Acta Neurochir Suppl (2008) 101:145-149
?9 Springer-Verlag 2008
Printed in Austria
Effect of hyperbaric oxygen on patients with traumatic brain injury
J. W. Lin 1'2, J. T. Tsai 3,4, L. M. Lee 5,6, C. M. Lin 1,2, C. C. Hung 7, K. S. Hung 1, W. Y. Chen 1,
L. Weil, Co P. Kol, y. K. Sul , W. T. Chiu 1,6
1 Department of Neurosurgery, Taipei Medical University-Wan Fang Hospital, Taipei, Taiwan
2 Graduate Institute of Clinical Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan
3 Institute of Biomedical Engineering, National Yang-Ming University, Taipei, Taiwan
4 Department of Radiation-oncology, Taipei Medical University-Wan Fang Hospital, Taipei, Taiwan
5 Department of Urology, Taipei Medical University-Wan Fang Hospital, Taipei, Taiwan
6 Graduate Institute of Injury Prevention and Control, Taipei Medical University, Taipei, Taiwan
7 Department of Neurosurgery, Taipei Medical University Hospital, Taipei, Taiwan
Hyperbaric oxygen therapy (HBOT) is the medical therapeutic use of
oxygen at a higher atmospheric pressure. The United States Food and
Drug Administration have approved several clinical applications for
HBOT, but HBOT in traumatic brain injury (TBI) patients has still
remained in controversial. The purpose of our study is to evaluate the
benefit of HBOT on the prognosis of subacute TBI patients. We pro-
spectively enrolled 44 patients with TBI from November 1, 2004 to
October 31, 2005. The study group randomly included 22 patients who
received HBOT after the patients' condition stabilization, and the other
22 corresponding condition patients were assigned into the matched
control group who were not treated with HBOT. The clinical conditions
of the patients were evaluated with the Glasgow Coma Scale (GCS) and
Glasgow Outcome Scale (GOS) before and 3 to 6 months after HBOT.
The GCS of the HBOT group was improved from 11.1 to 13.5 in
average, and from 10.4 to 11.5 (p <0.05) for control group. Among
those patients with GOS = 4 before the HBOT, significant GOS im-
provement was observed in the HBOT group 6 months after HBOT.
Based on this study, HBOT can provide some benefits for the subacute
TBI patients with minimal adverse side effects.
Keywords: Traumatic brain injury; hyperbaric oxygen; Glasgow
Coma Scale (GCS); Glasgow Outcome Scale (GOS).
Traumatic brain injury (TBI) is a major cause of death
and disability. Every year in the United States, there are
about one million head-injured people treated in hospital
emergency rooms, and roughly 50,000 people die from
Correspondence: Wen-Ta Chiu, M.D., Ph.D., Department of
Neurosurgery, Taipei Medical University-Wan Fang Hospital; Graduate
Institute of Injury Prevention and Control, Taipei Medical University;
111, Section 3, Hsing-Long Road, Taipei 116, Taiwan.
e-mail: wtchiu @tmu.edu.tw
TBI , 230,000 people are hospitalized, and 80,000
survive with significant disabilities. Because of the enor-
mous medical expenditure resulting from such injury,
many efforts have been devoted to minimize the influ-
ence of TBI.
Clinically, there are two mechanisms directly related
to the TBI outcome. The first one is the primary insult,
which results from the impact itself, and all the neuronal
damages are determined by the impact. As this insult has
already occurred before the patient comes to hospital,
there is little that a medical team can do for the patient.
The second mechanism is the delayed non-mechanical,
which results from tissue edema after the impact follow-
ed by ischemic change inside the brain. This theoretical-
ly preventable or treatable condition is the principal
target of treatment for TBI. All the medical treatment
should therefore be devoted to minimize edema and fa-
cilitate cerebral blood flow, to enhance cerebrovascular
autoregulation, to reduce cerebral metabolic dysfunc-
tion, and to adequately maintain cerebral oxygenation
. Furthermore, excitotoxic cell damage and inflamma-
tory process resulting from ischemia may also lead to
increased cell death . Generally speaking, about
80% of deaths in TBI result from hypoxia. Consequent-
ly, oxygen supplement in the initial resuscitation of a
TBI patient is of paramount importance.
Hyperbaric oxygen therapy (HBOT) is the medical
use of oxygen at a pressure exceeding atmospheric pres-
sure (ATA). The mechanism of HBOT consists in dras-
tically increasing oxygen partial pressure of the tissues,
146j. w. Lin et al.
and facilitating the oxygen transport by plasma. As a
result, HBOT can improve oxygen supply to the injured
brain and diminish the volume of brain that will necro-
tize during ischemia [5, 11].
Materials and methods
In this prospective cohort study, we intended to study the impact of HBOT
on moderate to severe TBI patients. The protocol of this study was ap-
proved by the Investigation Review Board-Wanfang Medical Center
(approval no. F950305). All the patients were enrolled under the regula-
tion of inclusion and exclusion criteria, and the criteria were as follows.
1. Age _> 16 y/o (Pediatric patient was excluded).
2. The patients were diagnosed to have moderate to severe TBI.
(Glasgow Coma Scale (GCS) from 3 to 12).
3. TBI condition stabilization.
4. Stable vital sign and spontaneous respiration without endotracheal
intubation or mechanical ventilation (tracheotomy was eligible for
5. No active infection or leucocytosis.
6. A hyperbaric oxygen department physician was consulted, and he
(she) agreed to treat the patient with HBOT.
7. Informed consent could be obtained from the patient's family.
1. Medical history with central nervous system disease (e.g. Parkinson
disease, dementia, congenital anomaly, stroke.., etc.)
2. Systemic disease history (e.g. diabetes, coronary artery disease, renal
insufficiency, COPD... etc.)
3. Multiple traumas (e.g. Chest contusion, abdominal blunt injury,
internal bleeding, pelvic fracture.., etc.)
4. Skull base fracture with CSF rhinorrhea or otorrhea.
5. Smoking or alcoholism.
6. Hemoglobin _< 10 gm/dl for female or < 12 gm/dl for male patients.
When the patients were enrolled, the assignment will be decided. If
one patient was chose randomly to be a study group candidate then there
will be another patient with corresponding condition (e.g. age, sex,
clinical course, severity and condition.., etc.) chose to be the matched
control group patients. If the patient was chose to be the study candidate,
the potential risk and benefit will be explained to the family for the
obtaining of the treatment consent. From Nov. 1, 2004 to Dec. 31, 2006,
there were total 62 patients enrolled into his study. Finally, there were 44
complete patients' data available for the further analysis.
In this research, we used a multi-user pressurized chamber (Model
no.: BTG/875/PV/02GOC-100, Apex Process Technologies (S) FFE
LTD, Singapore) to treat patients. The HBOT protocol was to apply two-
hour, two ATA pressures in the process. We increased the air chamber
pressure slowly to 2 ATA over 15 min and maintained this pressure for
90min. Patients were given 100% oxygen with 02 masks. Then we
depressurized to normal ATA over 15 min. The full treatment course
was defined as once a day for 20 days over a 4 week period.  During
the HBOT the patients' condition and vital sign were closely monitored.
If there was any complication (e.g. hypotension, short of breath, seizure,
unstable vital sign (blood pressure increased or decreased larger than
20mmHg), hypoxia (SaO2 <95%)... etc.) happened during HBOT, the
treatment for this patient will be discontinued, and this patient and the
corresponding controlled one will be excluded. The purpose of our study
was to clarify the influence on HBOT in subacute TBI patients by
analyzing patients' demographic information, GCS changes, and com-
paring variables such as the GCS, the injury severity, and the length of
time of HBOT, with regard to the subacute TBI outcome. And the TBI
outcomes were evaluated with the Glasgow Outcome Scale (GOS) con-
sisting of 5 levels (1: Death, 2: Vegetative status, 3: Severe disability, 4:
Moderate disability, and 5: Good recovery).
After data collection, we analyzed the result with SPSS 11.0 software.
We compared the two groups by means of demographic information,
including sex, age, body weight, injury timing, the severity of head
injury, duration of hospital stay, treatment with received surgery or that
without surgery, and length of HBOT treatment time. We compared the
GCS and GOS scores of both groups at different times with the Chi-
square test to assess difference between the two groups. The patients
were stratified with different GOSb levels (GOS level before the HBOT)
(GOSb - 2, GOSb = 3, and GOSb -- 4). There was no GOSb - 1 and the
GOSb- 5 patients enrolled because GOSb = 1 patients died before the
Table 1. Demographic information for patients
HBOT group Control group
Total 22 22
M:F 19"3 19:3
Body weight (kg) 61.75 65.12
GCS The patient initial GCS during admission.
GOSb The patient's GOS before HBOT or the same time for the control
group (mean 27.5 days after head trauma).
All variables have p > 0.05.
EDH Epidural hematoma, SDH subdural hematoma, ICH intracerebral
hemorrhage, SAH subarachnoid hemorrhage, DAI diffuse axonal injury.
Table 2. GCS improvement for the patients after HBOT
* p < 0.05 with significant difference.
Effect of hyperbaric oxygen on patients with TBI 147
Table 3. GOS outcome for the patients 3 and 6 months after HBOT
GOSb -- 2 HBOT
GOSb = 3
GOSb = 4
p < GOSb The patient's GOS before the HBOT or the same timing for the control group.
GOS3a The patient's GOS at 3-month post injury or at the same timing for control group.
GOS6a The patient's GOS at 6-month post injury or at the same time for control group.
p < 0.05 with significant difference.
HBOT and there was nothing could be improved for GOSb = 5 patients.
We recorded the GOS scores before (GOSb) and 3 (GOS3a) and 6
(GOS6a) months after HBOT or at the same time for the control group
patients to evaluate performance of the patients.
As showed in Table 1, 22 patients were enrolled in each
group. The M:F sex ratio was the same: 19:3 in both
groups. Most of the patients were aged between 25 and
64 years, which was also the most common range of age
for head injury. The average interval from injury to re-
ceiving HBOT was 27.5 + 5.8 days. This was also the
timing for the first GOS evaluation for both groups. If
the patients received HBOT, the average treatment times
were 24.4-+-7.8 times. In this table, no significant dif-
ference was found in age, sex, body weight, GCS severity,
presence or absence of surgical intervention, or GOS
severity between HBOT and control groups (all p > 0.05).
The average initial GCS scores for both groups' pa-
tients on arrival were 8.0 and 7.9, respectively. After
admission, surgical and/or medical treatments were ap-
plied to these patients, and the GCS recovery from 8.0 to
11.1 and from 7.9 to 10.4, respectively. We applied
HBOT to the patients after their traumatic condition
stabilization, and there was considerable improvement
in the HBOT group, from 11.1 to 13.5. In the control
group, GCS improved only from 10.4 to 11.5. Even in
this subacute stage of TBI, HBOT showed beneficial
effects on GCS improvement for moderate or severe
TBI patients (p<0.05) (Table 2).
The patients in both groups were stratified with the
GOSb level (GOSb = 2, 3, and 4) to evaluate the HBOT
effects on TBI patients. The outcome at the third and
sixth months after the HBOT was evaluated and analyzed.
In third month evaluation (Table 3), even though there
was some improvement in patients with HBOT, the num-
bers were not sufficient for drawing significant difference
and conclusion between study and control groups.
In sixth month evaluation (Table 3), there were 12
patients with improvement in the HBOT group, and 9
patients in the control group, but the difference did not
reach statistically significance (p > 0.05) for GOSb- 2 or
3 patients. However, there was a significant difference be-
tween these two groups among patients with GOSb : 4,
and as a whole the GOS6a (6 months after HBOT) im-
provement was greater in the HBOT group than in the
control group (p < 0.05).
Two patients developed seizures during the first week of
HBOT, and the convulsions were controlled with antic-
onvulsants. Then the patient resumed HBOT 2 weeks
later. Two patients experienced severe ear pain, and re-
ceived tympanostomy. Thereafter the ear pain subsided,
and the patient completed the full course of HBOT
successfully. No pulmonary adverse event, unstable vital
sign, or cataract occurred during HBOT or within 6
months follow-up. However, all these 4 study candidates
and their corresponding control patients were excluded.
The US Food and Drug Administration have approved
several clinical applications for HBOT. They included
certain non-healing wounds, radiation necrosis of soft
tissue and radiation osteonecrosis, carbon monoxide poi-
soning, decompression sickness, acute arterial ischemia,
and some sports injuries. These approvals did not in-
clude TBI. However, in the literature review; we found
reports showing some supports for HBOT application to
148j. w. Lin et al.
In 2004, the Agency for Healthcare Research and
Quality reviewed two fair-quality trials [1, 10], showing
fair evidence that HBOT might reduce mortality or the
duration of coma in severe TBI patients. But in one of
the trials, HBOT also implicated an increased chance for
poor functional outcome. Therefore, the evidences were
conflicting. Although these two trials are cited frequent-
ly, the methodologies of these two trials are also criti-
cized . In the past, HBOT was used under the
concepts of improving TBI patients' outcome and miti-
gating social economical expenditure . Previous stud-
ies have been focused on the immediate use of HBOT
after head trauma. However, during the initial period of
TBI, patients are often ventilator-dependent and may
have other associated injuries, such as lung contusion.
Under such situation, it is not convenient to treat TBI
patients with HBO early. With SPECT to show blood
flow improvement and to analyze them with different
age groups, Golden et al, reported that HBOT could im-
prove cerebral metabolism in the chronic stage of TBI
. There are only limited data in the literature to sup-
port beneficial effects of HBOT on TBI patients with
different degrees of severity.
In some animal study , using 2.5 ATA HBOT could
reach maximum microcirculatory hemoglobin oxygen
saturation and 2 fold of normal hemoglobin circulation
but there was also higher complication, such as pulmonary
system barotrauma, cataract, glaucoma, seizure ... etc.
[6, 10, 11]. The Rockswold et al. reported that using
1.5 ATA HBOT for 60 min is relatively safe without any
oxygen toxicity [5, 11]. In our series, we used HBOT
with 2 ATA for 90 min every day for a total 20 times.
What the optimal oxygen atmospheric pressure and
duration used in HBOT is need further clarification.
The timing of using HBOT around one month (27.5 +
5.8 days) after TBI is more practical in clinical condi-
tion. At that stage, patients have become more stable for
their cardiopulmonary function and often have received
intensive rehabilitation. As a consequence, the frequen-
cy of HBOT-related respiratory complications will be
reduced. Furthermore, the synergistic effect of rehabili-
tation with HBOT conceivably triggers the improvement
of the patient's GOS in the sixth month.
The adverse events in this trial were rare. Only two
patients had seizures during the initial period of HBOT,
and another two had middle ear barotrauma. The sei-
zure incidence in the previous reports was 2.4 per 10,000
patient-treatments [ 13]. Because the population base was
different between our results and this report, and we
focused on the head trauma patients only, there should
be higher incidence of seizure attack during the HBOT.
Besides, the sample size of our patients was too small
for statistical comparison. The tympanic membrane tear
is common in HBOT . However, in our HBOT center,
tympanoplasty was not routinely performed before
HBOT; this simple procedure should be considered
and performed before the HBOT to reduce the patient's
suffering. As the minor side effects, such as tinnitus,
aural fullness, disequilibrium, and vertigo and or nau-
sea, these side effects were all well tolerated by the
In this trial, we demonstrated that the GCS of TBI
patients in the HBOT group recovered significantly bet-
ter than in the control group (p <0.05) (Table 2). This
result would indicate that HBOT has a positive benefit in
GCS recovery of TBI patients. For the GOS improve-
ment, there was no obvious difference, especially for the
3 months follow-up (Table 3). Why did GCS improve so
much, and GOS did not? GOS is widely applied for TBI
patient outcome evaluation, but the intervals used for
GOS scores are too rough. In recent studies, an extended
GOS was used to evaluate the outcome of the TBI
patients, and more detailed evaluation might help our
For the patients receiving HBOT, there was no im-
provement of GOS in the third month follow-up, but 6
months after HBOT, GOSb=4 group got some im-
provement (Table 3). This situation could be explained
by the delayed effect of HBOT. That means HBOT
needs some more time to express the effects.
The TBI patients with GOSb-- 4 showed signifi-
cant GOS improvement six months after HBOT in the
study group (p<0.05, Table 3). But there was no such
difference in the GOSb-2
GOSb = 2 (vegetative state) and GOSb = 3 (severe dis-
ability) patients, there should be severe parenchymal
damages in the cerebral cortex. HBOT can not re-
generate necrotic neurons, but can only improve reox-
ygenation of the brain parenchyma. With incorporation
of rehabilitation, HBOT can help patients with mild
neurological deficits to recover and return to normal
life. In this prospective study, we can conclude that
HBOT can help TBI patients in GCS recovery and also
help patients with mild functional disability to lead a
or GOSb=3 groups. In
The authors would like to acknowledge the grant from the Department
of Health, Executive Yuan, Taiwan (DOH-TD-B-111-002), the grant
Effect of hyperbaric oxygen on patients with TBI 149
(NHRI-EX97-9707PI, NHRI-CN-HD9503S) from National Health Re-
search Institute, Taiwan, the grant (NSC 96-2314-B-038-010-MY2)
from National Science Council, Taiwan, and the grant (96001-62-010)
from Department of Health, Taipei, Taiwan.
1. Artru F, Chacornac R, Deleuze R (1976) Hyperbaric oxygenation
for severe head injuries: preliminary results of a controlled study.
Eur Neurol Res 14:310-318
2. Bennett MH, Trytko B, Jonker B (2004) Hyperbaric oxygen therapy
for the adjunctive treatment of traumatic brain injury. Cochrane
Database Syst Rev Oct 18(4): CD004609
3. Fiesseler FW, Silverman ME, Riggs RL, Szucs PA (2006)
Indication for hyperbaric oxygen treatment as a predictor of
tympanostomy tube placement. Undersea Hyperb Med 33(4):
4. Golden ZL, Neubauer R, Golden C J, Greene L, Marsh J, Mleko A
(2002) Improvement in cerebral metabolism in chronic brain
injury after hyperbaric oxygen therapy. Int J Neurosci 112(2):
5. Hardy P, Johnston KM, De Beaumont L, Montgomery DL, Lecomte
JM, Soucy JP, Bourbonnais D, Lassonde M (2007) Pilot case study
of the therapeutic potential of hyperbaric oxygen therapy on chronic
brain injury. J Neurol Sci 253(1-2): 94-105
6. McDonagh M, Helfand M, Carson S, Russman BS (2004)
Hyperbaric oxygen therapy for traumatic brain injury: a sys-
tematic review of the evidence. Arch Phys Med Rehabil Jul
7. Meirovithz E, Sonn J, Mayevsky A (2007) Effect of hyperbaric
oxygenation on brain hemodynamics, hemoglobin oxygenation and
mitochondrial NADH. Brain Res Rev $4:294-304
8. NIH, Traumatic Brain Injury: Hope Through Research 2002
9. Horn CE, Himel HN, Selesnick SH (2005) Hyperbaric oxygen
therapy for sudden sensorineural hearing loss: a prospective trial
of patients failing steroid and antiviral treatment. Otol Neurotol
Sep 26(5): 882-889
10. Rockswold GL, Ford SE, Anderson DC, Bergman TA, Sherman RE
(1992) Results of a prospective randomized trial for treatment of
severely brain-injured patients with hyperbaric oxygen. J Neurosurg
11. Rockswold SB, Rockswold GL, Defillo A (2007) Hyperbaric
oxygen in traumatic brain injury. Neurol Res 2:162-172
12. Werner C, Engelhard K (2007) Pathophysiology of traumatic brain
injury. Br J Anaesth 99(1): 4-9
13. Yildiz S, Aktas S, Cimsit M, Ay H, To~rol E (2004) Seizure
incidence in 80,000 patient treatments with hyperbaric oxygen.
Aviat Space Environ Med 75(11): 992-994