One Learning Objective in Seven Medical Schools

Abstract clinical-education motif: a dark training room in perspective, rows of empty seat backs receding toward a bright wall screen showing a line drawing of a hyperbaric chamber, one front-row seat lit amber, faint dust in the projector beam, on a deep navy field

Of the eight recommendations the US Food and Drug Administration sent to hyperbaric oxygen providers in August 2025, one is a single sentence: ensure proper training is provided and maintained for staff with use of HBOT devices. The letter is precise about nearly everything else. It names the device class, the clearance route, the product code, two technical standards and the fabrics patients should not wear inside a chamber. It never says what proper training consists of, how long it takes, who is qualified to deliver it, or how a facility should know whether it worked.

That gap is not rhetorical. The same letter describes fires involving hyperbaric devices that caused serious injuries and deaths, states that the root cause of those events is not known, and adds that serious adverse events are rare. A word carrying that much weight — proper — is doing structural work in American hyperbaric safety policy with nothing published underneath it.

This month a systematic review tried to build the floor. A team at Xiangya Hospital of Central South University in Changsha, led by Shirui Pei, searched eight databases including PubMed, Embase, Scopus, the Cochrane Library and three Chinese sources for every study evaluating any hyperbaric teaching intervention, through June 2026. What they found is a field that has produced a great deal of enthusiasm and almost no measurement.

What thirty studies amount to

The review located 312 records, screened 221 after de-duplication, assessed 65 in full text and included 30. Those 30 were published between 2001 and 2025, ranged from 15 to 240 learners, and covered medical students, residents, postgraduate trainees, nurses and multidisciplinary teams. Nine were in English and 21 in Chinese.

The design census is the story. Four studies used randomised allocation. One was quasi-experimental. Six were single-group pre-post evaluations. The remaining nineteen were descriptive curriculum reports, training-needs surveys, cross-sectional assessments or a post-course evaluation. Risk of bias ran from some concerns in the four trials — largely because blinding learners to an educational intervention is close to impossible — to serious or critical in the non-randomised comparative studies, to high risk across the descriptive work. That heterogeneity ruled out a meta-analysis, so the authors synthesised narratively under the SWiM guideline and registered the protocol retrospectively.

Only two of the 30 studies reported enough quantitative detail to compute a standardised effect size. A 2022 Chinese randomised trial of problem-based learning on teaching ward rounds, with 32 trainees, reported case-analysis scores of 88.4 against 78.8, which works out to a Hedges’ g of 2.11 with a confidence interval of 1.24 to 2.98. A 2025 Italian pilot of a web-based hyperbaric chamber simulator, with 27 participants in a single group, produced a paired Cohen’s d of 2.36. The review’s authors do not celebrate either number. They call both implausibly large, note that the wider medical education literature typically reports effects in the 0.3 to 0.8 range for technology-enhanced teaching, and attribute the gap to small samples, absent blinding and practice effects.

Their conclusion is more restrained than the individual papers. Technology-enhanced modalities look promising, the reviewers write, but no study yet demonstrates consistent superiority over traditional didactic teaching.

The evidence base has widened without maturing

Pei and colleagues position their work against a paper most people in this niche have never read. In 2019, Sylvain Boet and colleagues at the University of Ottawa set out to review the evidence for simulation-based education in hyperbaric medicine and found none. Their search of MEDLINE, Embase and the Cochrane Central Register of Controlled Trials returned no English-language article assessing its impact at all; the only potentially relevant paper they could locate was in German.

Seven years on, the comparison is unflattering in a specific way. The number of comparative studies is still small, and the newer technology — web simulators, virtual reality, blended learning, scenario drills — has arrived almost entirely in single-group or quasi-experimental designs concentrated in Chinese-language journals. The review states it plainly: the field has expanded in breadth but not in methodological rigour.

That is not a verdict on the teaching. Simulation for central nervous system oxygen toxicity, the catastrophe every chamber team fears most, has been written up properly as a curriculum — Sadler and colleagues published a scenario covering toxicity during Treatment Table 6 alongside an intubated intensive care patient with mucous plugging — and Boet’s group has since run a Delphi process to identify which scenarios an international hyperbaric simulation curriculum ought to contain. The material exists. The trial that would show whether it saves anyone does not.

Upstream, the subject is barely taught at all

Before anyone can be trained to a standard, somebody has to teach them the subject exists. A 2023 audit by Zachary Talbot, Adriana Lee and Boet asked the bluntest possible question of Canadian medical schools: does your curriculum mention hyperbaric medicine at all. Seven of the country’s 17 schools supplied their pre-clerkship and clerkship learning objectives. Across all seven, the authors found exactly one objective related to hyperbaric medicine. In the other six schools’ objectives, the subject was absent entirely.

The sample is small and the response rate is a fair criticism. But the finding is consistent with surveys from other countries, and it carries an obvious consequence. Hyperbaric oxygen is a prescribed therapy with a defined indication list, and the doctors most likely to encounter a patient who needs it — emergency physicians, intensivists, wound-care and rehabilitation specialists — are the ones whose training does not mention it.

What doctors actually know

The review’s descriptive studies converge on the same measurement. In a 2020 survey published in Undersea and Hyperbaric Medicine, Kurt Magri and colleagues questioned 152 doctors licensed in Malta about hyperbaric indications and the location of the local chamber. Respondents who had visited a hyperbaric unit scored significantly better, and so did those who had attended a lecture on the subject. A previous unit visit, working in a hospital and a personal history of scuba diving all predicted willingness to refer a patient in future.

The direction of that effect is worth pausing on. Knowledge tracked exposure, and exposure tracked nothing in particular — not a curriculum requirement, not a licensing standard, not a mandatory rotation. A 2017 pilot study in the same journal took the more direct route of testing knowledge of indications outright. Both belong to a literature that keeps rediscovering the same deficit without anyone being assigned to fix it.

The frameworks exist; the measurement does not

It would be wrong to conclude that hyperbaric medicine has no standards for who may run a chamber. It has several, and they are not casual documents.

The European Committee for Hyperbaric Medicine maintains a code of good practice, first produced by a COST Action working group in 2004 and revised in 2022 by Jacek Kot and colleagues, which sets out requirements for facilities and personnel across the continent. The American College of Hyperbaric Medicine published a consensus statement on physician credentialing in 2014, addressed explicitly to hospital credentialing committees and intended to apply to non-hospital-affiliated centres as well. The National Board of Diving and Hyperbaric Medical Technology, a non-profit certifying body, examines and certifies hyperbaric technologists, hyperbaric nurses and diver medics, and lists among its purposes the establishment of standards for introductory training curricula.

That board’s most recent position statement, issued in October 2025 and revised the following month, is a useful measure of how far the professional consensus goes. It holds that a hyperbaric-medicine-trained provider — physician, nurse practitioner or physician assistant — must be immediately available to personally attend and support the certified technologist during chamber operations, with availability for hospital-based facilities set against the CMS final rule of November 2009 and the expectations of Medicare administrative contractors.

Federal payment policy reflects the same instinct without supplying the detail. Medicare’s national coverage determination for hyperbaric oxygen therapy, at version effective April 2017, lists the conditions it will pay for and classifies the service as one incident to a physician’s professional service. Its revision history records a 1999 transmittal that clarified the physician supervision requirement. The operative text sets no training standard, because coverage determinations are not written to.

So the frameworks are real, and what is missing sits between them: no validated instrument to assess whether a technologist or a physician actually acquired the competence, no consensus on what the core competence is, and therefore no way to audit any of it.

Why this matters beyond the curriculum

The FDA’s involvement reframes the question. When the agency names training as one of eight fire-prevention recommendations, it is treating the human operator as part of the device’s safety envelope, alongside grounding, monitoring, prohibited electrical items and maintenance intervals. Read the letter’s own language and the logic is unmistakable: the recommendations exist to reduce risk from a therapy the agency simultaneously describes as rarely producing serious adverse events.

Against that, the review’s most consequential finding is not the missing trials. It is that in the one place where it measured something, the evidence base produced numbers too large to believe. A field that reports a d of 2.36 from 27 trainees in a single group is a field that has not yet built the habit of scepticism its subject demands.

Two boundaries need stating, because this is a subject where overclaiming is the default failure mode. The first is that weak education evidence is not evidence that training fails. Nothing in these 30 studies suggests that a well-drilled chamber team is worse than an untrained one; it suggests that nobody has run the comparison properly, and that measuring education is genuinely hard. The second is that training cannot manufacture an indication. Hyperbaric oxygen is cleared for a specific list — decompression sickness, carbon monoxide poisoning, certain non-healing diabetic wounds, late radiation tissue injury, gas gangrene, severe anaemia, selected grafts and flaps, crush injury, thermal burns, idiopathic sudden sensorineural hearing loss and intracranial abscess among them — and a superbly trained operator delivering it for something else is still delivering an unproven treatment. The Undersea and Hyperbaric Medical Society is blunt on this point in its own indications document, describing minimally pressurised chambers in wellness centres and health spas as unproven, generally neither physician-prescribed nor supervised, and as putting those exposed at risk for serious injury and even death.

Which is where training and regulation meet. The settings where the fire risk is highest and the supervision is thinnest are the ones least likely to be reached by a credentialing consensus statement written for hospital committees.

What would settle the question

The review proposes five priorities, and only the last of them is the one that would change policy. The first four are methodological: adequately powered randomised trials with concealed allocation and blinded outcome assessment, a psychometrically validated hyperbaric knowledge and skills test, a Delphi consensus on a core competency framework, and head-to-head comparisons of virtual reality against web-based simulation against blended learning.

The fifth is longitudinal work linking educational outcomes to clinical safety metrics. Nobody in this literature has connected what a team was taught to what happened to the patient afterwards. Until someone does, proper training will remain a phrase in a federal letter, one sentence long, with thirty studies and four small trials standing behind it and not one of them able to say whether the staff in the room were ready.

The FDA, for its part, has said it will keep working with manufacturers and professional societies to encourage training and safe use of the devices. Encourage is the operative verb. The agency has not proposed to require anything, and on the present evidence it is not obvious what it would require if it did.

Sources

← Back to blog