Sitting the patient up drains the surgical field, opens the anatomy, and quietly subtracts a fixed number of millimetres of mercury from the pressure reaching the brain. The subtraction is arithmetic, it is invisible on the monitor unless you do one thing, and it is a bigger threat to most patients than the complication this position is famous for.
For every 1.25 cm the head is elevated above the heart, local arterial pressure falls by about 1 mm Hg. A head 25 cm above the transducer therefore sees roughly 20 mm Hg less than the monitor is showing, and a mean of 65 displayed is a mean of 45 at the circle of Willis. The fix is one habit — zero the arterial transducer at the level of the skull base, and re-zero with every change of position. The famous complication of this position is venous air embolism, which is detected in up to 60% of cases but is clinically significant in about 2.7%. The pressure error happens in every case.
Every position in neurosurgery is a trade. This one is the most explicit trade in the book, and it is worth being able to state both sides of it before you are asked to agree to it.
The posterior fossa is the deepest of the three cranial fossae, between the tentorium and the foramen magnum, holding the cerebellum and the brainstem. In adults the commonest suboccipital craniotomy is for a cerebellopontine angle tumour — acoustic neuroma or meningioma — with metastases, aneurysms and haemorrhage behind it.
Elevating the head reduces dural sinus pressure by up to 10 mm Hg at 90 degrees. That single fact produces both halves of the trade at once: less venous bleeding, and a pressure gradient that draws air into any open venous channel.
The advantages that belong specifically to the anaesthetist are worth listing, because they are usually left out of the argument:
Improved postoperative cranial nerve function has been reported after acoustic neuroma resection in the sitting position compared with horizontal ones.
The costs are cardiovascular and hydrostatic: increased pulmonary and systemic vascular resistance, decreased cardiac output, decreased venous return, decreased cerebral perfusion pressure. Vital capacity and functional residual capacity improve, but hypovolaemia can underperfuse the upper lung and create ventilation–perfusion mismatch.
Relative contraindications are specific: known intracardiac septal defects, known pulmonary arteriovenous malformations, severe hypovolaemia, cachexia, and severe hydrocephalus.
This matters because angles get quoted without the position they belong to, and then they collide.
The full sitting position puts the head in a three-pin holder with the trunk upright; the source’s hydrostatic figure of a 10 mm Hg fall in dural sinus pressure is quoted for 90 degrees. The “lounge chair” modification raises the thoracic cage 30 to 45 degrees and is used for lateral lesions, though access to midline structures may be limited by the neck flexion required. The park bench lateral allows the head to be lowered rapidly to left lateral decubitus if hypotension or persistent air embolism occurs — which is a design feature, not a coincidence.
And none of these is the 15 to 30 degrees of head elevation used to aid venous drainage in a supine or lateral craniotomy, which is a different manoeuvre in a different operation.
So: name the position before quoting the angle. Two correct numbers from two different positions read as a contradiction if the position is dropped, and this is one of the places that reliably happens.
The details here are not fussiness. Each one prevents a named injury.
Zero the arterial transducer at the skull base, during positioning and throughout. That is today’s pitfall.
At least 2.5 cm between chin and chest. Excessive neck flexion stretches the cervical cord — midcervical quadriplegia is a described complication of this position — and obstructs venous drainage from the face and tongue, which produces the macroglossia and facial and neck oedema that cause postoperative airway obstruction. For the same reason, no large oral airway or bite block in the pharynx.
Avoid excessive neck rotation, particularly in older patients.
Avoid excessive knee flexion toward the chest, which compresses the abdomen, risks lower limb ischaemia and injures the sciatic nerve.
Pad the elbows, avoid brachial plexus stretch, and keep pressure off the common peroneal nerve just distal and lateral to the head of the fibula.
Check the tracheal tube after positioning. Neck flexion or extension displaces the tube caudad or cephalad by as much as 2 cm. Palpating the cuff above the sternal notch confirms the tip has not descended toward the carina, though this is not possible with every tube design.
Seal the venous entry sites. Central catheters inserted via the neck or subclavian route should have their insertion sites sealed with bacteriostatic ointment and dressing to minimise air entrainment. Place and remove central lines supine or slightly head-down — air embolism has been reported at line removal with the head elevated. And prolonged head-down positioning or head rotation during internal jugular cannulation should be minimised, because both reduce cerebral perfusion.
A patent foramen ovale exists in 20 to 30% of the population on autopsy, and transoesophageal echocardiography detects it in 10 to 30%. Specificity is 64 to 100% — but sensitivity is poor, and non-detection does not guarantee absence. Some institutions treat a known patent foramen ovale as a relative contraindication to sitting.
The reason to look is paradoxical air embolism, and the honest position is that it is far rarer than the arithmetic predicts. In 200 patients screened by post-induction echocardiography, 52 (26%) had a patent foramen ovale and 54% of those had a venous air embolism, but only one had significant clinical manifestations and none had neurological sequelae. A systematic review of four observational studies and 977 patients found 82 with a patent foramen ovale, air embolism in 33 of them (40.2%), and no signs of paradoxical embolism in any.
Two qualifications keep that from being reassuring in the wrong way. The route does not have to be intracardiac — transpulmonary passage occurs in humans, and cerebral air emboli have been reported without any cardiac defect. And the gradient can reverse during the case: up to half of patients may reverse an existing left-to-right atrial pressure gradient after an hour in the sitting position, creating the conditions for shunting that were not present at induction.
Which produces one practical rule that applies regardless of position: paradoxical embolism to the cerebral circulation should be suspected in any patient with an unexpected neurological deficit after a procedure carrying a risk of air embolism.
Intravenous fluid loading is recommended to reduce the likelihood of right-to-left shunting. Patients with pre-existing pulmonary hypertension or right ventricular failure may not tolerate even small emboli.
Why the posterior fossa. Venous channels there are stented open by the surrounding structures, so they do not collapse when pressure falls. Air embolism is more common and more severe in posterior fossa craniotomy than in laminectomy. Air also enters through burr holes and through pin-site wounds from the head holder, especially with the head elevated.
How much matters. More than 50 mL has been retrieved from patients showing clinical change — falls in blood pressure, dysrhythmia, ECG changes. The estimated lethal volume in humans is over 300 mL. In a collection of 93 early case reports, 37 of 40 untreated patients died, which is the historical reason this is taken so seriously.
How often depends on how you looked. Reported incidence is 25 to 60% with the head above the heart, and it tracks the monitor: transoesophageal echocardiography 54%, precordial Doppler 43%, end-tidal CO₂ 9 to 28%. A 2017 review of 1792 sitting procedures counted clinically significant embolism — those requiring anaesthetic intervention — and found 2.7% in sitting suboccipital craniotomy.
Hold both. Air is common; trouble from air is not; and quoting one number for the other is how this conversation goes wrong.
Reducing it. No manoeuvre is 100% effective while a gradient exists. Incidence and severity fall with controlled positive-pressure ventilation, adequate hydration, the lowest head elevation that still gives exposure, careful dissection and liberal bone wax, avoidance of nitrous oxide in known intracardiac defects, and avoidance of drugs that increase venous capacitance such as nitroglycerin.
Hypovolaemia is the modifiable villain. It lowers central venous pressure and widens the very gradient the position creates, and it has been proposed as a predisposing factor to both venous and paradoxical embolism.
The trigger to act is a change on precordial Doppler, or a fall in end-tidal CO₂ of more than 2 mm Hg. The surgeon is informed immediately.
The intraoperative goals are three: stop further air entry, remove the air already there, and correct hypotension, hypoxaemia and hypercapnia. The sequence that serves them:
Jugular compression works but is not benign. It raises dural sinus pressure in both supine and sitting patients. The concerns are cerebral venous outflow obstruction with reduced flow, carotid compression or dislodgement of atherosclerotic plaque, venous engorgement with brain swelling, and carotid sinus bradycardia. Maintain cerebral perfusion pressure above 50 mm Hg in normotensive patients, higher in chronic hypertension — which you cannot do if the transducer is at the wrong height.
Three corrections to things people reach for:
And one thing that does work in extremis: external cardiac massage disrupts a large air lock.
Afterwards: supplemental oxygen, ECG, chest radiograph, serial arterial blood gases. Retinal air bubbles on fundoscopy are a described diagnostic sign of cerebral air embolism; CT may help but can be normal initially and changes over time. Hyperbaric oxygen for suspected arterial air embolism, with better neurological outcome when started within 6 hours of the insult — so if it is being considered, it is being considered now. A patient with neurologically significant cerebral air detected after extubation may benefit from reintubation and ventilation with 100% oxygen.
Air enters wherever the dura is breached and cerebrospinal fluid has left a space for it. In the sitting position the hydrostatics favour cerebrospinal fluid egress through the durotomy and air trapping in the dead space, so the incidence is extraordinary: 100% of sitting intracranial procedures, 72% park bench, 57% prone.
Risk factors are the things you were doing for other reasons — long duration, older age, intraoperative osmotherapy, hyperventilation, cerebrospinal fluid drainage, and nitrous oxide.
Almost all of it is asymptomatic and resolves. Tension pneumocephalus is the rare exception: more than about 60 mL trapped, producing mass effect and possible herniation, with focal deficits, headache, altered mental status, loss of consciousness or seizures. It can be suspected intraoperatively from a fall in somatosensory evoked potentials and confirmed postoperatively on CT. Mild cases are treated supportively with 100% oxygen; severe cases need urgent burr hole evacuation or an external ventricular drain.
The debate about this position is usually conducted as though sitting were obviously more dangerous. The observational comparison is more interesting than that.
In 333 sitting and 246 horizontal posterior fossa craniectomies:
| Sitting | Horizontal | |
|---|---|---|
| Hypotension across the whole anaesthetic | 36% | 38% |
| Hypotension in patients with cardiac disease | 56% | 55% |
| Transfusion of more than 2 units | 3% | 13% |
| Average blood replacement | 359 mL | 507 mL |
| Cranial nerve function deteriorated | 22% | 34% |
The sitting position was not the hypotension villain, bled less, and had better cranial nerve outcomes. What it does carry, on the source’s own comparative table, is the highest risk of venous air embolism of any position and a higher risk of paradoxical embolism, cerebral and cervical cord ischaemia, sciatic and peroneal palsy, and a “compartment syndrome” the other positions do not produce.
That is what a trade looks like when it is stated honestly, and it is why the source says the appropriateness of the position remains a matter of debate rather than pretending otherwise.
Examine the cranial nerves — a lost gag reflex or laryngeal dysfunction raises aspiration risk and may make a rapid sequence induction necessary. Look for brainstem compression from mass effect or obstructive hydrocephalus, because respiratory dysregulation or altered consciousness predicts delayed extubation and postoperative ventilation. Ask about dysautonomia, which argues for a preinduction arterial line. Assess the cardiopulmonary system and the cerebrovascular history — including previous carotid endarterectomy and previous shunting, the latter raising the risk of subdural pneumocephalus. And assess the veins, because someone will need to place a right atrial catheter: obesity, poor vasculature from disease or chronic cannulation, and short thick necks all need to be identified early enough that there is time to do it properly.
The theme of that paragraph is the theme of the day. Almost every hazard of this position is decided before the first incision, by someone standing at the head of the bed with a spirit level and a list.
No randomized trial addresses this.
The hydrostatics are not a claim, they are arithmetic. A column of blood 1.25 cm high exerts about 1 mm Hg. Elevating the head above the heart therefore reduces arterial pressure at the brain and reduces dural sinus pressure — by up to 10 mm Hg at 90 degrees — which is simultaneously the reason the field is dry and the reason air is drawn in. Nothing about that is contested and nothing about it depends on a study.
The complication rates come from observational series, and they disagree with each other for a reason. Venous air embolism is reported in 25 to 60% of head-up cases, but the figure tracks the sensitivity of the monitor used — 54% with transoesophageal echocardiography, 43% with precordial Doppler, 9 to 28% with end-tidal CO₂. A 2017 review of 1792 sitting procedures found clinically significant embolism — defined as requiring anaesthetic intervention — in 2.7% of sitting suboccipital craniotomies. These are different quantities, not conflicting estimates.
Whether to use the position at all is openly unsettled. The source states plainly that its appropriateness remains a matter of debate among both neurosurgeons and neuroanaesthetists, because alternative positions exist and air embolism is more common and more severe in the sitting position. That debate has never been resolved by a randomized trial and probably will not be.
And the observational comparison is less one-sided than the debate suggests. In a series of 333 sitting and 246 horizontal posterior fossa craniectomies, hypotension across the whole anaesthetic occurred in 36% and 38% respectively — essentially identical, and identical again when analysed by presence of cardiac disease. The sitting group needed fewer transfusions of more than two units (3% versus 13%), lost less blood on average (359 versus 507 mL), and had cranial nerve function deteriorate less often (22% versus 34%). The position's costs are real, but they are not the ones usually named.
The one place a randomized result exists, it is negative. Fifty percent nitrous oxide did not increase the incidence or severity of venous air embolism in sitting neurosurgical patients, provided it was discontinued when air was detected on Doppler.
The patient is sitting. The arterial line was placed before induction, on the trolley, transduced at the level of the heart — as it is for every other case you do. The monitor reads a mean arterial pressure of 65, and 65 is a number you have accepted a thousand times.
The head is 25 cm above the transducer.
Local arterial pressure falls by about 1 mm Hg for every 1.25 cm of elevation above the heart. Twenty-five centimetres is twenty millimetres of mercury. The mean pressure at the circle of Willis is 45, and the cerebral perfusion pressure is lower still. Nothing on the monitor will tell you this. There is no waveform change, no alarm, no trend. The error is a constant, it is present from the moment the table is raised until the moment it is lowered, and it is proportional to how well the surgeon has positioned the head.
The whole of the fix is one sentence in the source: the arterial pressure transducer is zeroed at the skull base during positioning and throughout the procedure. Re-zero with every adjustment of the table.
Three things make this worth a whole pitfall rather than a footnote.
It compounds with everything else the position does. Sitting up increases pulmonary and systemic vascular resistance and decreases cardiac output, venous return and cerebral perfusion pressure. So the pressure you are reading is already lower than it would be supine, and then it is overstated by the height of the head on top of that.
The patients who tolerate it least are the ones you will have. The source names them precisely: altered limits of cerebral autoregulation, impaired cerebral perfusion, or abnormal baroreceptor function from chronic hypertension, cardiovascular disease, cerebrovascular insufficiency — or prior carotid endarterectomy. That last one is yesterday's patient. In these patients, hypotension in the head-up position is described as especially detrimental.
And the arithmetic reappears in the emergency. During treatment of a venous air embolism you may be asked to compress the jugular veins. That works by raising dural sinus pressure, but it also risks obstructing cerebral venous outflow. The instruction is to maintain a cerebral perfusion pressure above 50 mm Hg in normotensive patients and higher in chronic hypertension — a threshold you cannot honour if your zero is at the wrong height. A displayed 65 becomes a real 45, and the number you are defending against is 50.
So the error is not merely a measurement problem. It converts a stated safety threshold into one you are already below while believing you are above it.
One more thing worth naming, because it is the same mistake in a different coordinate. Neck flexion and extension move the endotracheal tube caudad or cephalad by as much as 2 cm. Palpating the cuff above the sternal notch after positioning is the equivalent habit for the airway: something that was correct before the patient was moved and is not necessarily correct afterwards.
Full text · Preston goes through LibKey to whatever Preston holds — the PDF where there is one — and works on a phone. The doi link always reaches the publisher's record, whether or not we hold a subscription. If you are reading on a computer, LibKey Nomad adds the same one-click PDF button to PubMed and publisher pages everywhere else; it is a desktop browser extension, so it does nothing on a phone or tablet.
A patient is in the sitting position for a suboccipital craniotomy. The arterial transducer was zeroed at the mid-axillary line before induction and not moved. The head is approximately 25 cm above the transducer and the monitor displays a mean arterial pressure of 65 mm Hg. What is the approximate mean pressure at the level of the brain?
A colleague says venous air embolism occurs in 50% of sitting posterior fossa cases and another says it occurs in under 3%. Who is right?
A large venous air embolism occurs during a sitting posterior fossa craniotomy. Which intervention should be avoided?
That is the whole orientation rule. Fat is bright on T1 and intermediate on T2, cortical bone gives no signal, and flowing blood produces a signal void. The practical summary is that anatomy is better seen on T1 and pathology on T2 — and post-contrast sequences are T1 weighted for exactly the reason given.
Ferromagnetic aneurysm clips are a contraindication; non-ferromagnetic ones are not. The risk from a ferromagnetic object is magnetically induced movement, with dislodgement and injury, and heating — image artefact is the secondary concern. The information you need is in the old operation note, not on any current imaging.