HBOT for Long COVID: Comparing the Global Evidence Efrati (Israel) vs HOT-LoCO (Sweden) vs Netherlands Registry

Hyperbaric oxygen therapy (HBOT) has become one of the more controversial and closely watched interventions in Long COVID research. What makes the field confusing is not a lack of data—but the fact that different countries have studied very different protocols, populations, and outcome measures, leading to seemingly conflicting results.

To understand the picture properly, it helps to look at the three most discussed evidence streams side by side: the Israeli Efrati group, the Swedish HOT-LoCO trial, and the Netherlands prospective registry data.

1- Israel: The Efrati Group and the “Neuroplasticity Protocol”

The Israeli studies led by Prof. Shai Efrati represent the most biologically detailed HBOT research in Long COVID.

Unlike many clinical trials that simply ask “does the patient feel better?”, these studies also attempt to measure what is changing in the brain itself.

Participants

• 73 adults

• Long Covid with ongoing symptoms (especially neurocognitive) for at least 3 months after infection.

The protocol

• 40 HBOT sessions 5x per week

• 2.0 ATA, 90 mins with 5 min air breaks every 20 mins

• Structured as a neuro-rehabilitation-style intervention

• Focus on persistent neurological Long COVID symptoms

 What they found

Patients showed improvements in:

• Cognitive function (memory, attention, executive function)

• Fatigue and sleep quality

• Emotional and quality-of-life measures

But what really made this work stand out was the imaging data:

• Increased cerebral blood flow in key brain regions

• Changes in white matter microstructure

• Signs consistent with neuroplastic reorganisation

 The interpretation

Efrati’s group frames HBOT not just as oxygen delivery, but as a trigger for neuroplasticity—a kind of “brain reboot environment” where oxygen, vascular changes, and metabolic signalling interact.

 Bottom line

This is the most mechanistically detailed and “biologically optimistic” dataset in the HBOT–Long COVID space.

2- Sweden: The HOT-LoCO Trial 

The Swedish HOT-LoCO trial is often considered the most methodologically rigorous HBOT Long COVID study to date—and its results are very different.

Participants

• 80 adults (18-60 years)

• Long Covid with reduced quality of life and physical function

The protocol

• Double-blind, placebo-controlled design

• Only 10 HBOT sessions over 6 weeks

• 2.4 ATA pressure

• Strong sham control group

What they found

Both the HBOT group and the placebo group improved over time.

But critically:

There was no statistically significant difference between HBOT and sham treatment.

What this suggests

The interpretation is not necessarily that HBOT “does nothing,” but rather that:

• Long COVID symptoms may fluctuate naturally over time

• Placebo and expectation effects can be strong

• Ten sessions may be too few to produce measurable physiological change

• 2.4 ATA may be too strong

3- Netherlands: Real-World Registry Data

The Netherlands data (from clinical HBOT centres and prospective registries) represents a completely different type of evidence: real-world clinical experience rather than controlled experimentation.

 The design

• 232 Long COVID patients treated in 6 clinical settings

• No placebo group

• Patient-reported outcome measures (quality of life, fatigue, physical function)

Patient Characteristics

• Long-term ill

• Functionally impaired

• Unable to work

• Median duration of illness was 20 months

What patients reported

Across follow-up periods, many patients reported:

• Improved energy levels

• Better physical endurance

• Reduced brain fog

• Improved quality-of-life scores

Protocol

• Pressure 2.4 ATA - 2.5 ATA

• Session duration 90-110 mins

• 5x days per week

• Usually 40 sessions

• Air breaks for 5 mins every 20 mins

Main Findings

Many patients improved - but not all.

Clinical meaningful improvement at 3 months were 56-63%, clinical meaningful worsening was 13-19%.

Possible reasons:

• HBOT may help some long-covid patients, but may worsen symptoms in others

• The protocols may have been too aggressive and produced oxidative stress in patients

 Putting the three together: Why the results look inconsistent

When viewed in isolation, these studies appear to conflict. But when compared side by side, a pattern starts to emerge.

1. Dose may be critical

• Sweden: 10 sessions → no clear difference

• Israel: ~40 sessions → measurable neurocognitive change

• Netherlands: often 20–60 sessions → subjective improvement

This raises an important possibility:

HBOT for Long COVID may be dose-dependent

2. Outcome measures matter

• RCTs (like HOT-LoCO) rely on strict statistical endpoints

• Registry studies capture subjective, lived improvement

• Efrati’s work combines both symptoms + brain imaging

So depending on what you measure, you may see:

• No difference (strict RCT endpoints)

• Clear improvement (patient-reported outcomes)

• Biological change (imaging data)

3. Long COVID is not one condition

A key emerging idea is that “Long COVID” is not a single uniform disease.

It likely includes multiple overlapping phenotypes:

• Neurocognitive (brain fog, memory issues)

• Autonomic (POTS-like symptoms)

• Fatigue-dominant

• Vascular/endothelial dysfunction

• Mixed systemic patterns

Different studies include different mixes of these groups, which can heavily influence results.

Final takeaway

The current evidence does not give a simple yes or no answer.

Instead, it suggests:

HBOT may have meaningful effects in certain Long COVID subgroups and dosing regimens—but the treatment response is highly dependent on protocol intensity, patient selection, and outcome measurement.

In other words, the science is not settled—but it is evolving quickly, and the differences between studies may be as important as the results themselves.

- Samantha Winters