Spinal Cord Decompression Sickness: The Myelopathy That Can Walk In

Stylised sagittal MRI-style illustration of the thoracic spine: stacked vertebral bodies in greyscale with a narrow longitudinal amber band running along the posterior spinal cord, teal oxygen molecules diffusing inward from surrounding vessels, on a deep navy background

Twenty-four minutes after he reached the surface, the pain started between his shoulder blades and ran down both legs. Then came the numbness, and the sensation that his bladder was full when it was not.

The diver was 43 and experienced enough to hold Trimix and closed-circuit rebreather certifications. Thirty-five minutes into a wreck dive at roughly 90 feet, out of sight of his buddy, his rebreather failed and he aspirated seawater into the loop — a “caustic cocktail.” He inflated his buoyancy compensator and rode it up in about a minute, bypassing a decompression stop that should have taken ten to fifteen. He exhaled continuously the whole way and never lost consciousness.

What makes his case worth reading, though, is the first neurological examination. Mental status intact. Cranial nerves intact. Full strength in every muscle group. And then: sensation blunted to light touch, vibration and proprioception below the T10–L1 dermatome, reflexes 3+ at the knees and ankles with bilateral ankle clonus, extensor plantar responses, an unsteady gait that was plainly sensory ataxia rather than weakness — and a bladder that would not empty, requiring a catheter.

That constellation is a spinal cord emergency. It is also the kind of presentation that gets briefly mistaken for a back strain, a panic reaction, or a diver who came up too fast and should sleep it off.

Why a neurologist needs to know this

Pisano and colleagues at the University of Pennsylvania — neurology, with co-authors from the division of undersea and hyperbaric medicine — published the case in The Neurohospitalist on 16 October as a teaching exercise for the consultant most likely to be called first and least likely to have seen a case before.

The epidemiology explains the gap. Decompression sickness in recreational diving is uncommon, on the order of two to four cases per 10,000 dives; neurologically significant involvement is rarer still, under one per 10,000. Rare enough that no individual clinician accumulates pattern recognition. Common enough that a busy hyperbaric unit sees it every few weeks.

The thoracic cord is the classic target, and the reason is physical before it is anatomical. Nitrogen dissolves into tissue in proportion to ambient pressure and is roughly five times more soluble in fat than in water. White matter is lipid-rich and comparatively poorly perfused, so it holds inert gas and releases it slowly. Add the thoracic cord’s watershed arterial supply and the valveless epidural venous plexus, where bubbles can collect, and you get the familiar picture: dorsal column deficits, a sensory level, lower-limb involvement, bladder dysfunction. Damage itself runs through three channels at once — direct compression, vascular obstruction, and inflammatory and coagulation activation at the bubble–endothelium interface.

What the scan shows, and what it cannot

Imaging was done on the third hospital day, seventy-three hours after symptom onset and after the first chamber run. It showed T2/STIR hyperintensity in the dorsal thoracic cord extending from T3 to T10-11, without contrast enhancement, without T1 hypointensity and without cord expansion.

The authors are unusually careful about what that means. The distribution supports spinal decompression sickness in the clinical context but is not specific, and a normal MRI early on does not exclude it. Compressive myelopathy, spinal cord infarction and transverse myelitis all sit in the differential, and what separates them is the exposure history and the tempo, not the magnet.

They are equally careful about prognosis. Symptom onset within thirty minutes, involvement of more than six vertebral levels and bladder dysfunction have each been associated with incomplete recovery in observational series, and the Boussuges severity score is used to quantify that risk. This patient was on the wrong side of two of those markers — roughly eight levels, retention, onset at twenty-four minutes — and still recovered well. Prognostic features describe groups, not outcomes.

Four hours and forty-five minutes, and the hours around them

He was treated on a U.S. Navy Treatment Table 6: 100% oxygen at 2.8 atmospheres absolute with intermittent air breaks and staged decompression, four hours and forty-five minutes in total, started the day he arrived. Three further sessions followed — a second Table 6, a Table 5, and an elective run — for persistent numbness and urinary retention. After the fourth, the pain was gone, the numbness had retreated to the distal legs, the clonus had resolved and he was walking unaided. He passed a trial of void on day five with a mildly elevated residual around 200 mL, and went home on day six.

The chamber was not the whole protocol, and the review is scrupulous about the difference between physiology and proof. Consensus guidance adds near-100% oxygen at the surface, supine or horizontal positioning with Trendelenburg explicitly avoided, thermal protection, oral hydration when it is safe, and intravenous isotonic crystalloid. The rationale for volume is plausible — immersion drives a natriuresis and diuresis, dehydration worsened severe decompression sickness in a swine model, and pre-dive hydration reduced post-dive venous bubbles in a small crossover study — but no DCS-specific fluid target exists. An eight-subject physiology study found greater nitrogen elimination when subjects were warm; that is a reason to keep a hypothermic diver warm, not a demonstration that warming improves neurological recovery.

The thin evidence under a universal standard

Here is the part the paper says out loud, and it deserves repeating. Recompression is the universally accepted standard treatment for decompression illness, and per the Cochrane review by Bennett and colleagues there is no randomised controlled trial evidence for its use. Two trials of adjunctive therapy met inclusion criteria, enrolling 268 patients between them. Adding the NSAID tenoxicam reduced the number of recompressions needed to reach clinical endpoints from three to two (P = 0.01) without improving recovery at six weeks (RR 1.04, 95% CI 0.90–1.20). A helium–oxygen table lowered the odds of needing multiple recompressions (RR 0.56, 95% CI 0.31–1.00) without improving the odds of recovery. The reviewers’ own summary: benefits “may be largely economic.”

Corticosteroids are not routinely recommended. Cerebrospinal fluid drainage is not standard of care — it is extrapolated from non-diving spinal ischaemia and aortic surgery. NSAIDs are reasonable and unproven for neurological outcome. The chamber stands on mechanism, on the behaviour of bubbles under pressure, and on eight decades of practice.

Where the recovery story usually stops

Six weeks before the Pennsylvania case appeared, a French group published what amounts to the other half of the picture. Dubois and colleagues, spanning neuro-urology at Pitié-Salpêtrière, rehabilitation medicine at Le Vésinet and hyperbaric medicine at the Sainte-Anne military teaching hospital in Toulon, searched international and French decompression-illness and spinal cord injury guidelines plus PubMed back to 1996. Their first finding was an absence: no specific long-term follow-up protocols for spinal cord decompression illness exist. They adapted spinal cord injury domains instead — lower urinary tract and gastrointestinal function, thromboembolic risk, autonomic dysreflexia, pulmonary function, pain, spasticity — and recommended systematic neurological and urological screening even in patients who appear to have recovered, with risk-adapted follow-up across the first two years. Sequelae, they write, are frequent and may be underdiagnosed.

The published cases bear that out with uncomfortable precision. A U.S. Navy diver treated twenty hours after surfacing, with a 2.2 cm T4 lesion visible on MRI four days in, achieved almost complete return of motor and sensory function over two years of follow-up — and continued to have faecal incontinence and an abnormal post-void residual. In the Pennsylvania case, the residual at discharge was already mildly elevated. Motor recovery and autonomic recovery are not the same recovery, and the visible one is the one everybody celebrates.

Rehabilitation may need to start earlier than it usually does. A case report from Cyprus in Acta Neurologica Taiwanica described a 65-year-old with an incomplete T4 injury after a 35-minute dive who began high-intensity physiotherapy on day one of admission alongside hyperbaric treatment, six sessions a week for three months and then four a week for three more, with clinically important gains in strength, sensation, balance and ambulation. One patient, no control group — an argument for a trial, not a protocol.

The regulatory line this sits on

None of this is a wellness claim, and the distinction is worth stating plainly in a field where it often is not. Decompression sickness is one of the conditions hyperbaric oxygen therapy was cleared to treat: HBOT devices are Class II medical devices cleared through the 510(k) pathway under Product Code CBF, CMS’s national coverage determination for hyperbaric oxygen therapy lists decompression sickness among covered conditions, and the Undersea and Hyperbaric Medical Society treats it as a core indication.

The same hardware, though, is marketed for sleep, focus, recovery and longevity, uses the FDA has not cleared — and the agency’s August 2025 letter to health care providers remains the standing warning for every one of them. It reports fires with HBOT devices that caused serious injuries and deaths, states that the root cause is not currently known, and asks facilities to follow manufacturer instructions on fire prevention, staff training, monitoring for the entire treatment, maintenance intervals, prohibited electrical and static devices, and hyperbaric-compatible clothing such as cotton.

Read that list against the clinical reality of decompression sickness and it stops looking like paperwork. These are emergency treatments, frequently begun at night, in patients arriving by helicopter after a delay nobody planned. The pre-treatment safety check is exactly the step that a rushed handover invites a team to skip.

For divers, the practical lesson is timing: symptoms that begin within half an hour of surfacing, back or chest pain that radiates, a limb that feels wrong, or a bladder that will not empty are reasons to call for hyperbaric advice, not to lie down and reassess. For clinicians, it is to examine proprioception and ask about voiding in any diver with a bad back. And for the field, it is the quieter problem the French authors identified — that the appointment nobody schedules, two years after the chamber, may be the one that determines what recovery actually looks like.

Sources

← Back to blog