HBOT for Long COVID: One Positive Sham-Controlled Trial, One Negative One

What is established about HBOT itself

Hyperbaric oxygen therapy means breathing high-concentration oxygen inside a pressurized chamber. It is a mainstream, physician-supervised procedure with long-standing, well-evidenced uses: treating decompression sickness in divers, carbon monoxide poisoning, certain non-healing wounds, and radiation injury after cancer treatment. Those indications are backed by decades of clinical use and clear physiological rationale, and are not in question here.

What is not established is whether the same procedure, applied to a very different problem, namely the fatigue, brain fog, and reduced function that persist for months after a COVID-19 infection, produces a real benefit. That is a separate clinical question with its own, much thinner evidence base, and it is the question this review addresses.

What the trial evidence for long COVID actually shows

Two randomized, sham-controlled, double-blind trials have tested HBOT specifically for long COVID, and they reached opposite conclusions. The first, an Israeli trial in 73 patients, gave 40 HBOT sessions over roughly two months and compared them against a sham chamber protocol designed to feel similar to patients. It found statistically significant improvements in global cognition, attention, executive function, energy, sleep quality, and psychiatric symptoms in the HBOT group, along with changes visible on brain perfusion imaging.[1]

The second, a Swedish trial known as HOT-LoCO, randomized 79 patients to 10 HBOT sessions or sham treatment and measured physical functioning at 13 weeks. It found no significant difference between groups. Both the HBOT and sham groups improved over the study period by a similar amount, which is itself informative: it suggests that at least part of what looks like a treatment effect in uncontrolled long COVID studies is really the natural course of the condition, plus a real placebo response, plus regression to the mean.[2]

Both studies were properly randomized, properly blinded, and used a genuine sham comparator, which is the strongest design available for a therapy where the setting itself (sitting in a chamber, being cared for, expecting benefit) could plausibly produce non-specific improvement. Neither trial can be dismissed as poorly designed. They simply disagree.[1][2]

The most likely explanation is dose, but that is a hypothesis, not a finding

The most obvious difference between the two trials is intensity: 40 sessions versus 10. If HBOT has a genuine biological effect on post-viral brain function, it is plausible that effect requires a threshold dose to appear, and that 10 sessions simply is not enough. This is a coherent hypothesis and matches how HBOT is dosed in some of its established indications, where longer courses are used for slower-healing problems.[1][2]

But the two trials also measured different things. The positive trial's primary signal was neurocognitive: attention, executive function, memory. The negative trial's primary outcome was physical functioning, a different domain that may simply respond less to this intervention regardless of dose. A patient reading both studies should notice that 'HBOT did not help physical functioning at 10 sessions' is not the same claim as 'HBOT does not help cognition,' and neither trial answers both questions in both directions.[1][2]

What remains genuinely uncertain

There is no long-term follow-up in either trial. The positive trial assessed patients only one to three weeks after their final session, so whether the cognitive gains persisted at three months, six months, or a year is simply not known. A benefit that fades quickly is a much less compelling reason to undergo dozens of chamber sessions than a durable one.[1]

Neither trial is large enough, on its own, to define which patients might respond. Long COVID is not one condition; it covers people with predominant fatigue, predominant cognitive complaints, predominant autonomic symptoms, and combinations of all three. A trial of 73 or 79 patients cannot reliably tell you whether HBOT helps the cognitive-predominant subgroup specifically, even if it shows an average effect across the whole group. Until a larger, dose-ranging trial with longer follow-up is done, the honest answer is that the evidence is mixed and the reasons for the mismatch are not fully resolved.[1][2]

What this means for you

This is exactly the kind of split evidence base where a physician needs to look at the individual patient rather than at a headline claim of 'HBOT cures long COVID' or a dismissal of 'HBOT does nothing for long COVID.' Both of those statements go beyond what the two trials actually show.

If you ask about HBOT for post-COVID symptoms, the conversation starts with diagnostics, not a chamber booking: a clinical assessment of which symptom domain is dominant (cognitive, physical, autonomic), a review of other explanations for persistent fatigue or brain fog, and a look at what the sham-controlled evidence does and does not show. If HBOT looks appropriate for you, your physician decides the course and tracks your response against your own baseline. There is no fixed protocol for long COVID HBOT, no promised session count, and no claim that this is a proven cure. The evidence does not support that framing, and neither do we.[1][2]