IV Ozone Therapy Risks: What the Published Case Reports Actually Show

Why gas embolism is the risk that matters most

Ozone (O₃) is a reactive, unstable gas. In medical ozone therapy, it is generated on demand and mixed with oxygen, then used in one of several ways. The technique offered here is major autohemotherapy, or MAH: a set volume of the patient’s own blood is drawn out, mixed with the ozone-oxygen gas in a closed container, and reinfused after the ozone reacts with blood components.

The theory behind MAH is that the ozone reacts with and is consumed by the blood before reinfusion, so no free gas goes back into the vein. In practice, that depends entirely on technique: proper reaction time, a fully closed system, and no residual undissolved gas making it into the syringe or line. When that fails, even partially, gas can enter the bloodstream as a vein-borne bubble. From there it can travel to the lungs, the heart, or, if it crosses into the arterial side, the brain.

That crossing point matters more than most patients realize. A patent foramen ovale, a small flap-like opening between the heart’s upper chambers, is present in roughly one in four adults, and most people who have one do not know it. It typically causes no symptoms and is never diagnosed unless someone looks for it. In someone with a PFO, a gas bubble that would normally get filtered out by the lungs can instead cross directly into arterial circulation and reach the brain. This is not a rare anatomical curiosity. It is a risk factor that a meaningful share of otherwise healthy people carry without any way of knowing it in advance.

The case that should not be softened

In January 2025, Archives of Academic Emergency Medicine published a case report titled “Neurological Crisis Following Intravenous Ozone Therapy.” A previously healthy 36-year-old woman received an autotransfusion of 150 mL of ozone-infused blood, with heparin added, at a private clinic. Within minutes she had severe chest pain, fainted, and then had a witnessed generalized seizure lasting about three minutes. She arrived at the emergency department confused and agitated, with a Glasgow Coma Scale score of 9, a level consistent with a serious brain injury.

An MRI eight hours later showed multiple ischemic infarcts in the left thalamus and right cerebellum, plus smaller areas of injury in several other lobes, a pattern the authors describe as consistent with an embolic event. She was intubated and admitted to intensive care. An echocardiogram found a small patent foramen ovale, the kind of shunt described above, offering a plausible path for gas to reach her brain.

She improved enough to be extubated by day three and reached a normal consciousness level by day four. But the case did not end there. At six months she still had impaired speech, memory, and comprehension. At nine months she was still working through attention and concentration problems and frequent migraines, with ongoing speech and occupational therapy. This was not a near-miss with a quick recovery. It was a lasting neurological injury in someone who was healthy before the infusion.

This case is not the only one in the literature. A separate 2024 case report documented cerebral gas embolism and a multifocal ischemic stroke in a 58-year-old woman following an intradiscal (spinal disc) ozone injection, a different route entirely. The pattern across both is the same: gas introduced near the vascular system, reaching the brain, causing stroke-like injury. Gas embolism from ozone therapy is not confined to one clinic’s error or one delivery method.

What real screening and informed consent should involve

A clinic that takes this seriously does more than get a signature on a form.

Informed consent for ozone therapy should state, in plain language, that the FDA has not approved ozone gas for any therapeutic use and that gas embolism, including the kind that caused lasting neurological injury in the case above, is a recognized risk of the procedure. It should not soften that into something like “rare side effects may include discomfort.”

Screening should ask directly about any known cardiac shunt, prior stroke, or unexplained neurological event, since a diagnosed PFO or similar finding changes the risk calculation meaningfully. It should also be honest that most people with a PFO do not know they have one, so a clean screening history does not eliminate the risk, it only rules out what is already known.

Administration should happen with a physician’s oversight, in a setting equipped to recognize and respond to a neurological or cardiopulmonary event immediately, not somewhere a bad reaction means waiting for someone to notice. Our medical oversight page describes the physician-led standard we hold every IV service to, including this one. For how major autohemotherapy is actually performed, step by step, see what is major autohemotherapy. For a fuller look at what the published research does and does not support, our what the evidence shows page and our IV ozone therapy overview go further than this post does.