This article is part of the HBOT Made Simple series, where we explain how hyperbaric oxygen therapy works — clearly, honestly, and without the jargon.
Brought to you by Brain Spa Hyperbaric
You've been reading about hyperbaric oxygen therapy. Maybe you've seen that professional athletes swear by it. Maybe someone in a Facebook group told you they bought a high-pressure chamber for their home and love it. Maybe you've found a deal on a 2.0 ATA chamber — hard-shell or even a soft-shell rated to clinical pressures — and the price seems surprisingly reasonable.
And you're thinking: why not skip the clinic and go straight to the "real thing"?
Here's the short answer: because at clinical pressures with medical-grade oxygen, things can go wrong that you cannot fix by yourself. And in a sealed chamber — whether it's steel, acrylic, or heavy-duty fabric — "by yourself" is exactly where you'll be.
This article walks through three specific risks — oxygen toxicity, fire, and the practical reality of being alone in a pressurized chamber when something goes wrong. These aren't theoretical scare stories. They're documented events with real numbers behind them.
First: what makes a high-pressure chamber different from a mild one
When we say "high-pressure" or "clinical-pressure" in this article, we mean any chamber — hard-shell or soft-shell — capable of reaching 2.0 ATA or higher, typically used with medical-grade oxygen from compressed gas cylinders or high-output concentrators. These are the chambers you find in hospitals and hyperbaric clinics. They're increasingly also being sold directly to consumers for home use.
A mild hyperbaric chamber (mHBOT) is a different category. It typically operates at 1.3–1.5 ATA. Some people use them with supplemental oxygen from a concentrator (which delivers about 93% oxygen, not 100%). Others use them with compressed air only — no supplemental oxygen at all. Both approaches are used.
The difference isn't just a number on a gauge. It's a fundamentally different risk profile — and by the end of this article, you'll understand exactly why.
Oxygen toxicity: what it is and why it matters
Your body needs oxygen. But too much oxygen, at too high a pressure, for too long, becomes harmful — particularly to your brain. This is called CNS (central nervous system) oxygen toxicity, and it's been known since the 1870s.
The way it works is straightforward. What matters isn't just how much oxygen you breathe — it's the partial pressure of oxygen (ppO₂) your tissues are exposed to. Partial pressure combines two factors: the fraction of oxygen in the gas you breathe, multiplied by the total pressure.
You're probably used to seeing ATA as a measure of chamber pressure — 1.3 ATA, 1.5 ATA, 2.0 ATA. But ATA is also the unit used to measure oxygen partial pressure (ppO₂). That can be confusing, because they look the same but mean different things. Chamber pressure tells you how much the air is compressed. Oxygen partial pressure tells you how much oxygen your tissues are actually absorbing — and that's what determines whether you're in the danger zone. A chamber at 2.0 ATA filled with normal air (21% oxygen) produces a ppO₂ of only 0.42 ATA. The same chamber with 100% oxygen produces a ppO₂ of 2.0 ATA. Same chamber pressure, completely different oxygen load on your body.
At sea level, breathing normal air: ppO₂ = 21% × 1.0 ATA = 0.21 ATA. Perfectly normal.
In a clinical chamber at 2.0 ATA, breathing 100% oxygen: ppO₂ = 100% × 2.0 = 2.0 ATA. That's nearly ten times normal.
The widely recognized threshold where CNS oxygen toxicity becomes a concern is around 1.6 ATA. Above this, your brain can be overwhelmed by reactive oxygen species — highly reactive molecules that damage cells. The most dramatic result is a seizure: you lose consciousness, your muscles convulse, and you are completely unable to help yourself.
Not a headache. Not a warning you can act on. A full tonic-clonic seizure that typically lasts 1–3 minutes and usually resolves on its own — once the oxygen source is removed.
Someone has to do that for you — because you'll be unconscious.
How often do seizures actually happen?
This isn't guesswork. Researchers have tracked seizure rates across hundreds of thousands of supervised clinical HBOT sessions. The numbers are remarkably consistent.
A Portuguese naval hospital tracked 188,335 sessions over 20 years. They recorded 43 seizures — a rate of about 1 in 4,400 sessions. An Australian and New Zealand study covering nearly 97,000 sessions found almost the same rate at 2.4 ATA.
Per session, that sounds rare. And it is.
But here's what the numbers look like from the patient's perspective. In the Portuguese study, the per-patient risk across an entire treatment course was about 1 in 220. That means roughly one out of every 220 people who undergo a full course of clinical HBOT will experience a seizure at some point.
And pressure matters enormously. One study found zero seizures in over 16,000 sessions at 2.0 ATA, while all seizures in their dataset occurred at 2.4 ATA and above. Higher pressure = dramatically higher risk. But "zero at 2.0 ATA" in one dataset doesn't mean it can't happen — other studies have documented seizures at and below 2.0 ATA.
Air breaks: the safety measure you can't afford to skip
Air breaks — short intervals of breathing normal air during an oxygen session — are one of the most effective tools for reducing acute seizure risk. Periodic breaks from high-concentration oxygen give your brain's antioxidant systems a brief window to recover before oxidative stress reaches dangerous levels.
The Costa et al. study that tracked 188,335 sessions found that 5-minute air breaks significantly reduced seizure frequency. In a clinic, a trained technician handles the timing — typically 20 minutes of oxygen followed by 5 minutes of air. You don't have to think about it.
At home, you do. And the failure modes are predictable: you lose track of time, you fall asleep, you skip a break because you're feeling fine. Some setups include automated timers that switch gas supply at programmed intervals — if you're operating at clinical pressures, this isn't a luxury feature, it's a critical safety system.
But air breaks reduce seizure risk — they don't eliminate it. You're still operating above the 1.6 ATA toxicity threshold during the oxygen phases. And they don't address fire risk, self-rescue incapacity, or any of the other issues in this article.
The finding that should worry every home user most
Here's the part that creates a false sense of safety.
In the Portuguese study, the median seizure occurred on the 21st session. The range was 1 to 126 sessions, but half of all seizures happened after session 20.
Only 4 out of 38 affected patients seized on their first treatment.
This means the typical pattern is: you complete 20 sessions, everything goes perfectly, you get comfortable, you stop worrying. And then session 21 — or 35, or 50 — is the one where things go wrong.
Previous success does not protect you. It simply hasn't happened yet.
When oxidative stress builds up slowly: Dr. Harch's 2024 findings
In 2024, Dr. Paul Harch — one of the most published researchers in hyperbaric medicine — published a study in Frontiers in Neurology that added an entirely new dimension to this picture.
Reviewing 60 cases over 33 years, Harch identified something the field had largely overlooked: chronic CNS oxidative stress from repeated HBOT sessions. This isn't a sudden seizure. It's a gradual accumulation of oxidative damage — caused by reactive oxygen species (ROS) — that builds up over many sessions.
Think of it this way: every HBOT session generates some degree of oxidative stress. Your body's antioxidant systems usually handle it. But when sessions are repeated frequently, intensively, or for too long — especially without adequate breaks — the oxidative burden can exceed your body's ability to recover. The nervous system starts to show signs of strain.
The symptoms Harch documented weren't dramatic seizures. They were subtler: increasing fatigue, cognitive decline, personality changes, regression of earlier improvements. The kind of changes that someone treating themselves at home might attribute to their underlying condition rather than to the treatment itself.
The average onset was after 103–116 atmosphere-hours of cumulative oxygen exposure — roughly 46–52 typical sessions. And a critical finding: once the oxidative stress was recognized and treatment was paused, most patients recovered. But when treatment continued despite worsening symptoms — which is exactly what happens when there's no clinical oversight — the situation got significantly worse.
Harch's dataset included reports from outside sources of three deaths and one patient who was institutionalized. These were not Harch's own patients — they were cases reported to him by other practitioners and families. The 60 cases came from both his own clinical practice and decades of correspondence from physicians and families seeking his expertise.
Importantly, two deaths and the institutionalization occurred specifically in home hard-shell chambers — cases where individuals continued intensive treatment protocols without professional monitoring to recognize the warning signs.
What actually happens when you have a seizure and no one is there
In a hospital or clinic, an oxygen toxicity seizure is a manageable event. A trained technician is watching. They see the early warning signs — facial twitching, sudden agitation — and act immediately: switch the gas supply from oxygen to air, begin controlled decompression, and prepare emergency support.
At home, alone, the sequence is very different.
You lose consciousness. A tonic-clonic seizure means you cannot think, move purposefully, press a button, or call for help. You are completely helpless.
The oxygen keeps flowing. Nobody turns it off. Nobody removes your mask. Your brain continues to be exposed to exactly the conditions that caused the seizure.
The seizure may repeat. Because the cause — high oxygen partial pressure — hasn't been removed, another seizure can follow. And another. Repeated seizures (status epilepticus) is a life-threatening emergency that requires immediate medical intervention.
You may vomit. Seizures commonly trigger vomiting. An unconscious person who vomits can aspirate — inhale vomit into their lungs. This alone can be fatal.
Nobody can reach you quickly. Even if someone in your house realizes something is wrong, they can't simply open a hard-shell chamber door. Internal pressure holds it shut with thousands of pounds of force. The chamber must be decompressed first — 10–15 minutes at safe rates. Decompress too fast and you risk barotrauma. During that entire time, you're sealed inside and unreachable.
Even with a soft-shell 2.0 ATA chamber, the situation is barely better. Higher-pressure soft chambers have reinforced closures that can't be unzipped instantly like a 1.3 ATA chamber, and the person inside is still unconscious and unable to self-rescue.
This isn't hypothetical. In 2020, a Dutch hospital published a case report about a 37-year-old man who died during his second HBOT session at 2.4 ATA. He seized, and despite an attendant being present and a physician arriving within three minutes, his body size (BMI of 52.6) made it impossible to reposition him in the confined chamber. He stopped breathing during the final minutes of decompression. CPR and intubation attempts failed. He died in a fully staffed clinical facility.
In another documented case, an 80-year-old man suffered a stroke following an oxygen toxicity seizure in a monoplace chamber. Because it was a monoplace unit pressurized with oxygen, the operators couldn't switch him to air — and the resulting prolonged oxygen exposure led to permanent brain damage.
If these outcomes can occur with professional supervision, what happens without it?
The fire risk: 140 deaths and a pattern that doesn't lie
Oxygen toxicity isn't the only danger. There's another one that's more sudden and more violent: fire.
Since 1923, at least 140 people have died in hyperbaric chamber fires worldwide. Fire accounts for roughly 90% of all hyperbaric chamber deaths.
The vast majority of these deaths — and the most catastrophic incidents — occurred in chambers filled with 100% pure oxygen. In a pure oxygen atmosphere, materials that barely burn in normal air ignite instantly and burn with terrifying intensity. Cotton. Wool. Human hair. A static spark is enough.
But fires haven't been limited to pure-oxygen environments. The deadliest single incident — Milan, Italy, 1997, 11 deaths — happened in a multiplace chamber pressurized with air, where patients breathed 100% oxygen through open masks. The exhaled oxygen gradually enriched the chamber atmosphere until a spark ignited it. All eleven died within seconds.
This "mask leakage" mechanism is important to understand. When you breathe high-concentration oxygen through any open mask, you exhale roughly 80% oxygen. That exhaled gas enriches the air around your face. Even if the overall chamber atmosphere is technically "air," the local environment near your mask is highly oxygen-enriched — and that's where ignition is most dangerous.
The most recent fatal incidents underscore how real this risk remains:
January 2025, Troy, Michigan: A five-year-old boy was killed when a monoplace chamber filled with 100% oxygen exploded at an unaccredited wellness center. The ignition source was static electricity from a pillow and blanket. Grounding straps were found unused in a drawer. No doctor on site, no licensed technician, no maintenance records. The center's CEO was charged with second-degree murder.
July 2025, Lake Havasu City, Arizona: A 43-year-old physical therapist died in a flash fire inside a chamber at his own facility. Electronic devices were found inside the chamber. The clinic was not UHMS-accredited.
Both incidents triggered an FDA safety communication and prompted Michigan to introduce the first state legislation requiring accreditation, physician oversight, and inspections for hard-shell chamber facilities.
The pattern across a century of data is unambiguous: every fatal chamber fire involved an oxygen-enriched atmosphere at clinical pressures. The higher the oxygen concentration and the higher the pressure, the more explosive the environment.
Why the clinical safety infrastructure exists
HBOT clinics aren't staffed out of tradition. Every layer of supervision exists because someone, somewhere, died without it.
Clinical HBOT facilities following UHMS (Undersea and Hyperbaric Medical Society) standards require a physician with hyperbaric training to be physically present — not just on-call by phone — during every treatment. Hyperbaric technologists must complete 40–60 hours of approved training plus 480 hours of supervised clinical experience before they can even sit for the certification exam.
Before each session, staff check your blood glucose (low blood sugar increases seizure risk), your temperature (fever increases oxygen toxicity risk), and your medications (some common drugs lower your seizure threshold without you realizing it). They verify you haven't brought prohibited items into the chamber. They monitor you continuously for early warning signs, using a well-known mnemonic: VENTID — Vision changes, Ears ringing, Nausea, Twitching, Irritability, Dizziness.
If they spot any of these signs, they can abort the session before a seizure occurs. You can't do this for yourself, because the most dangerous symptom — seizure — arrives without self-awareness. Your face may start twitching, but you might not feel it. Even if you do, the seizure may follow before you can act.
Fire safety follows NFPA 99 standards: fire suppression systems, prohibition of unapproved electronics, hyperbaric-compatible clothing, proper electrical grounding, and regular fire drills.
Online HBOT safety training courses do exist, and they're genuinely useful for understanding the risks. But knowing what to do and being physically capable of doing it while unconscious inside a sealed chamber are two very different things.
Who's most at risk — and you might not know you are
Several common conditions make you significantly more susceptible to oxygen toxicity. Some of them are so common you might not think twice:
Obesity and sleep apnea. These conditions cause chronic CO₂ retention, which dilates blood vessels in the brain and increases how much oxygen reaches brain tissue — exactly the mechanism that triggers toxicity. The Dutch man who died had a BMI of 52.6 and documented obstructive sleep apnea.
Fatigue and poor sleep. Sleep deprivation depletes your body's antioxidant defenses, lowering the threshold for oxidative stress. You might not even think about it before stepping into a chamber.
Fever. Even a mild fever increases your metabolic rate and lowers seizure threshold. A pre-session temperature check catches this in a clinic. At home, you might not bother.
Medications. Certain common drugs lower seizure threshold: some fluoroquinolone antibiotics, tramadol, and certain antidepressants. In a clinical setting, this is screened for. At home, you may not realize the interaction exists.
Low blood sugar. Hypoglycemia is a known seizure trigger that compounds oxygen toxicity. Easy to check, easy to skip when nobody's making you.
In a clinic, these factors are screened for before every session. At home, they're invisible — until they're not.
The math that changes everything: mild vs. clinical pressure
Now here's where the physics becomes your friend instead of your enemy.
Remember the critical threshold for CNS oxygen toxicity: a partial pressure of oxygen (ppO₂) around 1.6 ATA.
Let's start with the most common mild hyperbaric chamber setup.
In a mild chamber at 1.5 ATA, using an oxygen concentrator that delivers around 93% O₂ through a mask — though in practice, with mask dilution, the actual concentration reaching your lungs is likely somewhat lower:
ppO₂ = 93% × 1.5 ATA = 1.40 ATA (theoretical maximum)
That's well below the 1.6 ATA threshold. And in real-world use with some air mixing around the mask edges, the actual ppO₂ is lower still.
At 1.3 ATA with the same concentrator:
ppO₂ = 93% × 1.3 ATA = 1.21 ATA — even further from the threshold.
And if you're using the chamber with air only (no supplemental oxygen):
ppO₂ = 21% × 1.5 ATA = 0.31 ATA — essentially normal oxygen levels, just slightly concentrated. Nowhere near any toxicity concern.
Now compare that to a clinical chamber at 2.0 ATA with 100% oxygen — the kind being sold for home use:
ppO₂ = 100% × 2.0 ATA = 2.0 ATA
That's 25% above the toxicity threshold. Which is why clinical HBOT requires air breaks (breathing normal air every 20–25 minutes to give the brain a rest from oxidative stress), continuous monitoring, and trained staff ready to intervene.
This isn't a matter of being "more careful" at home. The numbers are on completely different sides of a physiological line. Mild hyperbaric pressures with concentrator oxygen physically cannot produce the oxygen partial pressures that cause acute CNS toxicity seizures.
A 2023 meta-analysis in Frontiers in Medicine confirmed this quantitatively: adverse events were statistically significant at 2.0 ATA and above, but not significant below 2.0 ATA.
The fire picture follows the same physics. In a mild chamber, the compressor pushes regular air, and the overall chamber atmosphere rises to roughly 25–33% oxygen when using a concentrator — elevated above normal (21%), but fundamentally different from 100% pure oxygen or the heavily enriched atmospheres in clinical chambers.
Every fatal hyperbaric chamber fire in recorded history occurred in an oxygen-enriched environment at clinical pressures. No fire has ever been documented in a mild hyperbaric chamber. Not one, across decades of worldwide use.
And if something did go wrong? You can exit a mild soft-shell chamber in about 60–90 seconds — just release the pressure and unzip. There's no decompression delay. No sealed door. The pressure differential of 0.3–0.5 ATM poses essentially zero risk from rapid exit.
Compare that to 10–15 minutes of controlled decompression from a 2.0+ ATA hard-shell chamber before anyone can even open the door.
A note on risk and responsibility
Articles like this one tend to get a predictable reaction: this is just fear-mongering to stop people from buying chambers.
It's a fair challenge. And it deserves a direct answer.
We sell chambers. I want people to buy them and use them at home — confidently, regularly, for years. That only works if they understand what they're working with. A customer who gets hurt is not a success story, no matter what was sold.
So let's be precise about what this article says and what it doesn't.
It does not say high-pressure HBOT doesn't work. It does not say you should avoid it. It says that at clinical pressures with medical-grade oxygen, the safety infrastructure in a professional facility — the trained technician, the pre-session screening, the ability to switch gas supply during a seizure, the fire suppression — exists because people were harmed without it. Every one of those layers was added after someone was injured or killed. Removing all of them at once by operating the same equipment at home, alone, is not a neutral decision.
The pushback usually comes down to a single idea: the risk is small, so why make a big deal of it?
Because low probability does not mean low consequences.
You don't wear a seatbelt because you expect to crash every time you drive. You wear one because the one time it matters, it really matters. The seizure rate in clinical HBOT is roughly 1 in 4,400 sessions. That's very low. But a seizure inside a sealed chamber, alone, with the oxygen still flowing and no one to decompress you for 10–15 minutes — that's not a minor inconvenience. It's a life-threatening emergency where you are physically unable to help yourself.
Acknowledging the severity of a consequence doesn't cancel out the rarity of the event. Both facts coexist. A responsible assessment includes both.
The difference between fear-mongering and safety education is simple: fear-mongering tells you to be afraid without telling you what to do about it. This article gives you the numbers, explains the mechanisms, and shows you exactly where the risk thresholds are.
Risk in hyperbaric therapy is not a binary. It's a continuum — and pressure, oxygen concentration, and supervision are the variables that determine where you sit on it. Moving down that scale reduces risk substantially. It doesn't eliminate it. Understanding the difference is not fear. It's the information you need to make a decision that's actually yours.
The bottom line
A high-pressure HBOT chamber — whether hard-shell or soft-shell — running at 2.0 ATA or above with medical-grade oxygen is a serious piece of medical equipment. In the right hands — in a properly accredited facility, with trained staff, fire suppression, grounding protocols, pre-treatment screening, and continuous monitoring — it can be used safely.
At home, alone, it's a fundamentally different situation. The oxygen toxicity risk is real and statistically predictable. The fire risk is real and has killed people. The inability to self-rescue during a seizure is absolute. And the false confidence that builds after 20 uneventful sessions is arguably the most dangerous factor of all.
Mild hyperbaric therapy at 1.3–1.5 ATA is a different category entirely. The oxygen partial pressures are physically below the acute toxicity threshold. The fire risk profile is fundamentally different. You can exit in seconds if anything feels wrong. And the most serious realistic risk is a compressor malfunction while you're asleep — something that proper equipment monitoring can catch, and that doesn't trap you inside a sealed chamber.
If you're considering hyperbaric therapy for home use, the question isn't "which one is more powerful?" The question is: "which one is safe to use without a medical team standing by?"
The physics answers that question clearly.
HBOT Made Simple is Brain Spa Hyperbaric's guide to understanding hyperbaric oxygen therapy — the technology, the terminology, and what to look for. Have a question we haven't covered? Contact us →
