Decompression Sickness Myelopathy: Case Report and Acute Management Review for the Neurologist

Abstract

Decompression sickness (DCS) is an uncommon but important cause of acute myelopathy that neurologists should recognize in the emergency setting. We report a 43-year-old experienced technical diver who developed acute sensory loss with a thoracic sensory level, gait dysfunction, upper motor neuron signs, and urinary retention after rapid ascent from approximately 90 feet below sea level. Thoracic spine MRI demonstrated longitudinal, posterior-predominant T2/STIR hyperintensity from T3 to T10–11 without pathologic enhancement, supporting spinal cord DCS. He received prompt hyperbaric oxygen therapy using U.S. Navy Treatment Table 6, along with 100% oxygen, horizontal positioning, and intravenous fluid resuscitation, followed by additional hyperbaric treatments for persistent sensory and autonomic symptoms. He improved rapidly, regained independent ambulation, and later returned to high-level physical activity, although intermittent sensory symptoms persisted during follow-up. This case illustrates the clinical presentation of spinal cord DCS and emphasizes the importance of early hyperbaric oxygen therapy alongside consensus-supported measures for oxygenation, intravascular volume repletion, thermal protection, and safe positioning. The independent effects of these supportive measures on neurologic recovery remain uncertain. Evidence for additional pharmacologic adjuncts, including NSAIDs and corticosteroids, is limited, and cerebrospinal fluid drainage remains unproven in DCS.

Introduction

Neurologic decompression sickness (DCS), or “the bends,” is a rare but potentially devastating condition resulting from inadequate decompression following exposure to increased ambient pressure. DCS in recreational diving is uncommon, with an estimated incidence of approximately 2-4 cases per 10 000 dives. ¹ A substantial proportion of clinically significant cases involve the nervous system. ² Registry and epidemiologic data suggest that neurologic DCS is rare among recreational divers, occurring at a rate on the order of less than 1 case per 10 000 dives.^1,3 Neurologists may be consulted for patients presenting with acute nontraumatic myelopathy after diving, making early recognition and treatment crucial.

Neurologic DCS often involves the posterior white matter of the thoracic spinal cord.^4,5 Prompt recognition and initiation of hyperbaric oxygen therapy (HBOT) remain central to management, alongside consensus-supported measures for oxygenation, intravascular volume repletion, and safe positioning.^2,6 This report presents a case of severe spinal cord DCS in an experienced diver to review the pathophysiology, clinical spectrum, diagnostic evaluation, and acute management principles relevant to the consulting neurologist.

Case Presentation

A 43-year-old male with no significant past medical history was brought by helicopter to the emergency department with acute onset chest-down numbness, severe radiating pain, bilateral lower extremity weakness, and urinary retention after a rapid ascent from depth during a scuba dive. The patient is a dive master with decades of experience and advanced certifications including Trimix and closed-circuit rebreather (CCR). The incident occurred approximately 35 minutes into a technical wreck dive on which he was operating a CCR system with compressed air. While at a depth of approximately 90 feet and out of visual line-of-sight of his dive partner, he suffered multiple critical equipment failures and ingested caustic byproducts due to ingress of seawater into the CCR system ('caustic cocktail'). ⁷ He was forced to make an emergency ascent via buoyancy control device inflation to surface in approximately one minute, bypassing a required 10-15 minute decompression stop. He was able to perform continuous exhalation and did not lose consciousness at any point. Twenty-four minutes after surfacing, he felt sudden intense pain from his mid-thoracic region down to his legs, quickly followed by numbness. He received 100% oxygen on the dive boat, his mouth was washed out with cola without any significant residual mouth or throat pain, and he was transported by Coast Guard helicopter for emergent hyperbaric oxygen therapy (HBOT).

Initial neurological consultation, prior to HBOT, revealed intact mental status and cranial nerve function. Motor examination showed full strength throughout. Sensory examination showed a deficit to light touch, vibration, and proprioception below the T10-L1 dermatome. Reflexes were 3+ at the patella and ankles with bilateral ankle clonus, and plantar responses were extensor. Gait was slow and unsteady, and he was unable to tandem walk. The gait impairment was likely related to sensory ataxia from impaired proprioception despite preserved strength on confrontational testing. He was unable to void despite a sensation of bladder fullness, requiring Foley catheter placement. Initial laboratory studies were unremarkable.

He was treated according to the U.S. Navy Treatment Table 6 (USN TT6) protocol, a standardized hyperbaric oxygen regimen developed by the U.S. Navy for the treatment of serious decompression illness and arterial gas embolism and widely adopted as the clinical standard of care. ⁸ He underwent the first session of HBOT on the day of presentation with inhalation of 100% O2 at a pressure of 2.8 atmospheres absolute with intermittent air breaks followed by staged decompression for a total treatment time of 4 hours 45 minutes. An MRI of the thoracic and lumbar spine performed on the third hospital day (73 hours after symptom onset), after the initial HBOT session, revealed T2/STIR hyperintensity in the dorsal thoracic cord from T3 to T10-11 without contrast enhancement, consistent with spinal cord DCS (Figure 1).

Figure 1.

Thoracic MRI reveals intramedullary T2 signal abnormalities. (A) Sagittal MRI of the thoracic spine obtained on hospital day 3 (73 hours after symptom onset), after the initial HBOT session, demonstrates T1-weighted (left), T2-weighted (middle), and STIR (short tau inversion recovery) sequences (right). Blue lines labeled B–E on the sagittal image indicate the slice levels for the corresponding axial panels; blue arrowheads in the axial panels identify intramedullary signal abnormalities. (B–E) Axial T2-weighted images show bilateral posterior-predominant, nonspecific intramedullary signal abnormality within the longitudinal lesion extending from approximately T3 through T10-11, without corresponding T1 hypointensity, cord expansion, or abnormal enhancement (not shown). This distribution is supportive of spinal DCS in the clinical context but is not specific

He completed 3 additional daily HBOT sessions (one USN TT6, one USN TT5, and one elective session) to address persistent leg numbness and urinary retention. After his fourth HBOT session, the patient showed substantial overall improvement. Pain had resolved, numbness was confined to the distal legs, clonus had resolved, and he was ambulating independently. He passed a trial-of-void on hospital day 5, though did still have mildly elevated post-void residuals around 200 mL. He was discharged to home on hospital day 6.

The presentation was most consistent with spinal DCS (Type II DCS) within the decompression illness (DCI) spectrum. The Boussuges score is a weighted clinical severity score that uses the most severe manifestations before recompression and incorporates repetitive diving, clinical course, objective sensory deficits, motor impairment, and urinary dysfunction. ⁹ The neurology team documented a score of 16 based on his most severe pretreatment manifestations that day, citing stable symptoms before recompression, sensory deficit, paresis, and bladder dysfunction, suggesting an increased risk of incomplete recovery.^10,11 At follow-up, the patient confirmed complete and rapid resolution of his urinary retention and leg weakness days following discharge, along with gradual improvement of his lower extremity paresthesias over several months. Between 6 and 8 months after discharge, he resumed running and completed a triathlon, but continued to experience intermittent sensory distortion in the feet and thighs. MRI of the thoracic spine at follow-up showed marked interval resolution of the previously noted longitudinal intramedullary T2/STIR hyperintensity (Figure 2), consistent with radiographic improvement paralleling his clinical recovery. At this interval imaging, a cervical MRI was also obtained and showed subtle T2 signal abnormality extending up to the mid-cervical level (Figure 3). Because dedicated cervical MRI was not obtained initially, these abnormalities cannot be confirmed as new or definitively attributed to this episode of DCI.

Figure 2.

Interval improvement of thoracic spinal cord signal abnormalities on follow-up MRI. Sagittal MRI of the thoracic spine (top row), obtained approximately 7 months after the initial study, demonstrates T1-weighted (left), T2-weighted (middle), and STIR (short tau inversion recovery) sequences (right). Compared with the prior examination, there is marked reduction of the previously noted longitudinal intramedullary T2/STIR hyperintensity. Vertebral levels are labeled for anatomical reference. Blue lines labeled B–E on the sagittal image indicate the slice levels for the corresponding axial panels. (B–E) Axial T2-weighted images demonstrate marked improvement of the posterior-predominant intramedullary signal abnormalities. These findings support interval radiographic improvement following spinal DCS

Figure 3.

Cervicothoracic spinal cord signal abnormalities on follow-up MRI. (A–C) Sagittal MRI of the cervical spine obtained at follow-up demonstrates T1-weighted (A), T2-weighted (B), and STIR (short tau inversion recovery) (C) sequences. There is longitudinal intramedullary T2/STIR hyperintensity extending from approximately C3 through the thoracic cord, without associated T1 hypointensity or cord expansion. (D) Axial T2-weighted image at the C6-7 disc level, corresponding to the blue slice-localization line in (C), demonstrates bilateral intramedullary signal abnormality without cord enlargement. These findings may reflect sequelae of spinal DCS. Without baseline dedicated cervical imaging, however, they cannot be confirmed as new or definitively attributed to this episode

Discussion

Decompression sickness (DCS) arises from the formation of inert gas bubbles within tissues and blood during or after a reduction in ambient pressure. Type II DCS includes neurologic and cardiopulmonary manifestations (Type I DCS symptoms are limited to skin, musculoskeletal, and lymphatic involvement). ¹² Arterial gas embolism (AGE) typically presents abruptly during ascent or shortly after surfacing, often with cerebral deficits.^2,13 Although the rapid ascent raised concern for AGE, symptom onset approximately 24 minutes after surfacing and a clinically isolated spinal cord syndrome favored spinal DCS. Preserved mental status and cranial nerve function, together with longitudinal posterior-predominant cord signal abnormalities, supported this interpretation but did not establish the underlying mechanism. AGE or mixed decompression illness cannot be excluded with certainty; this uncertainty does not diminish the urgency of recompression treatment.^2,13 MRI findings are supportive rather than specific, and normal MRI does not exclude spinal DCS (Table 1).^4,5

Table 1.

Evaluation of Suspected Neurologic Decompression Sickness. Key Clinical Features, Imaging Findings, Laboratory Considerations, and Prognostic Tools Relevant to the Early Identification and Risk Stratification of Spinal Cord Decompression Sickness in Patients Presenting With Acute Myelopathy After Diving

Domain	Component	Key elements	Clinical rationale
Recognition	Risk factors	Recent dive with rapid ascent, omitted decompression stops, equipment malfunction, technical/CCR dive	Establishes pre-test probability; neurologic DCS often presents immediately or within hours
	Symptom timing	Immediate or early onset (minutes–hours) after surfacing	Early onset associated with greater severity and worse prognosis
	Symptom pattern	Sensory level, paraparesis, gait instability, bladder/bowel dysfunction	Thoracic spinal cord and dorsal column involvement are characteristic
Neurologic exam	Motor	Weakness, UMN signs, hyperreflexia, clonus	Suggests spinal cord involvement (Type II DCS)
	Sensory	Dorsal column deficits (vibration, proprioception), sensory level	Typical of thoracic white matter injury
	Autonomic	Urinary retention, bowel dysfunction	Common in severe spinal DCS; prognostic significance
Imaging	MRI spine	T2/STIR hyperintensity, often dorsal thoracic cord; may lack enhancement	May support diagnosis but is not specific; multilevel involvement has been associated with outcome in small series.4,5
	Timing considerations	MRI may be normal early	Normal imaging does not exclude DCS
Laboratory workup	Exclusion labs	B12, copper, syphilis, inflammatory markers	Exclude alternative causes of acute myelopathy
Differential diagnosis	Key mimics	Compressive myelopathy, spinal cord infarction, transverse myelitis	Distinguished by acuity, exposure history, examination, and appropriate diagnostic testing
Prognostication	Clinical predictors	Early onset (<30 min), multilevel (>6) cord involvement, bladder dysfunction	Associated with incomplete recovery in observational series.4,11
	Severity scoring	Boussuges score	Quantifies risk of persistent neurologic deficit
Disposition planning	Specialty involvement	Hyperbaric medicine, neurology, urology, rehabilitation	Multidisciplinary care optimizes outcomes

The underlying principle of DCS is described by Henry’s law: the solubility of a gas in a liquid is directly proportional to its partial pressure.^2,12 During a dive, increased ambient pressure causes significant amounts of nitrogen, the primary constituent of compressed atmospheric air, to dissolve into the blood. While at depth, the dissolved nitrogen diffuses into tissues and forms reservoirs. Lipid-rich tissues, such as white matter and adipose tissue, are at particular risk given nitrogen is approximately 5 times more soluble in fat than in water, increasing its ability to diffuse across plasma membranes.^2,12 With a controlled ascent and decompression stops, nitrogen is gradually redistributed from tissue reservoirs into the blood and, as ambient pressure drops, is outgassed and exhaled harmlessly via the lungs. A rapid ascent, however, can cause dissolved nitrogen to precipitate directly into bubbles, leading to mechanical and vascular injury. ¹²

Prehospital management should begin immediately at the surface. Consensus guidance recommends administration of near-100% oxygen.^6,14 It also recommends placement in a horizontal or supine position, oral hydration when safe, thermal protection, and IV placement for isotonic crystalloid infusion if trained personnel are available. ⁶ Prompt evaluation to determine likelihood and severity of DCS (often via telemedicine consultation with Coast Guard, diving medicine specialist, or other emergency medical services) is key, with triage of all but mild cases of DCS for rapid evacuation to a facility with hyperbaric capability. ⁶ Mild DCS is defined by non-progressive limb pain, fatigue, cutaneous sensory changes, rash, and subcutaneous swelling, together with the exclusion of significant neurologic dysfunction. ⁶ In-water recompression (IWR), essentially returning the diver to deep water for recompression and subsequent controlled decompression stops, has been described but is not standard of care; it is considered only when a chamber is not reasonably accessible and only by trained teams with appropriate equipment. Most authorities advise surface oxygen and evacuation rather than ad hoc redescents for recreational divers.^6,15

In DCS, nitrogen bubbles cause damage via several mechanisms: (1) direct compression of neural structures, (2) vascular obstruction, and (3) activation of inflammatory and coagulation pathways at the bubble-endothelial interface. ² Proposed explanations for thoracic cord vulnerability include its watershed arterial supply, possible bubble trapping within the valveless epidural venous plexus, and the high lipid content and relatively lower perfusion of its white matter, which may favor inert gas retention.^2,4,5 This localization helps explain the frequent dorsal column and lower extremity involvement observed in spinal DCS.^4,5 A small retrospective MRI series reported both gray and white matter involvement, suggesting mixed arterial and venous injury patterns; symptom onset within 30 minutes and involvement of more than 6 spinal levels were associated with poorer outcomes. ⁴

Management of spinal cord DCS (Table 2) parallels principles applied to acute ischemic injury of the spinal cord. The goals are to eliminate the inciting bubbles and support perfusion of compromised tissue while avoiding secondary physiologic injury. The perfusion analogy rests on extrapolated physiology rather than direct evidence that spinal cord stroke protocols improve outcomes in DCS. Hyperbaric oxygen therapy (HBOT) remains the definitive treatment, working by driving gas bubbles back into solution, enhancing their diffusion gradient into the circulation for clearance in the lungs, and hyperoxygenating ischemic tissues. ² Additional HBOT sessions are individualized to persistent or recurrent symptoms and clinical response; evidence for specific repeat-treatment schedules is primarily observational. ¹⁹

Table 2.

Acute Management of Neurologic Decompression Sickness. Prehospital and In-Hospital Interventions for Spinal Cord Decompression Sickness, Including Definitive HBOT, Consensus-Supported Measures for Oxygenation and Perfusion Support, and Adjuncts With Limited or Unproven Clinical Benefit. Physiologic Rationale and Extrapolation Should Not be Interpreted as Established Evidence of Improved Neurologic Recovery in Spinal DCS

Phase	Intervention	Specific actions	Rationale/Evidence
Immediate (surface/ED)	Oxygen therapy	100% oxygen via non-rebreather or demand valve	Accelerates nitrogen elimination and improves tissue oxygenation
	Positioning	Supine, horizontal positioning; avoid Trendelenburg	Consensus-supported; may limit adverse bubble redistribution (physiologic rationale); avoid head-down positioning. 6
	Thermal management	Maintain thermal comfort; avoid temperature extremes. Rewarm if hypothermic, using active measures as indicated. 16	Warming may increase nitrogen elimination experimentally; neurologic benefit in DCS is unproven.6,17
	IV access	Early IV placement	Enables fluid replacement when indicated
Perfusion support and monitoring	Fluid resuscitation	Isotonic crystalloid replacement individualized to volume status and hemodynamics	Corrects volume depletion, including possible immersion-related losses; no validated DCS-specific volume or perfusion target.6,18
	Monitoring	Urine output, hemodynamics	Guides resuscitation; avoids volume overload
	Bladder management	Foley catheter if urinary retention	Prevents overdistension; common in thoracic DCS
Definitive therapy	Hyperbaric oxygen therapy	U.S. Navy treatment Table 6 (or equivalent)	Compresses bubbles, enhances nitrogen washout, hyperoxygenates ischemic tissue
	Repeat HBOT	Additional sessions individualized to persistent or recurrent symptoms and clinical response	Primarily observational evidence; optimal schedule and number remain uncertain 19
Adjunctive Therapies	NSAIDs	Consider (eg, tenoxicam)	May reduce HBOT sessions; improved neurologic outcome not demonstrated 20
	Corticosteroids	Not routinely recommended	Clinical benefit remains unproven2,8
	CSF drainage	Not standard of care	Extrapolated from non-diving spinal ischemia and aortic surgery; benefit in DCS unproven.21,22
Post-acute care	Rehabilitation	Early PT/OT	Supports neurologic recovery
	Follow-up	Neurology, urology	Addresses residual motor and autonomic dysfunction
	Diving counseling	Restrict diving until cleared	Prevents recurrence

Supportive measures extend beyond HBOT. Immersion shifts blood centrally and promotes renal sodium and water excretion (immersion-related natriuresis and diuresis), which can contribute to volume depletion. ¹⁸ Dehydration increased the risk of severe DCS in a swine model, ²³ and a small human crossover study found reduced post-dive venous bubble formation with pre-dive hydration. ²⁴ These observations provide a physiologic rationale for maintaining circulating volume, but do not establish that aggressive fluid resuscitation improves neurologic recovery after spinal DCS. Consensus guidance recommends early isotonic crystalloid replacement in more seriously affected divers, individualized to clinical volume status, hemodynamics, and urine output while avoiding fluid overload. ⁶ Specific fluid volumes or spinal cord perfusion targets have not been established for DCS.

NSAIDs may also play a role, but definitive data is still lacking. In one trial, tenoxicam reduced the number of HBOT sessions needed to achieve clinical endpoints without added complications, but did not significantly improve residual symptoms at follow-up. ²⁰ Thermal management should maintain comfort and avoid hypothermia or hyperthermia. ⁶ An eight-subject physiology study demonstrated greater nitrogen elimination under warm conditions and reduced elimination under cool conditions, providing a physiologic rationale for avoiding hypothermia. ¹⁷ This does not establish that deliberate warming improves neurologic recovery in spinal DCS. Rewarming a hypothermic patient, including active measures when clinically indicated, is distinct from warming a normothermic patient to enhance nitrogen elimination. ¹⁶

Patients should be maintained in a horizontal or supine position during transport and early management when practicable; Trendelenburg positioning is not advised. ⁶ Horizontal positioning may limit potentially adverse bubble redistribution, but this is a physiologic rationale rather than proof of improved spinal cord perfusion or neurologic outcome. ⁶

In severe spinal DCS, cord edema could theoretically create a compartment-like effect within the dural sheath and further compromise perfusion. This mechanism is extrapolated from traumatic spinal cord injury rather than demonstrated in DCS. ²⁵ Spinal cord perfusion pressure (SCPP) is commonly approximated as mean arterial pressure minus intrathecal pressure, providing a theoretical rationale for pressure augmentation or cerebrospinal fluid (CSF) drainage in refractory cases. ²¹ Perioperative CSF drainage reduced postoperative neurologic deficits in a randomized trial of extensive thoracoabdominal aortic repair, ²² and drainage with blood pressure elevation has been described in other non-diving spinal cord infarctions. ²¹ These findings cannot be assumed to apply to DCS: clinical benefit, safety, patient selection, and perfusion targets remain unestablished. Routine blood pressure augmentation or CSF drainage is therefore not supported for DCS. Corticosteroids are not routinely recommended because clinical benefit remains unproven.^2,8

Prognosis in spinal DCS is variable. The Boussuges score correlates with outcome but does not determine an individual patient’s recovery.^9-11 Higher scores are associated with incomplete recovery, and approximately one-quarter to one-third of patients in retrospective spinal DCS cohorts had residual deficits at 1 month despite recompression.^10,11 In a multicenter retrospective cohort, a Boussuges severity score greater than 7 was independently associated with unfavorable recovery after spinal DCI. ¹¹ There remains a need for better prognostic tools to address this clinical problem.

Conclusion

Spinal cord decompression sickness is a neurologic emergency that requires rapid recognition and immediate initiation of therapy. HBOT remains the cornerstone of treatment, directly targeting both intravascular and tissue reservoirs of harmful gas bubbles for elimination and improving tissue oxygenation. High-flow oxygen, individualized isotonic fluid replacement, thermal protection, and appropriate positioning are consensus-supported adjuncts, although their independent effects on neurologic recovery remain uncertain. Evidence for additional pharmacologic adjuncts is limited, and cerebrospinal fluid drainage remains unproven in DCS. Further research is needed to establish the safety and efficacy of these strategies, although the rarity of this clinical entity presents challenges. Conceptualizing spinal DCS as a “spinal cord stroke” provides a useful physiologic framework for neurologists to guide acute management. This analogy is extrapolative: direct evidence supporting stroke-derived perfusion protocols in DCS is lacking, and it does not establish a role for routine blood pressure augmentation or CSF drainage. Ultimately, a multidisciplinary approach incorporating hyperbaric medicine, neurology, and rehabilitation is essential to maximize recovery and functional outcomes.

ORCID iDs

Thomas J. Pisano https://orcid.org/0000-0002-8432-113X

Brendan Zotter https://orcid.org/0000-0002-6233-663X

David S. Lambert https://orcid.org/0000-0003-0523-0641

Consent to Participate

The patient provided consent for their case to be used in this report. Documentation of this consent is stored within the institution’s electronic medical record.