidural and subdural sensors are less accurate and are rarely the preferred answer when an EVD or parenchymal probe is available. Jugular bulb oximetry is a metabolic monitor, not an ICP substitute.

Choose an EVD when you need both a number and a treatment (hydrocephalus after aneurysmal subarachnoid hemorrhage, intraventricular hemorrhage, mass effect with obstructed CSF pathways). Choose a parenchymal probe when the goal is ICP data without a drain, ventricles are not safely cannulated, or coagulopathy makes a larger ventricular pass unattractive—understanding that you forgo CSF diversion.

When to place a monitor after TBI

BTF monitoring language is built for salvageable patients with severe TBI, defined as GCS 3–8 after resuscitation (GCS ≤8). Place an ICP monitor when that neurologic depth is accompanied by an abnormal head CT: hematomas, contusions, swelling, herniation, or compressed basal cisterns.

If the CT is normal, monitoring is still indicated when two or more of these are present at admission:

• Age >40 years
• Unilateral or bilateral motor posturing
• Systolic blood pressure <90 mmHg

Do not place a monitor solely because the patient is intubated, has a linear skull fracture, or has GCS 14 with a headache. Do not skip monitoring in a GCS 5 patient with obliterated cisterns because the family wants to wait. Exam stems often hide the two-risk-factor rule behind a CT described as unremarkable.

The BTF also states that managing severe TBI using information from ICP monitoring is recommended to reduce in-hospital and 2-week mortality (Level IIB). That is not a claim that every extra catheter-day is harmless, and it is not erased by BEST TRIP (Chesnut and colleagues, New England Journal of Medicine 2012), a strategy comparison in a specific resource setting. Treat BEST TRIP as a caution about one protocolized care pathway, not as a license to ignore herniation physiology.

ICP catheters are also used outside TBI—hydrocephalus after subarachnoid hemorrhage, intraventricular hemorrhage, selected meningitis with CSF outflow failure, and hepatic failure with cerebral edema—but the numbered TBI rules above are the ones the exam expects you to recite.

Zeroing, leveling, drain versus clamp

Level the EVD transducer at the external auditory meatus / tragus, the bedside proxy for the foramen of Monro. Do not zero a cranial drain at the phlebostatic axis (that landmark is for central venous and arterial transducers when you care about the heart). A height error of about 1 cm introduces roughly 0.7–0.8 mmHg of hydrostatic artifact; several centimeters of the bag looks fine on the bed rail can move a patient across the 22 mmHg treatment line.

Open drainage sends CSF into the drip chamber whenever ventricular pressure exceeds the height you set (for example, 10 or 15 cm H2O above the tragus). While the stopcock is open to the bag, the displayed number may reflect the circuit, not true intracranial pressure. To read ICP and inspect the waveform, briefly clamp (close the drain) so the fluid column communicates with the transducer and the intracranial space. Then reopen to the prescribed height if the therapeutic plan is drainage.

A continuously open drain without periodic clamped readings is a treatment, not a complete monitor. A permanently clamped EVD is a monitor that is not treating hydrocephalus—and it can clog. Document both the clamped ICP and the volume drained. Overdrainage can collapse ventricles, produce subdural collections, or, with a posterior fossa mass, contribute to upward herniation if you empty the supratentorial ventricles too aggressively.

Waveforms: P1, P2, P3, and Lundberg waves

A normal ICP pulse usually shows three peaks:

1. P1 (percussion wave) — arterial pulsation transmitted through the choroid plexus and large vessels. Normally the tallest peak.
2. P2 (tidal wave) — intracranial compliance. When P2 exceeds P1, compliance is reduced; the brain is on the steep part of the pressure–volume curve.
3. P3 (dicrotic wave) — related to aortic valve closure; usually the smallest.

A flattened or absent waveform suggests occlusion, air in the tubing, a clot in the catheter, a transducer problem, or a drain that is open and dumping. Troubleshoot the circuit before you give mannitol for a high ICP that is actually a damped tracing.

Lundberg A waves (plateau waves) are pathologic: ICP rises abruptly to roughly 50–80 mmHg (sometimes higher) for 5–20 minutes, then falls. They imply exhausted compensatory reserve and a high risk of ischemia or herniation. B waves occur at about 0.5–2 per minute with moderate ICP swings and suggest reduced compliance. C waves are faster (about 4–8 per minute) and often mirror Traube–Hering–Mayer blood-pressure oscillations; they are less ominous in isolation.

Complications you must name

Hemorrhage along the catheter tract is the feared immediate complication. Radiographic hemorrhage is more common than clinically important hemorrhage; risk rises with coagulopathy, antiplatelet drugs, and multiple passes. Correct platelets and INR according to your hemorrhage protocol before elective placement when time allows.

Infection / ventriculitis risk rises with catheter duration, CSF leaks around the site, and frequent sampling. Fever, unexplained CSF pleocytosis after hemorrhage, low CSF glucose, and a positive Gram stain or culture drive treatment. Antibiotic-impregnated catheters reduce infection in many series; hanging systemic antibiotics for the entire dwell without another indication is not a substitute for sterile technique and timely removal.

Malposition is diagnosed on CT: catheter in parenchyma, in the contralateral ventricle, looped, or not in CSF at all. A working drip chamber does not prove ventricular placement if you are draining bloody fluid from a hematoma cavity.

Bedside sequence the exam expects

When ICP is truly elevated on a clamped, zeroed tracing:

1. Confirm airway, head position (usually 30 degrees), neck venous drainage, and that the transducer is at the tragus.
2. Drain CSF if an EVD is in a ventricle and open drainage is safe.
3. Treat pain, agitation, fever, and seizures.
4. Osmotic therapy and brief hyperventilation as a bridge—not as a destination (avoid sustained PaCO2 <25 mmHg; crisis physiology is covered with intracranial hypertension as a disease state).
5. Reassess CPP: raising MAP is reasonable if CPP is below 60 mmHg and ICP is already being addressed; chasing CPP above 70 mmHg with volume and pressors is the ARDS trap.

ICP monitoring tells you when that ladder is failing so you can consider surgical decompression or other rescue therapy. It does not replace the neurologic examination or imaging.