Print Post Approved
ISSN 1833 3516
ABN 29 299 823 713
Volume 42 No. 1 March 2012
The Journal of the South Pacifi c Underwater Medicine Society (Incorporated in Victoria) A0020660B
and the European Underwater and Baromedical Society
Diving and Hyperbaric Medicine
Perfl uorocarbon emulsion for severe DCS
Direct effect of Co2 on apnea-induced haemoglobin increase
how consistent are doctors in assessing ‘fi tness to dive’?
The health of recreational dive masters and instructors
Risk factors for rapid ascent and buoyancy problems
Scuba diver’s pulmonary oedema can be fatal
ultrasound under pressure
SOUTH PACIFIC UNDERWATER
application,?go?to?the?Society’s?website: <www.spums.org.au> ?
EUROPEAN UNDERWATER AND
PURPOSES OF THE SOCIETIES
To promote and facilitate the study of all aspects of underwater and hyperbaric medicine
To provide information on underwater and hyperbaric medicine
To publish a journal and to convene members of each Society annually at a scientific conference
Immediate Past President?
Member at Large 2011?
Member at Large 2010
Member at Large 2009
Honorary Treasurer & Membership Secretary
Phone & Fax:?+44-(0)20-85001778
DIVING and HYPERBARIC MEDICINE
Diving and Hyperbaric Medicine is published jointly by the South Pacific Underwater Medicine Society
and the European Underwater and Baromedical Society
Submissions to the Journal should be sent to: <firstname.lastname@example.org>
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20121
The Editor’s offering
Undertaking research is a challenge, and never more so than
in primary health care. Whilst general practitioners may see
opportunities for clinical investigation amongst the range
of pathologies they deal with in their everyday practice and
have potential opportunities to contribute to epidemiological
studies, there are major barriers to turning these into a project
that can be seen through successfully. This is yet more so in
a narrow field such as diving medicine. Amongst the barriers
they face are lack of time (research for most must be done
in their own time), minimal resources, both financial and
professional, limited training in research methodology and
the ever present conflict with needing to earn a living for their
staff and themselves. You have to be either totally dedicated
or mad, perhaps a little of both. One way around the obstacles
is to enlist outside help from ‘experts’, both as mentors and
active participants in a project. Such arrangements may
be formal (e.g., supervising a project for a post-graduate
qualification) or an informal collegiate relationship.
A few years ago, Mike Bennett and I presented a session at
a Hyperbaric Technicians and Nurses Association Annual
Scientific Meeting on how to set about a research project.
As part of it, Mike discussed how to do a literature search
and what sorts of research might be achievable (Table 1),
whilst I discussed the components that make up a research
project (Table 2). These apply, in general, to all research.
Few people appreciate how many preparatory steps must
be taken before actually doing the research, and a common
reason for failure or for a less than satisfactory outcome is
lack of sufficient attention to these preliminaries.
In this issue, we have two good examples of research in
a primary health setting. Greg van der Hulst started his
project (towards a distance-learning Postgraduate Diploma
in Medical Science – Diving and Hyperbaric Medicine from
the University of Auckland) whilst he was a junior resident
in emergency medicine at Whangarei Hospital, completing it
subsequently whilst in a busy general practice in Northland,
New Zealand.1 In the process, he enlisted the help of David
Doolette, a physiologist at the US Naval Experimental
Diving Unit, Panama City, and whose methodology he
employed, and Peter Buzzacott, who at the time was a
doctoral candidate at the School of Sports Science, the
University of Western Australia. Whilst Chris Sames holds
a small part-time appointment at the Slark Hyperbaric Unit,
he is predominantly employed as a general practitioner (GP)
in the Naval Health Unit in Auckland, and his project was
conducted in his own time.2
Other examples of general practitioners publishing
independent research in the pages of Diving and Hyperbaric
Medicine within the past few years are Cathy Meehan, a GP
in Cairns (who enlisted Mike Bennett’s help) and Douglas
Walker with Project Stickybeak (now incorporated into the
DAN Dive Fatality Reporting Project).3,4 We encourage GPs
to pursue diving medicine topics of interest to them; there are
plenty of people within our two societies keen to help.
What types or classes of projects are achievable?
Magnitude – how big is the problem?
Therapy or intervention – what works?
Diagnosis – what is the best way to tell if someone has...?
Equipment – does this ‘thingy’ do what it should?
Quality – what works within our system, and why?
Cost – how much does it cost to achieve what we can do
in our system?
Teaching – how effective is the instruction process?
Elements of a research project
1. Asking a question
2. Doing a literature search
3. Understanding the literature
4. Making a plan
5. Finding somewhere to do it
6. Finding people to provide advice and help
7. Finding people/animals/stuff to do it on
8. Finding/costing equipment and materials
9. Writing a proposal
10. Obtaining ethical approval
11. Getting the money
12. Doing the work
13. Analysing the data
14. Presenting the results
15. Keeping everyone “sweet as”
1 van der Hulst GA, Buzzacott PL. Diver Health Survey
score and probability of decompression sickness among
occupational dive guides and instructors. Diving Hyperb Med.
Sames C, Gorman D, Mitchell S. Postal survey of fitness -
to- dive opinions of diving doctors and general practitioners.
Diving Hyperb Med. 2012;42:24-9.
Meehan CA, Bennett MH. Medical assessment of fitness to
dive – comparing a questionnaire and a medical interview-
based approach. Diving Hyperb Med. 2010;40:119-24.
Walker D, Lippmann J, Lawrence CL, Fock A, Wodak T,
Jamieson S. Provisional report on diving-related fatalities in
Australian waters 2005. Diving Hyperb Med. 2010;40:131-
The front page photo of Cairns professional musician
and diver Kirtley Leigh was taken by Bob Halstead, well
known to many members for his entertaining writings
in the diving magazines. In 2008, Bob was inducted into
the International Scuba Diving Hall of Fame.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20122
The Presidents’ pages
website is at
Members are encouraged to log in and to
keep their personal details up to date
In 2002, during the course of the Cooperation in Science
and Technology (COST) B14 Action, we had the opportunity
to develop and start a multicentre research protocol on the
treatment of idiopathic sudden sensorineural hearing loss
(ISSHL), more commonly called ‘sudden deafness’. The
COST Action was a European Commission-sponsored
consorted action, and the funds allowed us not to run
the trial itself but to coordinate this and other hyperbaric
oxygen therapy (HBOT) evidence-based and quality-related
issues (for a full overview of the COST Action, visit the
website <www.oxynet.org> or the European Committee for
Hyperbaric Medicine website: <www.echm.org>).
Only three hyperbaric centres (out of five involved) actually
enrolled patients in this study, which had a very ambitious
protocol. Current practice for treating ISSHL consists of
high-dose cortisone, a treatment that, together with the
spontaneous recovery rate within the first 10 days, results
in return of useful hearing in about 70% of cases. The
remaining patients have a poor likelihood of recovery, and
that was precisely our target group. Retrospective studies
had indicated that in this subgroup of patients, HBOT could
result in further improvement in about 40–50%.
The study was probably too ambitious. We recognised
that ISSHL has multiple causes, from vascular to viral
to auto-immune to trauma, and that what is considered
‘sudden deafness’ may only manifest itself as a minor
tonal audiogram change. We wanted to standardise our
study population as much as possible by maintaining strict
inclusion criteria. Then, patients were randomised to a
10-day course of ‘HBOT’ or ‘no HBOT’. Providing sham
hyperbaric treatments was technically and logistically not
possible in four out of the five centres, and, furthermore,
it was considered that sham compression would result in
possible side effects.
By 2007, it had become evident that less than 1% of cases
labelled as ISSHL were eligible for the study, making the
general applicability of the results questionable. Most
patients were presenting too late to be enrolled, but many
were excluded by their ENT surgeon on the basis of a
subjective feeling that the patient ‘should get all chances
possible’. Over the course of almost nine years, about 100
patients will be analysable, a task which will be undertaken
now. The results will be heavily criticised, no doubt – a pity,
because many were awaiting them anxiously.
A Sydney group, driven by the current SPUMS President,
Mike Bennett is now engaged in a similar study, with sham
compression and inclusion criteria that are much wider;
more importantly, they have a unique cooperation with the
ENT surgeons from the region, making inclusion of patients
possibly much easier.
In the meantime, the Hyperbaric Oxygen Committee of the
Underwater and Hyperbaric Medical Society (UHMS) has
officially recognised ISSHL as an indication for HBOT. I
am not quite sure whether to be happy or sad at this news.
On the one hand, many patients will now probably be
able to benefit from this treatment and add to the already
substantial database of retrospective studies. On the other
hand, I can already see the difficulties in convincing ENT
surgeons to participate in randomised prospective trials on
ISSHL: ‘Has it not been recognised as an indication? Is not
the UHMS one of the major players in the field of HBOT and
its evidence base?’ I fear the good intentions of the UHMS
Committee may make our task – to prove that HBOT can
contribute significantly to the treatment of ISSHL – more
difficult than before.
As you read this message, it is time to send your abstracts
and register for our Annual Scientific Meeting. This year,
the location is Belgrade (Serbia), and it will be preceded
by an ECHM Consensus Conference. The location and
organisation look excellent, and the registration fees and
accomodation prices are as low as we have not seen in years
– so there is no excuse not to attend (www.eubs2012.org).
Plans for the 2013 ASM are well underway too: a joint
EUBS–SPUMS Meeting, halfway between our continents
in the Indian Ocean (Reunion Island). The South African
Underwater and Hyperbaric Medical Association is keen to
join as well, so we are talking about a tri-continent meeting
on diving and hyperbaric medicine - and all are willing to
make this a big success !
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20123
Another year has passed in the life of SPUMS, and that life
continues to be full of interest. Committee work seems to
involve a lot of heads down burrowing through the detail,
so it is a great pleasure to step back and try to give you all
an overview of how things are going.
You will all be aware that this year we will hold our 41st
ASM at the Madang Resort just outside the town of Madang
on the north coast of Papua New Guinea. As I write, I am
happy to say the recent political crisis seems substantially
settled and all is looking good for our arrangements. This
is our second visit to Madang, and those who were there in
2001 will remember it very fondly (hard to believe it was that
long ago!). Cathy Meehan has done a great job getting it all
together and the programme is looking full of interest to our
members. The theme is “What lies beneath: the pleasures
and perils of our diving environment”. Cathy has organised
two world-class speakers in Associate Professor Jamie
Seymour (AKA ‘the jelly dude’) and Richard Fitzpatrick
(AKA ‘the shark guy’), both from James Cook University
in Cairns. I have seen some of their presentations in other
forums, and I can thoroughly recommend them to you. The
shark wrestling videos are particularly engaging! We will
also be continuing our popular diving and hyperbaric update
workshops. All details are on the SPUMS website <www.
SPUMS.org.au> along with the links to register and book
accommodation and flights to suit your purposes. I look
forward to seeing many of you there.
On the subject of ASMs, Cathy has also agreed to head up
our new ‘future meetings’ sub-committee. This is a group
constituted at our last AGM, and given the task of seeking out
interesting destinations for the Society, along with individual
members who would be willing to convene those meetings.
At present the sub-committee consists of Cathy, Janine
Gregson and Sue Paton, but if you are willing to assist with
your time or even simply to put an idea forward, you will be
welcomed with open arms. Please contact Cathy for more
detail. (NOTE: membership of this sub-committee does not
indicate you are willing to convene a meeting!)
For the immediate future, we are planning a joint meeting
with the EUBS and SAUHMA (South African Underwater
and Hyperbaric Medical Association) in Réunion in 2013
(date to be determined). Our secretary, Karen Richardson
has put her hand up to convene this meeting for us, so watch
the website and this journal for more information on what is
sure to be a true watershed meeting for all three societies.
The great and continuing project that is joint ‘ownership’ of
the Journal with the EUBS continues. The meeting in 2013
will be a great opportunity for members of both societies to
get together and discuss all those things that are of common
interest to us. The Journal continues to go from strength to
strength and must count as SPUMS’ greatest achievement of
the last few years – largely due to the continuing efforts of
our evergreen editor. More strength to him! The successful
listing on Medline is a dispassionate recognition of just how
far we have come. An agreement to continue joint ownership
of this Journal is accepted in principle, and the editorial
contract to cover 2013 onwards is now being prepared.
On a less rosy note, the Committee (and in particular our
Education Officer, David Smart) has been doing battle
on several bureaucratic fronts. Of most direct interest to
SPUMS members is the growing practice throughout most of
Australia for dive training agencies to drop the requirement
for a medical examination prior to dive training. Such a
medical remains a firm recommendation from this society
and we are vitally interested in hearing any comments from
our members – and particularly any experiences you have
had of direct consequences from this change in policy.
We are also fighting hard on two other fronts. Firstly, David
has formulated a very lengthy reply to proposed changes to
the Work Health and Safety Diving Regulations and their
wide implications for the safety of occupational divers in
Australia – these, along with the proposed abandonment
of local Standards in the area are likely to greatly impact
the future of professional diving in our region. Secondly,
both David and I are currently embroiled in the continuing
evaluation by the Medicare Services Advisory Committee
of hyperbaric oxygen indications. At the time of writing, we
are waiting to see a draft report from the Committee on the
continuing support for non-diabetic wounds and soft-tissue
radiation injuries. Watch this space…
So it is all go here, as ‘Punter’ (former Australian cricket
captain, Ricky Ponting) scores his first ‘ton’ for two years
and our new captain (‘Pup’ Clarke) has knocked up his first
triple ton. It is good to be alive in a Sydney summer. All the
best to all of you for the New Year and I look forward to
seeing many of you in Madang. If not there, then perhaps
website is at
Members are encouraged to log in and to
keep their personal details up to date
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20124
Effect of hypercapnia on spleen-related haemoglobin increase
Matt X Richardson, Harald K Engan, Angelica Lodin-Sundström and Erika Schagatay
(Richardson MX, Engan HK, Lodin-Sundström A, Schagatay E. Effect of hypercapnia on spleen-related haemoglobin
increase during apnea. Diving Hyperb Med. 2012;42(1):4-9.)
Background: Splenic contraction associated with apnea causes increased haemoglobin concentration and haematocrit (Hct),
an effect that may promote prolonged breath-holding. Hypoxia has been shown to augment this effect, but hypercapnic
influences have not been investigated previously.
Methods: Eight non-divers performed three series of apneas on separate days after inspiration of oxygen with different
carbon dioxide (CO2) levels. Each series consisted of three apneas 2 minutes apart: one with pre-breathing of 5% CO2 in
oxygen (O2, ‘Hypercapnia’); one with pre-breathing of 100% O2 (‘Normocapnia’); and one with hyperventilation of 100%
O2 (‘Hypocapnia’). The apnea durations were repeated identically in all trials, determined from the maximum duration
attained in the CO2 trial. A fourth trial, breathing 5 % CO2 in O2 for the same duration as these apneas was also performed
(‘Eupneic hypercapnia’). In three subjects, spleen size was measured using ultrasonic imaging.
Results: Haemoglobin concentration increased by 4% after apneas in the ‘Hypercapnia’ trial (P = 0.002) and by 3% in the
‘Normocapnia’ trial (P = 0.011), while the ‘Hypocapnia’ and ‘Eupneic hypercapnia’ trials showed no changes. The ‘easy’
phase of apnea, i.e., the period without involuntary breathing movements, was longest in the ‘Hypocapnia’ trial and shortest
in the ‘Hypercapnia’ trial. A decrease in spleen size was evident in the hypercapnic trial, whereas in the hypocapnia trial
spleen size increased, while only minor changes occurred in the other trials. No differences were observed between trials
in the cardiovascular diving response.
Conclusion: There appears to be a dose-response effect of CO2 on triggering splenic contraction during apnea in the
absence of hypoxia.
Breath-hold diving, carbon dioxide, hypercapnia, haematology, respiration, physiology
Apneic diving is associated with several physiological
adjustments in order to maintain brain and heart function
during interrupted gas exchange with the environment,
the best described of which is the cardiovascular ‘diving
response’ consisting of bradycardia and peripheral
vasoconstriction.1 The human diving response has been
found to be oxygen-conserving, likely owing to both the
reliance of non-perfused areas on anaerobic metabolism,
and to the bradycardia, limiting the oxygen demand of the
myocardium.2,3 The diving response is initiated by apnea
and may be modified by face immersion and possibly by
Recent work suggests that splenic contraction may also be a
protective response which serves to increase body gas storage
capacity by elevating circulating red cell mass.6,7 Increases
in haemoglobin concentration (Hb) and haematocrit (Hct)
have been demonstrated during both single and repeated
apneas performed within short intervals.7–9 The increases
in Hb and Hct are related to contraction of the spleen,
an effect that is maximised after three to five apneas and
reversed within 8–9 minutes after cessation of the series of
apneas.6,7,10 These changes may increase oxygen-carrying
capacity and carbon dioxide (CO2 ) buffering during apnea
and have been shown to prolong breath-hold time across a
series of apneas.7
The correlation between changes in Hb and Hct and splenic
contraction is strong, and it is estimated that approximately
60% of the change in these parameters during apnea
can be directly attributed to the emptying of the spleen’s
stored contents.7,11 This response does not appear to be
affected by face immersion, which makes it different to the
cardiovascular diving response, which is fortified by face
immersion.12,13 It has been shown that the magnitude of
the spleen-related Hb increase is augmented by hypoxia,14
but there may be other apnea-related components that
cause some contraction even in the absence of hypoxia. Of
these, hypercapnia is a strong candidate as it is a largely
unavoidable consequence of cessation of breathing.14 In
a recent study, we found that apnea or hypoxic breathing
resulted in different levels of splenic contraction despite
similar levels of arterial oxygen saturation (SaO2), with the
response to apnea being twice that of hypoxia breathing.15
One explanation for this could be the high partial pressure of
carbon dioxide (PaCO2) arising from apnea, but other apnea-
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20125
induced mechanisms could also be involved. It remains to be
tested whether PaCO2 has a separate initiation or modifying
effect on splenic contraction.
Previous research shows that reaching a threshold level of
CO2 initiates both the ‘struggle phase’, defined as the onset
of involuntary breathing movements, and the end point
of apnea, at least in novice apnea subjects.16 Therefore,
hyperventilation can prolong apneic duration by reducing
the CO2 content of the tissues and blood, so that the breaking
point of apnea is reached later, which is beneficial for the
diver when sufficient O2 exists. However, if CO2 has a role in
inducing spleen contraction, hyperventilation could prevent
the development of this apnea-prolonging response. In order
to reveal the separate role of the PaCO2 stimulus we examined
changes in haematological parameters and splenic volume
during apneas conducted at varying PaCO2 levels without
the influence of hypoxia.
Four male and four female subjects of mean (SD) age 28 (7)
years, weight 78 (19) kg and height 176 (11) cm volunteered
for the study. Mean vital capacity for the subjects was 5.0
(1.0) L. Subjects signed a consent form after being informed
of the experimental protocol, which was in accordance with
the Declaration of Helsinki and had been approved by the
regional human research ethics board at Umeå University,
Sweden. All were non-smokers although one subject used
snuff. Subjects were involved in physical exercise for an
average of 2.9 (2.7) h per week for general fitness. Subjects
had only limited lifetime experience in breath-holding, with
no current activity.
The subjects completed four experimental trials spaced
by at least 24 hours. Each trial consisted of three apneas
spaced by 2 minutes of rest. Hypoxia was eliminated by O2
breathing and apnea times held constant in all tests allowing
the capnic influence to vary independently. In order to reveal
any effect of hypercapnia without apnea, a fourth test using
eupneic hypercapnia was included. The individual apneic
times produced in the hypercapnia trial were repeated in
the following trials, which were performed in a randomised
order. The four trials were thus:
• Three maximal apneas after first breathing 100% O2 for
90 s and then 5% CO2 in O2 for 30 s (‘Hypercapnia’);
• Three fixed duration apneas after breathing 100% O2 at
a normal rate for 120 s (‘Normocapnia’);
• Three fixed duration apneas after first breathing 100% O2
at a normal rate for 90 s and then 30 s hyperventilation
on O2 (‘Hypocapnia’);
• Breathing of 100% O2 at a normal rate for 90 s, breathing
5% CO2 in O2 for 30 s and subsequently for a similar
period as the apneas in the other trials (‘Eupneic
Subjects were unaware of which gas was being inspired at
which time and during which trial.
Subjects reported to the laboratory fasted and without
caffeine for at least 2 hours prior to testing. Vital capacity
was measured via a spirometer (Compact II, Vitalograph,
Buckingham, England) and an intravenous catheter was
placed in the antecubital region using sterile technique.
Subjects lay prone for the duration of the trials, beginning
with a 20-minute period of prone horizontal rest. A nose
clip was placed prior to the first 2-minute countdown and
remained in place until 2 minutes after the final apnea.
Subjects were administered a normal-fitting mask for
breathing the gas mixtures with a flow rate of approximately
10 L min-1 during the 2-min countdown periods. At the end
of the countdown, the subject was instructed to exhale fully,
followed by a deep but not maximal inspiration and begin
the apnea. In previous studies, recordings of inspiratory
volume after this instruction have documented lung filling
to approximately 85% of vital capacity with low inter- and
intra-individual variance.16 Subjects were instructed to avoid
hyperventilation, with the exception of the final 30 s of the
countdowns in the ‘Hypocapnia’ trial. Upon completion
of the apnea, subjects expired fully into the mask and then
resumed normal breathing. In the ‘Hypercapnia’ trial, apneas
were conducted to maximum duration without time cues.
In the three time-limited trials, subjects terminated apneas
after a 5 s countdown.
Blood samples (2 ml) were taken via the intravenous catheter
2 min before the first apnea, immediately after the first and
third apneas and 10 min after the third apnea. Waste samples
of 1–2 ml preceded each blood sample and the catheter was
rinsed with 2 ml saline following each sample. The total
volume of blood (including waste volume) removed from
each subject was approximately 15 ml, and the injected saline
was approximately 12 ml. Blood samples were analysed
for Hb via an automated blood analysis unit (Micros 60
Analyzer, ABX Diagnostics, Montpellier, France).
From 2 min prior to each apnea until 2 min post-apnea, the
following parameters were measured continuously: arterial
haemoglobin saturation (SaO2) and heart rate (HR) via a
finger pulse oximeter (Biox 3700e, Ohmeda, Madison, WI,
USA), mean arterial pressure (MAP) via continuous finger
plethysmography (Finapres 2300, Ohmeda, Madison, WI,
USA), skin blood flow (SkBF) via laser-Doppler (Periflux
System 5000, Perimed, Järfälla, Sweden) on the thumb, and
breathing movements via a laboratory-developed pneumatic
chest bellows. Breath-by-breath CO2 was measured before
and after each apnea via a Normocap Oxy™ gas analyser,
(Datex-Ohmeda, Helsinki, Finland). Data were stored via a
BioPac MH100A CE multi-channel data acquisition system
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20126
(BioPac Systems Inc., Goleta, CA, USA). The continuous
monitoring of the cardiovascular parameters was done to
detect the diving response and for safety.
SaO2 values from the 30 s after each apnea were analysed to
determine if any desaturation occurred as a result of apnea,
and compared to both control and end-apneic SaO2 values.
Expired CO2 percentages from the first breath following each
apnea (and prior to gas mixture inhalation) were compared
between trials. Apneas were divided into an ‘easy’ phase
(prior to the onset of involuntary breathing movements) and
a ‘struggle phase’ (with involuntary breathing movements),
and durations compared between trials.17
Three subjects had triaxial measurement of spleen size
using ultrasonic imaging (Mindray DP-6600, Shenzhen
Mindray Bio-Medical Electronics Compan Ltd, Shenzhen,
China) simultaneously with all blood-sampling occasions,
for all four trials. Measurements of the maximal diameters
of spleen length (L), width (W) and thickness (T) were
used to calculate spleen volume according to the Pilström
Splenic volumes after the first and third apnea were
compared to the pre-apnea volume and with the 10-minute
post-apneic measurement. The ultrasonic imaging technique
was not available during the initiation of the study, and so
only three subjects were measured.
The Hb values obtained 2 min before the first apnea were
used as the control. Mean percentage changes from control
were used to compare changes within each trial, and pooled
mean changes from apneas were used to compare between
trials. Subjects served as their own controls, and effects
were expressed as percentage changes from control. All
variables were log-transformed before analysis to reduce
non-uniformity of error. Excel™ templates were used
for the calculations, purpose-designed for analyses using
physiological data.18 Comparisons were done using Student’s
t-tests with a level for acceptance of significant changes set
at P < 0.05. A Bonferroni correction was then applied for
multiple comparisons and significance was accepted at the
respective calculated level from the correction.
Results are reported as mean (SD) for point values, and
as mean (90% confidence intervals, CI) for comparisons.
One subject’s blood values were lost for the normocapnic
trial due to catheter failure, so this subject’s data were not
included in the blood analyses for this trial. The ‘missing’
subject was included in the analysis of the remaining trials
because the loss of one subject is compensated by a reduction
of the degrees of freedom in the calculations. As spleen
measurements were obtained from only three subjects, these
data are reported without statistical analysis.
DURATION OF APNEA
All subjects successfully repeated the following apnea
times (SD) in all trials: 216 (68) s for apnea 1, 222 (80) s
for apnea 2, and 245 (55) s for apnea 3. There was a trend
(P = 0.07) towards prolonged apneic duration from the
first to the third apnea. The ‘easy’ phase of the apneas was
shortest in the ‘Hypercapnic’ trial at 90 (27) s, followed
by the ‘Normocapnic’ trial at 103 (47) s and longest in the
‘Hypocapnic’ trial at 132 (43) s. The ‘easy’ phase was
significantly longer in the ‘Hypocapnic’ trial than in the
‘Hypercapnic’ trial (P = 0.012). There were no significant
differences in the ‘struggle phase’ duration: ‘Hypercapnic’
trial 134 (38) s; ‘Hypocapnic’ trial 118 (44) s and the
‘Normocapnic’ trial 112 (48) s.
ARTERIAL HAEMOGLOBIN SATURATION
Mean control values for SaO2 were above 98% for all trials,
and SaO2 did not change from control levels during any of
the apneas, or during post-apneic periods. There were no
differences between trials.
Baseline values of Hb before the apneas were the same for
all conditions. After the first apnea, Hb had increased in the
Changes in mean (SD) Hb from baseline after apnea 1 (A1),
after apnea 3 (A3), and 10 minutes following A3.
* P = 0.024 in ‘Hypercapnia’, P = 0.015 in ‘Normocapnia’
and ** P = 0.002 in ‘Hypercapnia’, P = 0.011 in
‘Normocapnia’; for comparisons with control, n = 7 for
‘Eupneic hypercapnic’ trial, n = 8 for all other trials
after A1after A3 after A3+10 min
Hb (% change from baseline)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20127
‘Hypercapnic’ (P = 0.024) and ‘Normocapnic’ (P = 0.015)
trials, while the ‘Hypocapnic’ and ‘Eupneic hypercapnic’
trials were similar to control values (Figure 1). After the
final apnea, Hb had further increased in the ‘Hypercapnic’
trial, to 4% above baseline (P = 0.002), and by 3 % in the
‘Normocapnic’ trial (P = 0.011), while the ‘Hypocapnic’ and
‘Eupneic hypercapnic’ trials showed no significant changes.
Ten minutes after apneas, Hb values were not different from
control values for any of the trials, nor were they different
among trials. A comparison of the magnitude of change from
baseline revealed no significant difference between trials.
The largest reduction in splenic volume after apnea 3 was
seen in the ‘Hypercapnia’ trial, at -33% from control,
followed by the ‘Normocapnia’ trial at -9% from control
(Figure 2). The ‘Hypocapnia’ and ‘Eupneic hypercapnia’
trials resulted in increases in spleen size of 30% and 13%
from control respectively. Ten minutes following the final
apnea, spleen volume tended to be restored in all trials.
Mean HR and SkBF, two main parameters of the
cardiovascular diving response, did not deviate from
control values or between trials. MAP was not different
between trials, but increased from control in all apnea
trials: by 35% in the ‘Hypercapnia’ trial (P = 0.0012), 15%
in the ‘Normocapnia trial’ (P = 0.06) and by 23% in the
‘Hypocapnia’ trial (P = 0.034). MAP remained unchanged
during the ‘Eupneic hypercapnia’ trial.
END-TIDAL CARBON DIOXIDE
Post-apneic expired CO2 was greatest in the ‘Hypercapnia’
trial at 7.6 (1.3)%, followed by the ‘Normocapnia’ trial at
7.4 (1.8)%, and the ‘Hypocapnia’ trial at 7.0 (1.5)%. In the
‘Eupneic hypercapnia’ trial, the expired CO2 level following
the breathing period equivalent to the apneic duration was
4.6 (1.0)%. Expired CO2 in the ‘Hypercapnia’ trial was
higher than in the ‘Hypocapnia’ trial (P = 0.029), and CO2
in the ‘Eupneic hypercapnia’ trial was lower than in all other
trials (‘Hypercapnia’ P = 0.001; ‘Normocapnia’ P = 0.001;
‘Hypocapnia’ P = 0.001).
In the absence of hypoxia (SaO2 ≥ 98% in all trials),
temporary increases in Hb across a series of apneas were
greatest in trials with an increased hypercapnic stimulus,
suggesting a role for hypercapnia in the elicitation of splenic
contraction. The three subjects studied with ultrasound also
demonstrated a greater degree of splenic contraction with
increased hypercapnic stimulus. This could explain why
apnea causes more splenic contraction than that seen with
eupneic hypoxia despite similar resulting levels of SaO2.15
A role of the apnea stimulus per se is supported by the
lack of response in the ‘Eupneic hypercapnia’ trial. A
greater stepwise influence of CO2 was also apparent in the
relative division of the ‘easy’ and ‘struggle’ phases during
apnea, where the ‘Hypercapnia’ trial had the shortest ‘easy’
phase and the ‘Hypocapnia’ trial had the longest, further
confirming a ‘pre-loading’ effect of CO2. Expired post-
apnea CO2 concentrations also indicated a similar, residual
stepwise pattern of systemic CO2 concentration. The lack
of change in SaO2 levels throughout the trials demonstrates
that hypercapnia acts as an independent stimulus for splenic
contraction during apnea.
The study cannot elucidate the neural or hormonal
mechanisms underlying this response. However, the
impact of inspired gas concentration just prior to apnea on
splenic contraction is likely to be mediated via both central
medullary and peripheral carotid body chemoreceptors for
CO2 and O2 respectively since changes in alveolar CO2 are
reflected in brain extracellular fluid pH over a time course
consistent with circulation time, i.e., a few seconds.19
Nevertheless, there are some ‘crossover’ effects whereby
peripheral receptors respond to CO2, and hypoxia affects
central chemoreception via alterations in cerebral blood
flow.20,21 In most individuals, hyperoxia (PO2 = 150 mmHg)
effectively silences the peripheral response to CO2.22,23
Therefore, the likely prevention of significant peripheral
chemoreceptor input, because of the sustained normoxia
in our protocol, makes it likely that the chemoreceptive
stimulus created by CO2 alone is sufficient to elicit a stimulus
leading to splenic contraction during apnea.
Changes in splenic volume from baseline after apnea 1, after
apnea 3 and 10 minutes following apnea 3; mean (SD) values
from three subjects
Controlafter A1 after A3after A3+10
(% change from baseline)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20128
Although the mechanisms leading to splenic contraction
are, as yet, only partially defined, they almost certainly
involve sympathetic innervation. The splenic nerve is mainly
adrenergic in composition, and has been shown to respond
powerfully to sympathetic discharge and related adrenergic
output.24–26 Hoka and associates also noted marked changes
in blood volume following hypercapnia in spleen-intact
dogs, whereas this response was considerably decreased in
splenectomised dogs.27 Similar sympathetic activity on the
part of the splenic nerve in humans is likely. Bradycardia,
a main component of the cardiovascular diving response,
was not significant in any trial nor different between trials,
suggesting that variations in CO2 levels do not affect this
response. This also illustrates the independent elicitation
of the splenic response, in accordance with previous
Both hypoxia and hypercapnia develop upon cessation of
breathing, and splenic contraction-induced blood boosting
may counteract, to some degree, these effects. Breath-hold
divers would likely benefit from a strong splenic contraction,
as the increase in circulating Hb would result in increased
oxygen storage capacity, increased CO2 buffering capacity
and a speedier recovery from hypoxia between apneas,
especially when these haematological effects remain across
several minutes between dives, whereas the cardiovascular
diving response reverses rapidly.28 Based on the observations
in this study, an increased capnic stimulus during apnea may
elicit a stronger or earlier spleen response and subsequent
Hb increase than apnea preceded by hyperventilation.
A direction for further research could be to focus on
whether there is a true dose-response relationship between
arterial CO2 content and the splenic contraction response, as
appears possible from this study. It would also be of interest
to compare the individual splenic responses to elevated
CO2 concentration of competition divers who employ
hyperventilation during ‘warm-up’ and divers without
‘warm-up’ practices before competition.29
The enhanced spleen-induced increase in Hb during
normoxic hypercapnia suggests a role of hypercapnia as a
trigger for splenic contraction during apnea. A separate role
of the apnea stimulus is suggested by the lack of response
in the ‘Eupneic hypercapnia’ trial.
We thank our subjects for participating in these studies,
and the temporary co-worker Robert de Bruijn for valuable
help during experiments. The study complies with Swedish
laws and ethical standards and was financed by the Swedish
National Centre for Research in Sports (CIF), and Mid
1 Gooden BA. Mechanisms of the human diving response. Integr
Physiol Behav Sci. 1994;29;6-16.
Andersson J, Schagatay E. Arterial oxygen desaturation during
apneas in humans. Undersea Hyperb Med. 1998;25:21-5.
Schagatay E, Andersson J. Diving response and apneic time
in humans. Undersea Hyperb Med. 1998;25:13-9.
Schuitema K, Holm B. The role of different facial areas in
eliciting human diving bradycardia. Acta Physiol Scand.
Lin YC, Shida KK, Hong SK. Effects of hypercapnia, hypoxia,
and rebreathing on heart rate response during apnea. J Appl
Hurford WE, Hong SK, Park YS, Ahn DW, Shiraki K, Mohri
M, et al. Splenic contraction during breath-hold diving in the
Korean Ama. J Appl Physiol. 1990;69:932-6.
Schagatay E, Andersson JPA, Hallén M, Pålsson B. Selected
contribution: role of spleen emptying in prolonging apneas in
humans. J Appl Physiol. 2000;90:1623-9.
Espersen K, Frandsen H, Lorentzen T, Kanstrup I-L,
Christensen NJ. The human spleen as an erythrocyte reservoir
in diving-related interventions. J Appl Physiol. 2002;92:2071-
Bakovic D, Eterovic D, Saratlija-Novakovic Z, Palada I, Valic
Z, Bilopavlovic N, et al. Effect of human splenic contraction on
variation in circulating blood cell counts. Clin Exp Pharmacol
10 Schagatay E, Haughey H, Reimers J. Speed of spleen
volume changes evoked by serial apneas. Eur J Appl Physiol.
11 Richardson MX, Lodin A, Reimers J, Schagatay E. Short-term
effects of normobaric hypoxia on the human spleen. Eur J
Appl Physiol. 2007;104:395-9.
12 Schagatay E, Andersson J, Nielsen B. Hematological response
and diving response during apnea and apnea with face
immersion. Eur J Appl Physiol. 2007;101:125-32.
13 Andersson JP, Liner MH, Runow E, Schagatay EK. Diving
response and arterial oxygen saturation during apnea and
exercise in breath-hold divers. J Appl Physiol. 2002;93:882-
14 Richardson MX, de Bruijn R, Schagatay E. Hypoxia augments
apnea-induced increase in haemoglobin concentration and
hematocrit. Eur J Appl Physiol. 2009;105:63-8.
15 Lodin-Sundström A, Schagatay E. Spleen contraction during
20 min normobaric hypoxia and 2 min apnea in humans. Aviat
Space Environ Med. 2010;81:545-9.
16 Schagatay E. The human diving response: effects of
temperature and training (dissertation). Lund, Sweden: Lund
17 Moore TO, Hong SK. Physiological and conventional breath-
hold breaking points. J Appl Physiol. 1974;37:291-6.
18 Hopkins W. A new view of statistics. Available from: Internet
Society for Sport Science http://www.sportsci.org/resource/
19 Ahmad HR, Loeschcke HH. Fast bicarbonate-chloride
exchange between plasma and brain extracellular fluid at
maintained PCO2. Pflugers Arch. 1982;395:300-5.
20 Poulin MJ, Liang PJ, Robbins PA. Dynamics of the cerebral
blood flow response to step changes in end-tidal PCO2 and
PO2 in humans. J Appl Physiol. 1996;81:1084-95.
21 Vovk A, Cunningham DA, Kowalchuk JM, Paterson DH,
Duffin J. Cerebral blood flow responses to changes in oxygen
Diving and Hyperbaric Medicine Volume 42 No. 1 March 20129
and carbon dioxide in humans. Can J Physiol Pharmacol.
22 Lloyd BB, Jukes MG, Cunningham DJ. The relation between
alveolar oxygen pressure and the respiratory response to
carbon dioxide in man. Q J Exp Physiol Cogn Med Sci.
23 Mohan R, Duffin J. The effect of hypoxia on the ventilatory
response to carbon dioxide in man. Respir Physiol.
24 Ayers AB, Davies BN, Withrington PG. Responses of
the isolated, perfused human spleen to sympathetic
nerve stimulation, catecholamines and polypeptides. Br J
25 Davies BN, Powis DA, Withrington PG. The differential effect
of cooling on the responses of splenic capsular and vascular
smooth muscle to nerve stimulation and noradrenaline.
Pflugers Archiv. 1978;377:87-94.
26 Herman NL, Kostreva DR, Kampine JP. Splenic afferents and
some of their reflex responses. Am J Physiol. Regul Integr
Comp Physiol. 1982;242:R247-54.
27 Hoka S, Arimura H, Bosnjak ZJ, Kampine JP. Regional
venous outflow, blood volume, and sympathetic nerve activity
during hypercapnia and hypoxic hypercapnia. Can J Physiol
28 Schagatay E, Andersson JP, Nielsen B. Hematologial response
and diving response during apnea and apnea with face
immersion. Eur J Appl Physiol. 2007;101:125-32.
29 Schagatay E. Predicting performance in competitive apnea
diving. Part II: dynamic apnea. Diving Hyperb Med.
Submitted: 08 April 2011
Accepted: 18 October 2011
Matt X Richardson, PhD, Harald K Engan, MSc, Angelica
Lodin-Sundström, MSc, Erika Schagatay, PhD,
Environmental Physiology Group, Department of Engineering
and Sustainable Development, Mid Sweden University,
Professor Schagatay also works at the National Winter
Sports Research Centre, Östersund, Sweden.
Address for correspondence:
Harald K Engan
Environmental Physiology Group, Mid Sweden University
Department of Engineering and Sustainable Development
Akademigatan 1, SE 831 25
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 10
The effect of intravenous perfluorocarbon emulsions on whole-body
oxygenation after severe decompression sickness
Cameron R Smith, J Travis Parsons, Jiepei Zhu and Bruce D Spiess
Breathing compressed air increases the amount of nitrogen
(N2) dissolved in body fluids.1–3 Factors such as ambient
pressure and time at depth are the primary determinants
of the amount of N2 absorbed.1–4 As ambient pressure
decreases, dissolved gas tensions in tissue can exceed
ambient pressure. This supersaturated state may lead to the
formation and growth of gas bubbles, resulting in venous
gas emboli (VGE) and possible arterial gas emboli (AGE).4,5
It is believed that these bubbles within the vasculature
and tissues are the root cause of decompression sickness
(DCS).4,5 There are likely multiple pathophysiological
mechanisms at play in DCS, including impairment of
microcirculation by inert gas bubbles, increased blood
viscosity, endothelial damage and complement activation.6–10
The physicochemical discontinuity of the gas-blood interface
can also denature proteins promoting the release of fatty
acids from cell membranes leading to the formation of fat
emboli.4,5 When bubbles obstruct capillaries or venules,
ischaemia ensues followed by reperfusion-induced oxidative
Perfluorocarbon emulsions (PFCs) are emulsions of
fluorinated hydrocarbons within phospholipid micro-
particle micelles.12 PFCs have been developed in medicine
as intravenous oxygen (O2) therapeutics.12 However,
compared to how whole blood carries the majority of its O2,
the transport of O2 by PFCs is fundamentally different. O2
carried by PFCs is not bound, as with haemoglobin, rather it
is dissolved in the PFC. Pure perfluorocarbons can dissolve
up to 600 ml L-1 O2,
0.031 ml L-1 and whole blood at 150 gm L-1 haemoglobin can
contain up to 210 ml L-1 O2.12 The O2 dissolved in PFCs is
all available to tissue, whereas that bound by haemoglobin is
restricted (arterial pO2 would need to drop below 40 mmHg
for greater than 25% of bound O2 to be released).14
13 whereas plasma can only dissolve
Microcirculatory changes such as oedema, vasospasm,
white cell activation and vessel plugging result in decreased
erythrocyte delivery of O2 to watershed tissue beds, yet
plasma flow may continue without red cells.15 PFCs, due
to their extremely small particle size (~0.1–0.4 µm), can
be delivered in this trickle-flow of plasma.12,16,17 Plasma
O2 delivery by PFCs is enough to keep tissue alive, as seen
with Fluosol DA-20%, a PFC which reduced myocardial
infarction and garnered FDA approval.18,19
PFCs are also effective in treating DCS, AGE and VGE.20–26
Using a swine saturation dive model with direct ascent to the
surface, it was found that administration of intravenous (IV)
PFCs and 100% O2 post-decompression decreased mortality,
the incidence of DCS and the number of neurological
events compared to animals administered just 100% O2 or
room air.21 Also, PFC and 1 hour of 100% O2 given at the
onset of DCS significantly decreased mortality observed
24 hours post-dive compared to animals treated with saline
and 100% O2 in a swine model of rapid decompression.27
Smith CR, Parsons JT, Zhu J, Spiess BD. The effect of intravenous perfluorocarbon emulsions on whole-body oxygenation
after severe decompression sickness. Diving Hyperb Med. 2012;42(1):10-17.)
Introduction: Decompression sickness (DCS) results from a decrease in ambient pressure leading to supersaturation of
tissues with inert gas and bubble formation. Perfluorocarbons (PFCs) are able to dissolve vast amounts of non-polar gases.
Intravenous (IV) PFC emulsions reduce both morbidity and mortality associated with DCS, but the mechanism of this
protective effect has not yet been demonstrated.
Methods: Juvenile Dorper-cross sheep (n = 31) were anaesthetised and instrumented for physiological monitoring, IV fluid
administration and blood sampling. Animals were compressed in air in a hyperbaric chamber to 608 kPa for 30 minutes and
then rapidly decompressed. Upon decompression, animals were randomly assigned to receive 6 mL kg-1 of PFC or saline
over 10 minutes beginning immediately after chamber exit. Arterial and mixed venous bloods were drawn at 5, 10, 15,
30, 60 and 90 minutes to examine pH, partial pressures of oxygen and carbon dioxide, oxygen saturation and electrolytes.
Results: Compared to saline, PFC administration increased arterial oxygen content (16.33 ± 0.28 vs. 14.68 ± 0.26 ml dL-1,
P < 0.0001), mixed venous oxygen content (12.56 ± 0.28 vs. 11.62 ± 0.26 ml dL-1, P = 0.0167), oxygen delivery (14.83 ±
0.28 vs. 13.39 ± 0.26 ml min-1 kg-1, P = 0.0003) and tissue oxygen consumption (3.30 ± 0.15 vs. 2.78 ± 0.13 ml min-1 kg-1,
P = 0.0149) but did not increase the extraction ratio (0.22 ± 0.012 vs. 0.21 ± 0.011, P = 0.5343).
Conclusions: It is likely that the improved oxygenation explains, at least in part, the previously-observed therapeutic effects
of PFCs in DCS.
Blood substitutes, perfluorocarbons, decompression sickness, treatment, oxygen, oxygen consumption
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201211
Similarly it was found that IV PFCs improve outcomes after
massive VGE, cerebral AGE, and coronary AGE.23,24,28 IV
PFCs have also been shown to increase N2 washout after
VGE.26 PFC administration is of benefit in the treatment of
decompression illnesses, but the mechanism of this benefit
has not been elucidated. Is it the PFCs’ ability to increase
N2 washout and remove bubbles obstructing circulation,
a product of improving O2 supply and metabolic state
of tissue, or some combination of these? The research
described here was designed to investigate the effect of IV
PFCs administered acutely after surfacing on whole-body
oxygenation in an ovine model of severe DCS.
Materials and methods
All experiments were performed in accordance with the
National Institutes of Health Guide for the care and use of
laboratory animals, and were approved by the Department
of Defense and the Virginia Commonwealth University
Institutional Animal Care and Use Committees. Juvenile
Dorper-cross sheep of either sex (Robinson Services, Inc.,
Mocksville, NC) weighing 18.5 ± 2.6 kg were housed in
United States Department of Agriculture and Association for
Assessment and Accreditation of Laboratory Animal Care
International approved facilities in social flocks with free
access to food and water on a 12-hour light/dark cycle. Sheep
were allowed a minimum of three days for acclimatisation
and veterinary inspection prior to use in any experiment.
PREPARATION AND INSTRUMENTATION
Prior to the experiment, sheep were muzzled for a period of
48 hours in order to prevent access to food but to provide
free access to water while remaining with their flock to limit
animal stress. Sheep were sedated with ketamine/xylazine
(20.0/2.0 mg kg-1 IM) and placed supine on the surgical table.
Animals were intubated with a 9.0 mm internal diameter
cuffed endotracheal tube (Hudson RCI, Temecula, CA) and
ventilated with 50/50 N2/O2 using a Siemens 900C ventilator
(Siemens Corp., New York, NY) with a tidal volume of
approximately 10 ml kg-1 and a rate of approximately 15
breaths per minute adjusted to maintain arterial pCO2 at 40
± 5 mmHg. An orogastric tube fashioned from TYGON®
R-3603 tubing (Satin-Gobain Performance Plastics Corp.,
Akron, OH) was advanced into the rumen to allow for fluid
drainage and to allow gas accumulated in the gut during
the air dive to vent upon decompression. A MAC® 2-port
introducer sheath (Arrow International Inc., Reading, PA)
was placed in the right external jugular vein to allow for the
administration of fluids and anaesthetic cocktail.
Once IV access was secured, administration of ‘triple
drip’ anaesthetic cocktail (ketamine/xylazine/guaifenesin
2.0/0.1/50.0 mg ml-1 in 5% dextrose) was begun immediately
at 1.0–2.0 ml kg-1 hr-1 titrated to maintain a surgical plane
of anaesthesia using a Harvard Apparatus PHD 2000
syringe pump (Harvard Apparatus, Holliston, MA). The left
femoral artery was cannulated with an 18-gauge femoral
arterial catheter (Arrow International Inc., Reading, PA)
for monitoring of arterial pressure (AP) and arterial blood
sampling. The right femoral vein was cannulated with a
4-French double-lumen catheter (Arrow International Inc.,
Reading, PA) for the anaesthetic administration while in the
hyperbaric chamber and for study drug administration after
exiting the chamber. The left femoral vein was cannulated
for the placement of a 7.5 Fr CCOmbo® continuous
cardiac output (CCO) pulmonary artery catheter (Edwards
Lifesciences, Irvine, CA) to allow for CCO, central venous
(CVP) and pulmonary arterial pressure (PAP) monitoring
and mixed venous blood sampling. Respiratory gases were
monitored continuously using an MGA 1100 respiratory
mass spectrometer (Perkin-Elmer, Norwalk, CT).
Following surgical manipulations, all animals were allowed
to stabilise for 30 minutes. After stabilisation, animals were
weaned off the ventilator until capable of spontaneously
breathing prior to being placed inside the hyperbaric
chamber. Normal saline was administered intravenously
at a rate of approximately 1 ml min-1 throughout the
surgical procedure in order to ensure proper hydration of
SHEEP DRY-DIVE PROCEDURES
Once weaned from the ventilator, monitoring equipment was
disconnected and sheep (n = 31) were placed into a Reimers
Systems model #17-48-100 Research Hyperbaric Chamber
(Reimers Systems, Inc., Springfield, VA). During the dry
dive all animals breathed room air and general anaesthesia
was maintained using a continuous infusion of ‘triple drip’ as
described above. Sheep were subjected to the following dive
profile. The chamber was pressurised at a rate of 101.3 kPa
min-1 to a pressure of 203 kPa. From 203 kPa the chamber
was pressurised at a rate of 203 kPa min-1 to a pressure of
608 kPa. The pressure of 608 kPa was maintained for 27
minutes, after which sheep were immediately decompressed
to ambient pressure at a rate of 203 kPa min-1.
Upon complete decompression (time = 0) all animals
were quickly removed from the hyperbaric chamber
and monitoring equipment was reconnected, as was the
ventilator with settings and breathing gas unchanged from
pre-compression/decompression settings. Animals were
randomised using a computer-generated block randomisation
sequence such that for each eight animals, four were assigned
to receive IV infusion of 6.0 ml kg-1 PFC (n = 15, 60% w/v
tert-butyl perfluorocyclohexane) and four were assigned
to receive saline control (n = 16) as an infusion over 10
minutes. All animals were monitored for 90 minutes after
decompression, during which time both arterial and mixed
venous blood samples were drawn and analysed using a
Radiometer OSM 3 Hemoximeter and a Radiometer ABL
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201212
700 blood gas analyser (Radiometer America, Westlake, OH)
at 5, 10, 15, 30, 60 and 90 minutes after decompression. Data
from all instruments were recorded directly to hard drive
using the BioPac system with Acqknowledge 3.90 software
(BioPac Systems Inc., Goleta, CA). After 90 minutes all
animals were euthanased.
Later offline analyses were performed to determine arterial
and mixed venous blood oxygen content (CaO2, ),
oxygen delivery ( ), tissue oxygen consumption ( ),
and oxygen extraction ratio (ER). The formulae used for the
calculations are listed in Table 1.
Unless otherwise stated, all data were analysed using
repeated measures analysis of variance (ANOVA) with
cardiac index and PFC administration included as model
effects. If the ANOVA was found to be significant, post
hoc least squares means Student’s t-tests were applied to
determine if the PFC treatment and saline control groups
were significantly different. Data are presented as least
squares (LS) means ± SEM. Differences were considered
statistically significant with P values of less than 0.05. All
statistical calculations were performed using the JMP 8 from
SAS Institute (Cary, NC).
Prior to diving the sheep, all physiological parameters under
investigation were compared to ensure that differences
between PFC and saline-treated sheep observed post-dive
were not the result of pre-dive surgical manipulations. One-
way ANOVA performed on baseline data obtained during
the stabilisation period post-surgery and pre-dive indicated
that there were no significant differences between the PFC-
treated group and the saline controls on any of the variables
of interest (PFC vs. saline – CaO2, , , , ER and
cardiac index (indexed to body weight, CI)).
Since previous studies have indicated that split-hoofed
species can develop pulmonary hypertension severe enough
to interfere with CI after PFC administration, we examined
the effect of PFC administration on CI in this model.29 Figure
1A shows CI changes in saline- and PFC-treated sheep
during the 90-minute period post-dive (repeated measures
ANOVA, P < 0.0001). In PFC-treated animals, CI was lower
compared to saline and trended towards increasing over
time while remaining stable in saline-treated sheep. When
LS means were compared, CI was found to be significantly
lower by 19.4% in the PFC-treated group vs the saline
control group (Figure 1B). Because of the significant effect
PFC administration had on CI, CI was included as a model
effect in all further analyses.
Figure 2A illustrates the changes in CaO2 over the time course
of the experiment following the return to surface (repeated
measures ANOVA, P < 0.0001). CaO2 increased in both PFC-
and saline-treated animals. Likewise, CaO2 is higher in the
PFC-treated group vs. the saline control. Figure 2B shows
the results of the LS means post-hoc comparison. CaO2 was
significantly increased over saline control by 10.5%.
The effect of PFC treatment on over time post-
chamber was also investigated and found to be significant
as described by repeated measures ANOVA (Figure 3,
P = 0.0159). Both PFC- and saline-treated sheep displayed
a non-significant trend towards increasing over time. The
Arterial O2 content: CaO2 = (1.34 x Hb x SaO2) + [(0.0031 x PaO2 x a) + (0.01997 x PaO2 x ß)] (1)
Mixed venous O2 content:
Where Hb = haemoglobin concentration in mg dL-1; SaO2 = arterial O2 saturation fraction; PaO2 = arterial O2 tension in mmHg;
CO = cardiac output in L min-1; 0.0031 = O2 solubility coefficient in plasma in ml dL-1; 0.01997 = O2 solubility coefficient in 60% w/v
tert-butyl perfluorocyclohexane emulsion in ml dL-1; a = blood fraction of circulation volume; ß = PFC fraction of circulating volume
and 1.34 = O2-haemoglobin binding coefficient in ml g-1.
CO C O
Equations used to determine arterial and mixed venous blood oxygen (O2) content (ml dl-1), O2 delivery (L min-1 kg body
weight-1), tissue O2 consumption (L min-1 kg body weight-1), and oxygen extraction ratio
C OHb S O
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201213
The effect of perfluorocarbon administration on arterial oxygen content (CaO2)
A: Arterial oxygen content vs time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B: Least squares means of saline- and PFC-treated groups; PFC significantly increased CaO2 vs saline control (* P < 0.0001)
The effect of perfluorocarbon administration on mixed venous oxygen content
A: Mixed venous oxygen content vs time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B Least squares means of saline- and PFC-treated groups; PFC significantly increased
C O vs saline control (* P = 0.0159)
The effect of perfluorocarbon administration on cardiac index (CI)
A: Cardiac index plotted against time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B: Least squares means of saline- and PFC-treated groups; PFC significantly decreased CI vs saline control (* P < 0.0001)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 14
The effect of perfluorocarbon administration on oxygen delivery (
A: Oxygen delivery vs time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B: Least squares means of the saline- and PFC-treated groups; PFC significantly increased
DO vs saline control (* P = 0.0002)
The effect of perfluorocarbon administration on oxygen consumption (
A: Oxygen consumption vs time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B: Least squares means of saline- and PFC-treated groups; PFC significantly increased
VO vs the saline control (* P = 0.0122)
The effect of perfluorocarbon administration on extraction ratio (ER)
A: Extraction ratio vs time; solid line represents PFC, dashed line represents saline; chamber exit at time = 0
B: Least squares means of saline- and PFC-treated groups; PFC had no significant effect on ER vs saline control (P = 0.5190).
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 15
results of the LS means comparison are shown in Figure
3B. was found to be significantly higher in PFC
animals vs saline control by 6.7%.
Figures 4A and 5A demonstrate the changes in and
respectively, following decompression (repeated
measures ANOVA, P < 0.0001 for both). Both and
are higher in the PFC sheep compared to the saline-
treated animals. Also, and remain stable in animals
treated with PFC over the 90 min period while appearing to
decrease in sheep administered saline. Figures 4B and 5B
show the results of the LS means comparisons of and
respectively. It can be seen that is 10.3 % higher
and is elevated some 22.1% over saline controls.
Additionally, the effect of PFC treatment on ER was
investigated (Figure 6A, repeated measures ANOVA,
P < 0.0001). The data reveal that ER for PFC-treated sheep
was not different to animals given saline during the 90
min observation period. However, ER for both PFC- and
saline-treated sheep trended toward increasing throughout
the post-chamber examination window. Figure 6B shows the
results of the LS means comparison. ER was not significantly
increased in the PFC-treated group vs the saline control.
Finally, in order to present a more complete picture of
the animals’ condition following decompression, several
haemodynamic parameters were analysed (Table 2). Arterial
pressure (systolic, diastolic, mean), pulmonary arterial
pressure, and arterial PCO2 in both PFC- and saline-treated
sheep were all found to be decreasing over time post-chamber
(repeated measures ANOVA, P < 0.0001 for all except PAP,
P = 0.0007). LS means comparison showed that all variables
were significantly higher in PFC- vs saline-treated animals.
Central venous pressure and heart rate in both PFC and
saline sheep were stable over time following decompression
(repeated measures ANOVA: CVP P = 0.2696; HR
P = 0.2371 HR). LS means comparison revealed that both
variables were significantly lower in PFC- vs saline-treated
animals. Arterial pH showed a non-significant trend towards
increasing in both PFC and saline sheep over time after the
dry dive (P = 0.0554), and was significantly lower in PFC-
vs saline-treated animals when LS means were compared.
Arterial pO2 was found to be increasing over time in both
groups (P = 0.0188), but LS means revealed no significant
difference between PFC- and saline-treated animals.
Taken together, the data presented show that sheep subjected
to decompression stress and treated with PFC immediately
following return to surface displayed significantly greater
arterial oxygen content, oxygen delivery, and oxygen
consumption compared to animals exposed to decompression
stress and given saline.
Further analysis using the repeated measures ANOVA
model showed both PaO2 and SaO2 were significantly lower
in the PFC-treated vs saline-treated group (200.03 ± 10.77
vs 238.72 ± 9.90 mmHg, P = 0.0109 and 93.52 ± 0.89 vs.
97.39 ± 0.81 %, P = 0.0021 respectively). Haemoglobin (Hb)
was elevated in the PFC-treated group vs saline control after
compression/decompression (12.22 ± 0.17 vs. 10.72 ± 0.15
mg dL-1, P < 0.0001) but not at baseline (one-way ANOVA
11.19 ± 0.38 vs 10.61 ± 0.37 mg dL-1, P = 0.2931).
As has been seen before in other split-hoofed animal
models, IV administration of PFCs resulted in decreased
CI.29 Previous work conducted in pigs reported pulmonary
hypertension to be the cause of the observed decrease in CI
and similar observations were made here.29 When analysed
using repeated measures ANOVA, pulmonary arterial
pressure was found to be nearly doubled in the PFC-treated
animals vs the saline control (see Table 2). This suggests that
the problem of pulmonary hypertension leading to decreased
CI will likely be present in all split-hoofed species.
It is clear from this study that IV PFC administration results
in increased CaO2. CaO2 was elevated nearly 11% over
control with PFC. Even if the oxygen carried directly by
the PFC is removed from the calculations, CaO2 was still
significantly higher in the PFC-treated group (P = 0.0019).
PFC appears to do more than simply carry more O2, but
exactly what PFC does in addition to its own O2-carrying
ability is unclear. It is possible that the presence of the PFC is
inducing the release of erythrocytes from the spleen or other
storage, accounting for the higher Hb, and contributing to
the higher CaO2 in the PFC-treated group. It is also possible
that the presence of free gas bubbles in the microcirculation
results in some vessel injury, followed by inflammation and
leakage of plasma out of the intravascular space resulting in
an apparent haemoconcentration. These possibilities warrant
further investigation in order to elucidate their exact cause,
and could be tested by examining spun haematocrit values,
The effect of perfluorocarbon (PFC) administration on
haemodynamic parameters and arterial gas partial pressures
Arterial Pressure (mmHg)
Central Venous Pressure (mmHg)
Pulmonary Arterial Pressure (mmHg)
Heart Rate (bpm) 121.7 (2.80) 113.2 (2.97)
pH 7.43 (0.008)
PaO2 217.3 (20.08) 209.9 (21.31)
PaCO2 39.3 (0.65)
Least squares means (SEM) P value
93.6 (3.20) 105.3 (3.40)
12.1 (1.17) 7.9 (2.24) 0.0158
25.5 (1.18) <0.0001
7.39 (0.007) <0.0001
43.6 (0.69) <0.0001
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 16
plasma protein content and/or by conducting tagged RBC
The observation that PFC administration results in increases
in both and of 10% and 22%, respectively,
demonstrates that the PFC was able to not only increase the
amount of O2 present in the blood, but to improve tissue
access to that O2. This suggests that the mechanism whereby
IV PFC improves tissue oxygenation is not simply a result
of its ability to carry greater quantities of O2, but that it
facilitates O2 delivery to cells. This may take the form of the
PFC extravasating in capillary beds, taking dissolved oxygen
with it. Alternatively, the PFC emulsion particles, being
approximately 1/100th–1/1000th the size of an erythrocyte,
may be able to pass through blood vessels where red cell flow
has been blocked by bubbles, but a trickle flow of plasma
remains.21,25 In this case the small amount of O2 carried in the
PFC may be sufficient to keep viable tissues that otherwise
might succumb to hypoxic injury.
More interestingly, PFC particles may act as a bridge,
facilitating the movement of O2 from erythrocytes into
tissues. This possibility has very intriguing implications.
As shown above, the amount of O2 actually dissolved in
PFC is relatively small. Haemoglobin binding O2 remains
the dominant mechanism for O2 transport. Once in capillary
beds, the greatest impediment to the offloading of O2 from
haemoglobin is the plasma.30 O2 is very insoluble in plasma,
and much more soluble in PFC. Therefore, PFC could act
as a transport vessel for O2, ferrying it from erythrocytes to
tissues, a mechanism somewhat akin to facilitated diffusion
across cell membranes. These possible mechanisms should
be explored further in future studies.
These results demonstrate that improved tissue oxygenation
at a whole-body level is likely responsible for at least a portion
of the beneficial effects offered by the IV administration of
PFC emulsions after decompression sickness.
The authors thank Drs Kevin Ward, R Wayne Barbee, and
Penny S Reynolds for their insight and suggestions during
the experimental design and data analysis.
Conflict of interest
Travis Parsons is an investor owning 90 shares of stock in
Oxygen Biotherapeutics, Inc., less than 0.0001% of public
Bruce Spiess is an investor owning 10,000 shares of stock
in Oxygen Biotherapeutics Inc., less than 0.01% of public
1 Guyton AC, Hall JE. Physiology of deep-sea diving and other
hyperbaric conditions. In: Guyton AC, Hall AC, editors.
Textbook of medical physiology, 10th ed. Philadelphia, PA:
Saunders; 2000. p. 504-9.
Moon RE. Treatment of diving emergencies. Crit Care Clin.
Replogle WH, Sanders SD, Keeton JE, Phillips DM. Scuba
diving injuries. Am Fam Physician. 1988;37:135-42.
DeGorordo A, Vallejo-Manzur F, Chanin K, Varon J. Diving
emergencies. Resuscitation. 2003;59:171-80.
Dutka AJ, Francis TJ. Pathophysiology of decompression
sickness. In: Bove AA, Davis JC, editors. Bove and Davis’
diving medicine, 3rd ed. Philadelphia: Saunders; 1997. p.
Dutka AJ, Kochanek PM, Hallenbeck JM. Influence of
granulocytopenia on canine cerebral ischemia induced by air
embolism. Stroke. 1989;20:390-5.
Hallenbeck JM, Bove AA, Elliot DH. The bubble as a non-
mechanical trigger in decompression sickness. In: Ackles
KN, editor. Proceedings of a symposium on blood bubble
interactions. Report # 73-CP-960. Downsview, Ontario:
Defence and Civil Institute for Environmental Medicine;
Hallenbeck JM, Bove AA, Moquin RB, Elliott DH. Accelerated
coagulation of whole blood and cell-free plasma by bubbling
in vitro. Aerosp Med. 1973;44:712-4.
Hills BA, James PB. Microbubble damage to the blood-brain
barrier: Relevance to decompression sickness. Undersea
Biomed Res. 1991;18:111-6.
10 Vane JR, Anggard EE, Botting RM. Regulatory functions of
the vascular endothelium. N Engl J Med. 1990;323:27-36.
11 Elliot DH, Moon RE. Manifestations of decompression
disorders. In: Bennett PB, Elliott DH, editors. The physiology
and medicine of diving, 4th ed. Philadelphia: Saunders; 1993.
12 Stowell CP, Levin J, Spiess BD, Winslow RM. Progress in
the development of rbc substitutes. Transfusion. 2001;41:
13 O’Brien RN, Langlais AJ, Seufert WD. Diffusion coefficients
of respiratory gases in a perfluorocarbon liquid. Science.
14 Leow MK. Configuration of the hemoglobin oxygen
dissociation curve demystified: A basic mathematical proof for
medical and biological sciences undergraduates. Adv Physiol
15 Theilen H, Schrock H, Kuschinsky W. Gross persistence
of capillary plasma perfusion after middle cerebral artery
occlusion in the rat brain. J Cereb Blood Flow Metab.
16 Biro GP. Perfluorocarbon-based red blood cell substitutes.
Transfus Med Rev. 1993;7:84-95.
17 Riess JG. Perfluorocarbon-based oxygen delivery. Artif Cells
Blood Substit Immobil Biotechnol. 2006;34:567-80.
18 Kerins DM. Role of the perfluorocarbon fluosol-da in coronary
angioplasty. Am J Med Sci. 1994;307:218-21.
19 Kent KM, Cleman MW, Cowley MJ, Forman MB, Jaffe
CC, Kaplan M, et al. Reduction of myocardial ischemia
during percutaneous transluminal coronary angioplasty with
oxygenated fluosol. Am J Cardiol. 1990;66:279-84.
20 Dainer H, Nelson J, Brass K, Montcalm-Smith E, Mahon R.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201217
Short oxygen prebreathing and intravenous perfluorocarbon
emulsion reduces morbidity and mortality in a swine
saturation model of decompression sickness. J Appl Physiol.
21 Dromsky DM, Spiess BD, Fahlman A. Treatment of
decompression sickness in swine with intravenous
perfluorocarbon emulsion. Aviat Space Environ Med.
22 Spiess BD. Perfluorocarbon emulsions: one approach to
intravenous artificial respiratory gas transport. Int Anesthesiol
23 Spiess BD, Braverman B, Woronowicz AW, Ivankovich AD.
Protection from cerebral air emboli with perfluorocarbons in
rabbits. Stroke. 1986;17:1146-9.
24 Spiess BD, McCarthy R, Piotrowski D, Ivankovich AD.
Protection from venous air embolism with fluorocarbon
emulsion fc-43. J Surg Res. 1986;41:439-44.
25 Spiess BD, McCarthy RJ, Tuman KJ, Woronowicz AW, Tool
KA, Ivankovich AD. Treatment of decompression sickness
with a perfluorocarbon emulsion (fc-43). Undersea Biomed
26 Zhu J, Hullett JB, Somera L, Barbee RW, Ward KR, Berger
BE, et al. Intravenous perfluorocarbon emulsion increases
nitrogen washout after venous gas emboli in rabbits. Undersea
Hyperb Med. 2007;34:7-20.
27 Mahon RT, Watanabe TT, Wilson MC, Auker CR. Intravenous
perfluorocarbon after onset of decompression sickness
decreases mortality in 20-kg swine. Aviat Space Environ Med.
28 Spiess BD, McCarthy RJ, Tuman KJ, Ivankovich AD.
Protection from coronary air embolism by a perfluorocarbon
emulsion (fc-43). J Cardiothorac Anesth. 1987;1:210-5.
29 Spiess BD, Zhu J, Pierce B, Weis R, Berger BE, Reses J, et al.
Effects of perfluorocarbon infusion in an anesthetized swine
decompression model. J Surg Res. 2009;153:83-94.
30 Popel AS. A finite-element model of oxygen diffusion in the
pulmonary capillaries. J Appl Physiol. 1997;82:1717-8.
Submitted: 18 August 2011
Accepted: 12 January 2012
Cameron R Smith, PhD1,3,5, J Travis Parsons, PhD4,5, Jiepei
Zhu, MD, PhD1,5, and Bruce D Spiess, MD1,2,5
Departments of Anesthesiology1, Emergency Medicine2,
Physiology3, Neurosurgery4, and the Virginia Commonwealth
University Reanimation Engineering Shock Center
(VCURES)5, Virginia Commonwealth University Medical
Center, Richmond, Virginia, USA.
Address for correspondence:
Cameron R Smith, PhD
PO Box 980695
Virginia 23298-0695, USA
This work was supported by a grant from the United States
Office of Naval Research (ONR) (N000140210399)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 18
Diver Health Survey score and probability of decompression
sickness among occupational dive guides and instructors
Greg A van der Hulst and Peter Lee Buzzacott
(van der Hulst GA, Buzzacott PL. Diver Health Survey score and probability of decompression sickness among occupational
dive guides and instructors. Diving Hyperb Med. 2012;42(1):18-23.)
Introduction: This study attempted to correlate self-reported post-dive Diver Health Survey (DHS) scores with computed
daily probability of decompression sickness (pDCS) values as a measure of decompression stress in occupational divers
in the recreational diving industry.
Methods: Divers completed the DHS form and their dive profiles were recorded electronically. The pDCS for each dive
was calculated using the LE1 probabilistic model. Data were analysed using a mixed effects model.
Results: DHS score was not significantly associated with pDCS. Mean DHS score on non-diving days was 1.6 and increased
by 0.8 for each dive made during any day. Mean number of daily dives was 1.9 and mean DHS score on diving days was
Conclusion: Utility of the DHS for monitoring daily decompression stress among occupational divers working in the
recreational diving industry in New Zealand remains unproven.
Occupational diving, occupational health, health surveillance, diving at work, decompression sickness, models
Decompression schedules for diving have progressively
evolved from those developed by Haldane in the early
1900s, all with the common goal of avoiding decompression
sickness (DCS).1 DCS is a multisystem condition that can be
protean in its manifestations. Both clinicians treating divers
and researchers testing decompression procedures have
historically utilised a binary classification system – DCS vs
no-DCS. However, it is also accepted that the physiological
processes responsible for the clinical manifestations of
DCS are active to a greater or lesser degree after all but the
most trivial exposures to pressure. Where to draw the line
for diagnosis of DCS depends on a number of factors but,
irrespective of the exact definition used, DCS remains a rare
event. This very low incidence of clinical DCS presents a
challenge to researchers in that a prohibitively large number
of trials need to be conducted before a decompression model
can be statistically shown to be effective at preventing such
a rare event.
Weathersby et al. pointed out the advantages of applying
maximum likelihood techniques to binary outcomes from
diving decompressions and proposed fitting a risk model
to profiles of depth-time-breathing gas with known DCS
outcomes.2 For a given dive profile, such ‘trained’ models
can predict the probability of DCS (pDCS). How accurate the
prediction is depends to a large extent on how well the dive
being assessed matches the original data set.3 Use of binary
outcome data (DCS/no-DCS) can limit the complexity of
the models that can be fitted because of the low incidence
of DCS within most diving data sets.4 Statistically based
decompression models have been fitted to Doppler venous
bubble scores and to binary DCS/no-DCS results with the
inclusion of ‘marginal’ cases to increase model degrees
of freedom.5,6 Regardless, many dives must be monitored
to detect enough DCS cases to allow fitting of complex
THE DIVER HEALTH SURVEY
An alternative approach to detecting DCS in the field is to
utilise self-reported health status measured in the form of a
questionnaire. Doolette suggested this approach commenting
that, if diving health outcome could be reliably measured in
the field, results could be matched to electronic depth-time
profiles and could provide an alternative source of data for
decompression model calibration.4 The Diver Health Survey
(DHS) was subsequently developed to measure self-reported
diver health status following decompression. The DHS tool
consists of a single-sided A4 post-dive questionnaire with
nine explicit items covering five general concepts indicative
of health status, (physical functioning, role limitation,
general health perception, bodily pain, and vitality), six
common symptoms of DCS, (pain, paraesthesia, weakness,
vitality, rash, and balance/dizziness), and time of onset of
symptoms relative to diving activity. A response to each
of the nine explicit items is chosen from four check boxes
with semantic anchors representing scores of 0 through 3;
the lower the score, the more normal is the health status. The
DHS has been described in detail elsewhere.7 Psychometric
testing of this survey tool suggested that it was a statistically
valid measure of decompression-related health outcome
and that it also appeared sufficiently reliable for collection
of grouped data for decompression model calibration.7
Advantages of the DHS were that it removed the need to
diagnose DCS in the field (replacing binomial DCS/no-DCS
with 30-point interval data, significantly increasing model
degrees of freedom), it was brief (nine questions + one free
response) and it was self-administered.7
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201219
The DHS was used initially on tuna farm divers in South
Australia to review their diving practices and the impact of
multi-day diving on reported post-dive health status.8,9 It
has also been used to measure perceived post-decompression
health status in hyperbaric chamber attendants following
standard medical hyperbaric exposures, health status
following dry chamber dives on nitrox, on a cave diving
expedition and on a small group of technical divers.10–13
The work on tuna farm divers comprises the only published
data correlating occupational diver health scores with
computed probability of DCS. The DHS is described
as a valid instrument for field assessment of DCS with
significant correlation of DHS scores and concurrent
medical diagnosis.7 The aim of this study was to assess if
the DHS correlated with computed daily pDCS values as a
measure of decompression stress in occupational divers in
the recreational diving industry.
Thirty-one occupational divers working in Tutukaka, New
Zealand were invited to participate and 25 (81%) agreed.
Participants were supplied with an information sheet
describing the study’s aims, the data to be collected and
the ultimate destination of the data. Participants then gave
signed consent. The research protocol was approved by the
University of Auckland Human Research Ethics Committee.
Participants completed the DHS form both on diving and
non-diving days. None reported previous DCS. DHS scores
were calculated and stored in an Excel spreadsheet matched
to each diver’s individual identifier (ID). Also recorded were
the consecutive number of days each diver had participated
(DAY), total daily dive duration in minutes (DUR), daily
maximum depth reached in metres’ sea water (MSW) and
the number of dives per day (NUM). All dives were made
Depth-time dive profile data were recorded by Sensus Ultra
loggers (Reefnet inc, Mississauga, Canada) or personal dive
computers (Suunto Oy, Finland; ScubaPro Uwatec, USA;
and DeltaP Technology, UK). The Sensus Ultra loggers had
a pressure resolution to 1 mbar, with an accuracy of +/-30
mbar, equivalent to 30 cms change in depth whilst immersed
in sea water. Variation in depth resolution between personal
dive computers was not measured. Depth-time profiles
were downloaded from each depth-time recorder directly
to a laptop PC using each unit’s proprietary interface and
software. Data were exported from each manufacturer’s
proprietary software in comma-delimited ASCII format,
before being transferred into a purpose-built spreadsheet
via an import routine programmed in Visual Basic for
Applications (Microsoft Excel 2002, Microsoft Corp,
Redmond, WA, USA).
Repetitive dives (defined as a surface interval of less than
18 h) were combined into a single depth-time profile linked
with the DHS score from the end of that day. Dive profile
data were analysed by Dr David Doolette to compute pDCS
for each ‘diving day’ employing the LE1 probabilistic model
calibrated to military air diving using the methods described
by Thalmann and co-workers in 1997.6 The resultant column
of daily pDCS values completed the dataset.
Six of the 25 participants were lost to follow-up when they
left the area at the end of the summer diving season without
returning their data collection booklets or dive data recorder.
A seventh experienced a dive computer malfunction which
rendered its data unusable, leaving 18 participants for
Data were analysed using SAS (ver. 9.2, Cary, NC).
Strengths of association with the dependent variable DHS
were evaluated using a linear mixed effects model. Mixed
effects models are particularly suited to the analysis of
repeated measures data involving randomly selected subjects
exhibiting inter-subject variability.14,15 Variance components
and parameters were estimated using maximum likelihood.
The full model before later variable selection was:
HSij = ß0i + ei + ß1pDCSij + ß2DURij + ß3MSWij + ß4NUMij + ei
where ß0 = the intercept of the regression which is dependent
upon the diver (subscript i) and e = random error, which
was associated with the diver (subscript i) and the day on
which data were collected (subscript j). Homoscedasticity
for individual residual variance was tested for using a
likelihood ratio test. In search of the most parsimonious
model, independent variables were manually removed from
the full model one at a time and the increasingly simplified
models fitted to the data. Models were evaluated using
Akaike Information Criteria (AIC), which bypasses the need
to specify a level of significance a priori to model building
unlike backwards elimination; smaller AIC indicates better
fit.15 Differences in fit between models pre- and post-variable
removal follow a chi-square distribution and were tested
for significance (P < 0.01) using a likelihood ratio test with
degrees of freedom equal to the number of explanatory
Eleven of the 18 divers were male. Mean diving experience
was 11.5 years with a median of 1,200 lifetime dives.
Participant characteristics are presented in Table 1; subjects
Body mass index (kg m-2)
Diving experience (years)
Number of lifetime dives
New Zealand occupational dive guide and instructor
demographic characteristics (n = 18)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201220
were primarily young, fit, experienced divers.
The mean delay between surfacing from the last dive of each
day and completing the DHS was 6.0 hours (SD 1.3). As
shown in Table 2, the mean DHS overall during diving days
(n = 359) was 3.1 (SD 2.0). Mean DHS during non-diving
days (n = 395) was 1.6 (SD 1.7).
Divers’ individual residuals were sufficiently different to
reject the assumption of homoscedasticity, (chi-square =
24.9, df = 1, P < 0.01), therefore, the effect of repeated
measures (ID) was retained within each model tested.
Though these are not shown in Tables 3 or 4, the range of
intercepts for ID in model 1 of Table 3 was -2.5 to +3.1.
Removal of DUR did not significantly improve the full
model (P = 0.16) nor did the removal of pDCS (P = 0.16).
By model 3, the AIC was the lowest value of any model but
the parameter estimate of MSW was so small as to affect
DHS by a score of -1 for every increase of 50 msw maximum
depth. Model 3 was significantly worse for the removal of
either NUM (model 4, P < 0.01) or MSW (model 5, P <
0.01). In keeping with the aim of the study model, model
6 was also tested and found to be significantly worse than
model 3 (P < 0.01), as was the null model comprising only
the intercept and random error (model 7, P < 0.01).
Taking into account Table 3, the delay in minutes between
surfacing from the last dive of each day and completing
the DHS (SUR2DHS) was added to the model and the AIC
process repeated for data recorded during diving days only
(n = 359). The fitting of the model including SUR2DHS is
presented in Table 4.
Fitting all data (n = 754) in Table 3, the lowest AIC was
calculated for model 3, in which the size of the effect of
MSW was negligible, and where the addition of pDCS
did not result in a significantly improved fit (model 3 vs
2, P = 0.16). Likewise, for the diving data alone (n = 359)
the removal of pDCS from the model with the lowest AIC
(model 3) did not result in a significantly worse fit (model
5 vs 3, P = 0.17). The fit of model 3 was not significantly
worsened for the removal of SUR2DHS and pDCS (model 8,
AIC 1265 vs 1261, P = 0.15), but it was significantly worse
for the removal of NUM (model 6, AIC 1300 vs. 1261, P <
0.001), suggesting that, among occupational divers in the
recreational industry, DHS is most closely linked to the daily
number of dives. An intercept of 0.8 (model 3) suggests an
increase in DHS of 0.8 for each additional dive made during
any day, as can be seen in Figure 1.
Repetitive dives 266
Depth (msw) Duration (min) Daily dives Probability of DCS
93 2.4 (1.5)
Overall 359 3.1 (2.0) 20.5 (7.7) 77.3 (29.3) 1.9 (0.7) 0.013 (0.017)
Diving data, showing medians (range) for individual Diver Health Survey (DHS) scores, depths, dive durations,
numbers of daily dives and computed pDCS (LE1) values and means (SD) for grouped data;
msw – metres’ seawater depth; DCS – decompression sickness
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 21
This review of the diving practices of occupational
dive guides and instructors suggests they manage their
decompression risk conservatively. There were no reported
incidences of DCS among the study participants and
their DHS scores were typically within the asymptomatic
range. However, DHS scores did not correlate highly with
computed pDCS values.
As with the Doolette study of tuna divers, the random effect
of diver ID had a significant effect upon the model AIC.9
Given the generalised nature of the health status indicators
used in the DHS, the capture of some non-diving-related
symptoms is expected. While this reduces the specificity of
the survey at the level of the individual diver, it maintains
sensitivity for the non-specific, generalised symptoms of
DCS, which is needed when collecting group data. Internal
consistency testing of the DHS has previously demonstrated
the survey items measure aspects of the same attribute
(established by concurrent validity testing for symptoms of
DCS).7 In this study, the intercept for ID ranged from -2.5
to +3.1 (range 5.6), similar to the variance among tuna divers
of 0.1 to 4.7 (range 4.6).8
The mean pDCS recorded in this study during 359 diving
days was 0.013, which was higher than recorded during
383 occupational tuna diving days (pDCS = 0.005).9 Of the
359 diver-days in this study, 293 (82%) exceeded a pDCS
of 0.005. The LE1 model used to compute pDCS in this
study may not be a good predictor of DCS in occupational
dive guides and instructors. A mean pDCS of 0.013 over
359 diving days equates to 4.67 predicted incidents. There
were no reported cases of DCS and only two diving days
with DHS > 8, which has been associated previously with
clinical DCS.7 The dataset used to calibrate the LE1 model
contained only 8% repetitive air dives; whereas this study
recorded 266/359 (74%) repetitive air dives and this may
also have affected the pDCS. The LE1 model has previously
under-estimated pDCS for repetitive air dives.6
Parameter Likelihood ratio
chi square (df)
2593.2 -1257 2 vs 1 2 (1) 0.157
3 vs 1
3 vs 2
1 vs 4
3 vs 4
1 vs 5
3 vs 5
1 vs 6
3 vs 6
1 vs 7
pDCS – probability of decompression sickness; DUR – dive duration (minutes); MSW – maximum depth in metres of sea water;
NUM – number of daily dives; AIC – Akaike Information Criteria; LL – log likelihood; df – degrees of freedom
Model improvement through variable removal and fitting to all data (n = 754)
Diver Health Score by daily number of dives
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201222
The mean depth of non-repetitive dives of 22 msw and
mean dive time of 45 minutes approaches the no-stop limit
of the DCEIM tables, which has a pDCS ≥ 0.0156.9 One
daily dive schedule did exceed that no-stop limit (pDCS
= 0.159) resulting in an unremarkable health outcome
(DHS score 1). Overall, this study found a mean depth of
20 msw, mean total daily duration underwater of 77 min,
spread over 1.9 dives per day (Table 2). This contrasts with
occupational tuna divers who recorded a mean depth of 17
msw, a mean dive time of 23 min and a mean of 1.4 dives
per day.8 Though the divers in this study recorded greater
mean depth, total bottom time and daily number of dives
than occupational tuna divers, these parameters may not
adequately portray overall decompression stress because of
potential differences in dive profiles, for example multi-level
vs square-wave. That the DHS was insensitive among New
Zealand recreational dive guides and instructors, yet useful
as a measure of decompression stress among Australian tuna
farm divers, may be (at least in part) due to these differences
in diving profiles. Caution is, therefore, advised before
generalising these findings to other occupational recreational
It is also possible these results may have been influenced
by a degree of response bias. The South Australian tuna
farm divers studied by Doolette were predominantly
company employees with attendant benefits under Australian
employment law,9 whereas the recreational divers surveyed
in this study were predominantly employed on short-
term casual contracts in New Zealand. Though data were
collected from the recreational group independently of
their employers, the lack of sick leave provisions for many
chi square (df)
1261.9 -631 1 v 2 1.9 (1) 0.168
1261.3 -591 3 v 1 2.5 (2) 0.287
4 v 1
4 v 3
5 v 1
5 v 3
1 v 6
3 v 6
1298.2 -611 1 v 7 34.4 (4) < 0.01
8 v 1
8 v 3
1 v 9
1306.6 -616 1 v 10 42.8 (5) < 0.01
pDCS – probability of decompression sickness; DUR – dive duration (minutes); MSW – maximum depth in metres of sea water;
NUM – number of daily dives; SUR2DHS – delay between surfacing from last dive and completing the DHS; AIC – Akaike Information
Criteria; LL – log likelihood; df – degrees of freedom
Model improvement through variable removal and fitting to data on diving days only (n = 359)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201223
individuals may have influenced reporting of post-dive
symptoms, as previously found in other occupational diver
groups.17 Better correlation may be achieved by comparing
DHS scores to pDCS computed using a predictive model
developed using repetitive, multi-level air diving data.
The divers in this study were a relatively young, fit group
with a relatively high number of annual dives. This suggests
the possibilities of, firstly, selection bias whereby less fit
dive professionals may drop out of the industry or move
elsewhere leaving behind only the most suited and, secondly,
the potential for an acclimatisation to these elevated levels
of diving stress resulting in lower reported DHS score.
The potential advantages of the DHS as a tool for self-
assessment of post-dive health status both logistically in
terms of data collection and statistically when modelling
the results are substantial. The acquisition of field data to
complement laboratory dives used in the development of
decompression models remains an important goal, though
how well the DHS correlates with pDCS among other diving
cohorts remains to be seen.
The DHS score was most strongly associated with the daily
number of dives, increasing by 0.8 for each additional dive
made in a day, but did not correlate highly with pDCS values
calculated using the LE1 model. Reasons for this may be
that the LE1 model is a poor predictor of decompression
stress in this population of divers, the DHS tool may be too
insensitive to detect variation in decompression stress or sub-
clinical DCS in this group, or the DHS may not be a good
outcome measure in this population. Utility of the DHS for
measuring daily decompression stress among occupational
divers working in the recreational diving industry in New
Zealand remains unproven.
Our sincere thanks are extended to Dr David Doolette
for advice and for performing the correlations and to the
recreational dive professionals who took part in this study.
1 Boycott A, Damant G, Haldane J. The prevention of
compressed air illness. J Hyg (Lond). 1908;8:101.
Weathersby P, Homer L, Flynn E. On the likelihood of
decompression sickness. J Appl Physiol. 1984;57:15-25.
Weathersby PK, Survanshi S. Data quality for decompression
modeling. In: Sterk W,. Hamilton RW, editors. Operational
Dive and Decompression Data: Collection and Analysis.
EUBS Report (DATA) 17-8-90. Amsterdam: Foundation for
Hyperbaric Medicine; 1991. p. 94-9.
Doolette D. Field identification of decompression sickness.
SPUMS Journal. 2000;30:203-5.
Gault K, Tikuisis P, Nishi R. Calibration of a bubble evolution
model to observed bubble incidence in divers. Undersea
Hyperb Med. 1995;22:249-62.
Thalmann E, Parker E, Survanshi S, Weathersby P. Improved
probabilistic decompression model risk predicitions
using linear-exponential kinetics. Undersea Hyperb Med.
Doolette DJ. Psychometric testing of a health survey for field
reporting of decompression outcome. Undersea Hyperb Med.
Doolette DJ. Health outcome following multi-day occupational
air diving. Undersea Hyperb Med. 2003;30:127-34.
Doolette DJ, Gorman DF. Evaluation of decompression
safety in an occupational diving group using self reported
diving exposure and health status. J Occup Environ Med.
10 Doolette DJ, Goble S, Pirone C. Health outcome of hyperbaric-
chamber inside attendants following compressed-air exposure
and oxygen decompression. SPUMS Journal. 2004;34:63-7.
11 Harris RJ, Doolette DJ, Wilkinson DC, Williams DJ.
Measurement of fatigue following 18 msw dry chamber dives
breathing air or enriched air nitrox. Undersea Hyperb Med.
12 Doolette DJ. Decompression practice and health outcome
during a technical diving project. SPUMS Journal.
13 Fock A. Health status and diving practices of a technical diving
expedition. Diving Hyperb Med. 2006;36:179-85.
14 Wolfinger R, Chang M, editors. Comparing the SAS GLM and
MIXED procedures for repeated measures. 20th Annual SAS
Users Group Conference; 1995 April; Orlando, Florida.
15 Ngo L, Brand R. Model selection in linear mixed effects
models using SAS Proc Mixed. SUGI 22: Cary, NC: SAS
Institute; 1997. p. 1335-41.
16 Bauer D, Sterba S, Hallfors D. Evaluating group-based
interventions when control participants are ungrouped.
(Supplementary material: SAS Proc MIXED syntax for
evaluating treatment and covariate effects with partially nested
data). Multivariate Behav Res. 2008;43:210-36.
17 Ross J, Macdiarmid J, Oman L, Watt S, Lawson, A. Health-
related quality of life in former North Sea divers (letter). J
Occup Med. 2007;57:611-2.
Submitted: 11 April 2011
Accepted: 30 November 2011
Greg van der Hulst, BSurv, MBChB, PGDipMedSci,
FRNZCGP, is a general practitioner and rural hospital
doctor in Northland, New Zealand.
Peter Buzzacott, BA, MPH, PhD, is a research associate
at the School of Sports Science, Exercise and Health, The
University of Western Australia, Perth, Australia.
Address for correspondence:
Dr Greg van der Hulst
Dargaville Medical Centre
PO Box 257
Dargaville 0340, New Zealand
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201224
Postal survey of fitness-to-dive opinions of diving doctors and
Chris Sames, Des Gorman and Simon Mitchell
(Sames C, Gorman D, Mitchell S. Postal survey of fitness-to-dive opinions of diving doctors and general practitioners.
Diving Hyperb Med. 2012;42(1):24-29.)
Aim: To determine the consensus and concordance with published standards and expert opinion among New Zealand’s
designated diving doctors (DDDs) and general practitioners (GPs) regarding medical fitness-to-dive.
Methods: A postal survey canvassed doctors’ opinions regarding fitness to dive of 20 ‘real-life’ applicants with potentially
relevant medical conditions. In 17 cases, a ‘desired response’ was identified as expert opinion and the relevant published
Standard concurred; the remaining three cases were excluded from analysis. Consensus was measured between the groups of
doctors, and concordance measured against the ‘desired response’. The performance of the DDDs was also correlated with
both the number of diver medical assessments conducted annually and time since completing a diving medicine course.
Results: Seventy-seven of 98 DDDs (79%) and 75 of 200 GPs (38%) responded to the questionnaire. The mean concordance
was 60% and 50% for DDDs and GPs respectively. Consensus between DDDs and GPs was generally high, but was low
between these groups and the ‘desired response’. DDDs’ concordance was negatively correlated (r = -0.3) with time since
undertaking a diving medicine course, but was positively correlated (r = 0.2) with their annual number of dive medical
assessments. Both groups were more likely to differ from the ‘desired response’ by considering an ‘unfit’ diver as ‘fit’ than
Conclusions: There is poor concordance between doctors assessing fitness to dive and both published recommendations
and expert opinion when there is a relevant medical condition. This supports the current New Zealand practice of centralised
audit of occupational diver medical fitness prior to certification.
Fitness to dive, medical examinations, compressed-gas divers, scuba divers, recreational divers, occupational divers
In New Zealand (NZ), the estimated compressed-gas diver
fatality rate was 5.8 deaths per 100,000 divers per year
during 1996–2000,1 or a mean death rate of 6 per year from
1980–2006.1,2 This figure represents only about 5% of
drowning fatalities and suggests that diving is a relatively
safe occupation or pastime. However, of the 40 diver deaths
in NZ from 2000–2006, 12 should have been disqualified
from diving on medical grounds and, although the
relationship between the medical condition and the accident
was often unclear, these pre-existing medical conditions
were considered by the coroner to be either causative or
contributory to their deaths.2
Recreational divers in NZ are required to undergo a medical
examination conducted by a medical practitioner prior
to concluding training. There is no requirement for the
examining doctor to have undergone training in diving
medicine, and there is no ongoing health surveillance for
these divers. In contrast, occupational divers undergo a
five-yearly medical examination conducted by a ‘designated
diving doctor’ (DDD) who has undertaken post-graduate
training in diving medicine recognised by the South Pacific
Underwater Medicine Society (SPUMS). In intervening
years, the divers complete an annual health questionnaire.
Both the medical examination documentation and the
annual health questionnaires are independently reviewed
by an expert medical panel. This system has been shown
to be reliable, but controversy periodically arises about the
justification for expert and independent review of the medical
One reason for such a review is the potential for inconsistency
in decision making, even between doctors trained in diving
medicine. A previous study of doctors in Queensland,
Australia, who had training in diving medicine, showed a
low level of consensus in regard to the impact of certain
medical conditions on ‘fitness’ to dive.4 Similar problems
were found in a review of the process used to certify civil
pilots fit to fly in NZ.5,6
The present study re-examined this issue in NZ; the aim
was to determine consensus and concordance with expert
opinion among NZ DDDs and general practitioners
(GPs) regarding fitness for diving (both occupational and
recreational), to consequently see if there is an ongoing need
for independent review or arbitration of occupational diving
medical evaluations and to identify possible improvements
to recreational diving medical evaluations.
A questionnaire describing 20 compressed-gas diving
candidates who had a medical condition that could affect
diving fitness was mailed, along with a reply-paid envelope,
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201225
to two groups of doctors. The first was the cohort of DDDs
currently registered with the NZ Department of Labour for
the conduct of occupational diving medical evaluations (n =
98). The second group comprised GPs selected alternately
from the local (Auckland area) telephone book (n = 200),
who were asked to complete the survey if they conducted
diving medical fitness examinations for recreational divers
as part of their normal practice, but only if they had not
done a course in diving medicine. The questionnaires were
anonymous, but coded by administrative staff for later
identification to enable feedback. Incentive to complete the
questionnaire was offered in the form of Continuing Medical
Education (CME) points (RNZCGP), and for the DDDs, the
completion was a requirement to retain registration.
The cases were selected by one of us from recreational
diver candidate clinical records and the NZ occupational
diver medical database on the basis that there was a medical
condition that could adversely impact risk in compressed-gas
diving. The case set was then culled to a final set of 20 to
obtain a mix of organ system issues and to obtain a set where
the ‘certification outcome’ would include a selection of
positive, uncertain (where further investigations were needed
to better define the level of individual risk) and negative
responses (see Table 1). Two of us (DG and SM), both of
whom are certified in diving medicine by the Australian
and New Zealand College of Anaesthetists, represented the
‘expert review panel’.
Respondents were asked to categorise the medical fitness
for compressed-gas diving for each of the 20 scenario
candidates into one of three categories: medically fit to
dive in accordance with the standards that apply in New
Zealand; uncertain medical fitness for compressed-gas
diving or as being medically unfit for compressed-gas diving.
Respondents were also asked to write brief comments to
justify their answers.
The DDDs were also asked to provide additional information
in the form of an estimate of the number of dive medicals that
they conducted per year, and the number of years that had
elapsed since they had completed a diving medicine course
that would entitle them to DDD recognition.
Responses were compared to the opinion of the expert
panel and on the outcome likely from a consideration of the
Australian and New Zealand Standards for compressed-gas
divers.7–9 Expert opinion differed in three cases (scenarios
10, 11 and 19), which were therefore excluded from further
analysis. The expert opinion for the remaining 17 cases
was also predictable from a consideration of the Standard
and hence is used here as the ‘desired response’. Unless
specifically stated, the scenarios were assumed to refer to
recreational divers. For each respondent, the ‘concordance
score’ was the percentage of scenarios where there was
agreement with the ‘desired response’. For each scenario,
the ‘concordance score’ was the percentage of respondents
agreeing with the ‘desired response’. We have used the term
‘consensus’ to describe agreement within or between groups,
whereas ‘concordance’ is used to describe agreement of an
individual or group with a reference standard.
Statistical analysis was completed using SPSS software.
Randolph’s free-marginal kappa values (k) were derived to
demonstrate consensus within each group of assessors and
account for agreement by chance. To compare the DDDs
with the GPs, both having been measured against the ‘desired
response’, Student’s t-test of means (two-tailed) was used.
To describe the correlation between concordance with the
‘desired response’ and time since completing a dive medicine
course or number of dive medicals annually, Pearson’s
correlation coefficient (r) was derived.
The responses to the 20 scenarios are shown in Table 1, as
well as the ‘desired response’ and the relevant Standards
sections. Seventy-seven of 98 DDDs (79%) and 75 of
200 GPs (38%) responded to the questionnaire. The mean
concordance score was 60% (range 24–88%) and 50% (range
12–82%) for DDDs and GPs respectively. By scenario, the
mean concordance was 61% (range 26–94%) and 50%
(range 19–89%) for DDDs and GPs respectively (Figure 1).
Consensus within each group was 52% (k = 0.28) and 46%
(k = 0.18), for the DDDs and GPs respectively. Although
both groups scored poorly, Student’s t-tests of means showed
DDDs were significantly more likely to express concordance
with the ‘desired response’ than GPs (t = 3.88, 150 df, P
= 0.0002). For those DDDs who provided the additional
information (n = 51), there was a negative correlation (r =
-0.3, P = 0.03) between their concordance score and the time
elapsed since they completed a designated dive medicine
course, and a positive correlation (r = 0.2, P = 0.03) with
the number of dive medicals they did each year.
The probability of assessing an ‘unfit’ diver as ‘fit’ was higher
for GPs than DDDs (17.3% versus 11.7% respectively), and
Concordance of responses of doctors with basic training
in diving medicine (DDD) and non-trained general
practitioners (GP) with Standard responses to fitness-to-dive
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201226
Twenty fitness-to-dive case scenarios with responses from doctors with basic training in diving medicine (DDDs; n = 77)
and non-trained general practitioners (GPs; n = 75)
Case Scenario description, 'desired response' and relevant Standards sections Group Fit Unfit Unsure
1 A 23-yr-old female with bipolar affective disorder and a history of psychotic
symptoms, well controlled on Lithium.
Refs: 7) A4.14b 8) A4.9 9) K4.15d,g
A 32-yr-old female who has a history of 2 spontaneous left-sided
pneumothoraces, but who has had corrective surgery to the apex of her left
lung; spirometry normal.
Refs: 7) A4.10b,ii 8) A4.10b,ii 9) K4.11ii
A 190 cm 31-yr-old customs diver with an FVC of 7L but an FEV1/FVC of
0.69; chest X-ray, hypertonic saline challenge results and exercise tolerance
Refs: 7) A4.10c 8) A4.10d 9) K4.11c
A fit 21-yr-old male who has Mobitz type 1 (Wenckebach) second degree
heart block on resting ECG, but a normal exercise ECG.
Refs: 7) A4.9 8) A4.9a 9) K4.10
A fit, asymptomatic 25-yr-old female with a soft systolic cardiac murmur
heard best in the aortic region.
Refs: 7) A4.9a 8) A4.9a 9) K4.10
A 20-yr-old female with a history of ‘wheezy bronchitis’ in childhood. She
used inhalers until she was 12 yrs old but has not used any since then. Plain
spirometry results are normal.
Refs: 7) A4.10b,iv 8) A4.10b,v 9) K4.11
7 A 54-yr-old male hypertensive controlled with a diuretic. He has a normal
exercise ECG and renal function.
Refs: 7) A4.9c 8) A4.9c 9) K4.10
A 24-yr-old male with cerebral palsy who is able to walk with the use of
Refs: 7) A4.3/A4.12 8) A4.3/A4.12 9) K4.13
An asymptomatic 45-yr-old male with atrial fibrillation diagnosed and fully
investigated 10 years ago. He remains on warfarin and has normal exercise
Refs: 7) A4.9 / 4.14b 8) A4.9a / 4.14a 9) K4.15d / K4.10
A 28-yr-old male with a BMI of 40. An exercise ECG to level 4 Bruce
protocol showed no ischaemic changes.
No agreement between ‘experts’
Refs: 7) A4.4 8) A4.4 9) K4.3
11 A 32-yr-old diver found on an epidemiological survey to have a patent
foramen ovale (bubble contrast echo). He has been a Navy operational diver
for 10 years without incident.
No agreement between ‘experts’
Refs: 7) A4.9 8) A4.9 9) K4.10
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 27
Table 1 (cont)
BMI – body mass index; CXR – chest X-ray; ECG – electrocardiogram; EEG – electroencephalogram; FEV1 – forced
expiratory volume in 1 s; FVC – forced vital capacity; MRI – magnetic resonance imaging
was also significantly higher for both GPs and DDDs than
the converse probability of assessing a ‘fit’ diver as ‘unfit’
(3.3% and 2.6% respectively).
Concordance scores varied by greater than 15% (mean
variance 27.7%) between DDDs and GPs (DDDs higher than
GPs) in six of the scenarios (1, 2, 5, 6, 13 and 14). For the
remaining 11 scenarios, the consensus between DDDs and
GPs was high (mean variance 3.9%). The concordance with
the ‘desired response’ was < 40% for both DDDs and GPs in
four of the 17 scenarios (three in common: scenarios 14, 15
and 20; DDDs in scenario 4, and GPs in scenario 13).
Case Scenario description, 'desired response' and relevant Standards sections Group Fit Unfit Unsure
12 A 19-yr-old male with a history of convulsions as an infant, for which he
was maintained for several years on phenobarbitone. The family GP has no
record of any fits.
Refs: 7) A4.8b 8) A4.8b 9) K4.9
A 25-yr-old male who had a chest drain inserted after he suffered broken
ribs and a haemo-pneumothorax three years ago in a car accident. He is back
playing club rugby. His CXR and spirometry are normal.
Refs: 7) A4.10b,ii 8) A4.10b,ii 9) K4.11a,ii
14 A 45 kg, 14-yr-old female school swimming champion.
Refs: 7) A4.2 8) A4.2 9) K4.2
15 A 35-yr-old female with asthma since her teens. She is well-controlled on
twice daily Fluticasone and last used her Salbutamol inhaler three months
ago. She had a normal result on a recent hypertonic saline challenge test.
Refs: 7) A4.10b,iv 8) A4.10b,v 9) K4.11a,iii
16 A 22-yr-old female with a history of severe head injury 5 years previously
with small subdural haematoma but no surgical intervention. She fitted at the
time. Was on Epilim for 2 years and has had no fits since discontinuing it.
Recent MRI and EEG normal. She has had ongoing minor cognitive deficits
Refs: 7) A4.8c 8) A4.8d 9) K4.9
A 29-yr-old female with a history of migraines. She has had no symptoms
for the past year on prophylactic medication, but suffered severe bifrontal
and occipital headaches during two familiarisation dives, the headaches
onset at depth.
Refs: 7) A4.8 8) A4.8c 9) K4.9
A 26-yr-old professional diver who was treated for neurological DCI 3
Refs: 7) A4.8 8) A4.8 9) K4.15j
A 49-yr-old male diabetic controlled by diet alone. He has mild diabetic
No agreement between ‘experts’
Refs: 7) A4.14 8) A4.14 and appdx D 9) K4.15
A 48-yr-old male with a past history of severe angina who has undergone
successful coronary vessel grafting three years ago; no angina now and good
Refs: 7) A4.9 8) A4.9 9) K4.10
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 28
The scenarios used in this survey were selected to include
important respiratory, cardiovascular and neurological health
issues for divers. Many of our ‘real-life’ cases were similar
to those used in the Queensland study, some of which were
fictitious and some real, emphasising that these are the
kind of medical conditions that arise relatively commonly
in assessing would-be divers.4 They were also selected to
present a challenge to the assessing doctors as compared
to more straightforward cases, which represent the great
majority of assessments. It follows that the current survey
does not represent the outcome likely from a random
selection of cases in which a much higher concordance
would be expected.
The overall 38% response rate for surveyed GPs is likely to
mask a much higher response rate for those GPs who fulfilled
the inclusion criteria (those who conduct recreational diving
medical fitness examinations but have not completed a diving
medicine course) as many GPs do not undertake diving
The published standards for fitness to dive are conservative,
and if strictly applied they may result in divers being
inappropriately denied medical clearance for diving.7–9
However, the finding that both DDDs and GPs were more
likely to assess an unfit or indeterminate diver as fit, rather
than the converse, suggests either disagreement with, or a
lack of familiarity with the published standards, as the bias
in the latter is in the opposite direction.
There was a wide range of opinions and a low mean
concordance with the ‘desired response’ for both DDDs and
GPs. This, together with the negative correlation between
concordance score and time since completing a designated
diving medicine course, suggests potential benefit could
arise from periodic refreshers and/or regular formative
assessments of DDDs and GPs. It also suggests that the most
reliable method of assessing someone’s medical fitness for
occupational diving involves an expert in diving medicine
and/or a risk evaluation conducted by a specifically trained
doctor who has ready access to expert advice. The problem
with either of these ‘solutions’ is that there are very few
diving medicine experts and hence access would be limited.
The central audit facility for employed divers that exists in
New Zealand is a workable solution to this problem and is
clearly independent and less vulnerable to diver-advocacy
bias. It is noteworthy that many divers who might otherwise
have been disqualified, have been able to continue a career
in diving, with specified constraints, due to the intervention
of this facility.
For recreational divers, there is evidence both supporting
and refuting the utility of a medical examination prior to
training.10–12 In the face of this controversy, most countries
have now adopted a self-declaration health questionnaire
for recreational scuba diving candidates in line with the ISO
standards.13 However, for occupational divers, there remains
a widespread reliance on annual medical examinations
conducted by doctors analogous to our DDDs. Our study
suggests that in the absence of independent review, there
is a strong possibility that candidates with significant
medical conditions who undergo such an examination will
receive a determination of fitness different to that which
an expert would deliver or that expected by consideration
of the relevant Standard. To the extent that we derived a
‘desired response’, this study suggests that independent
review by such experts is a valuable adjunct to the process
of occupational diver evaluation.
LIMITATIONS OF THE STUDY
The respondents, both DDDs and GPs, were asked only to
assess the diving candidates’ fitness to dive on the basis of the
brief vignette. There was no specification regarding fitness
for occupational versus recreational diving. Therefore, it is
possible that some of the respondents, especially the GPs,
may have applied a more liberal ‘informed risk acceptor’
approach in their decision making. It should be noted,
however, that there are very few differences between the
published standards for occupational and recreational
This study supports the need for better, iterative and
formative diving medical education for DDDs, and the
desirability of diving medical education for any GP who
wishes to conduct recreational dive medicals. The overall
low concordance of both DDDs and GPs with published
recommendations and expert opinion is mitigated for DDDs
performing occupational diving medicals in the New Zealand
setting by the existence of a central, independent and expert
Conflict of interest
Drs Des Gorman and Chris Sames are members of the
Department of Labour Diving Medical Directorate,
which is responsible to the Department of Labour for the
certification of the medical fitness of occupational divers in
1 Davis M, Warner M, Ward B. Snorkelling and scuba diving
deaths in New Zealand, 1980-2000. SPUMS Journal.
McClelland A. Diving-related deaths in New Zealand 2000-
2006. Diving Hyperb Med. 2007;37:174-88.
Sames C, Gorman D, Mitchell SJ, Gamble G. Utility of
regular medical examinations of occupational divers. Internal
Medicine Journal. 2009;39:763-6.
Simpson G, Roomes D. Scuba diving medical examinations
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201229
in practice: a postal survey. MJA. 1999;171:595-8.
Gorman DF, Scott PJ. The process of determining fitness to fly
aeroplanes in New Zealand: A review of current practice and
recommended changes. Wellington: Civil Aviation Authority
of New Zealand; 2001.
Gorman DF, Scott PJ. The process of determining fitness to
fly aeroplanes in New Zealand: A follow up audit report of
current practice and recommended changes. Wellington: Civil
Aviation Authority of New Zealand; 2003.
Training and certification of recreational divers – Part 1:
minimum entry-level SCUBA diving (AS 4005.1 – 2000), 2nd
ed. Sydney: Standards Australia; 2000.
Guidelines on medical risk assessment for recreational diving.
Melbourne: South Pacific Underwater Medicine Society;
2011. Available at: http://www.SPUMS.org.au.
Occupational diving operations – Part 1: standard operational
practice (AS/NZS 2299.1 – 2007), 2nd ed. Sydney/Wellington:
Standards Australia/Standards New Zealand; 2007.
10 Meehan CA, Bennett MH. Medical assessment of fitness to
dive – comparing a questionnaire and a medical interview-
based approach. Diving Hyperb Med. 2010;40:119-24.
11 Glen S, White S, Douglas J. Medical supervision of sport
diving in Scotland: reassessing the need for routine medical
examinations. Br J Sports Med. 2000;34:375-8.
12 Glen S. Three year follow up of a self certification system for
assessment of fitness to dive in Scotland. Br J Sports Med.
13 ISO 24801-2:2007. Recreational diving services – Safety
related minimum requirements for the training of recreational
scuba divers – Part 2: Level 2 – Autonomous diver. Geneva,
Switzerland: International Organization for Standardization;
Submitted: 01 June 2011
Accepted: 03 January 2012
Chris Sames, BSc, MBChB, MMedSc, is Hyperbaric Medical
Officer, Naval Health Unit, Royal New Zealand Navy,
Des Gorman, MD, PhD, is Professor of Medicine, University
of Auckland, Executive Chairman of Health Workforce New
Zealand and Director of Diving and Hyperbaric Medical
Simon Mitchell, PhD, FANZCA is Associate Professor, Head
of the Department of Anaesthesiology, University of Auckland
and Consultant Anaesthetist, Dept of Anaesthesiology,
Auckland District Health Board.
Address for correspondence:
Dr Chris Sames
Slark Hyperbaric Unit, Naval Health Unit
91 Calliope Road
Private Bag 32901
Devonport, Auckland 1309
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201230
Rapid ascent and buoyancy problems among Western Australian
certified recreational divers
Peter Buzzacott, Terri Pikora, Michael Rosenberg and Jane Heyworth
Buzzacott P, Pikora T, Rosenberg M, Heyworth J. Rapid ascent and buoyancy problems among Western Australian certified
recreational divers. Diving Hyperb Med. 2012;42(1):30-35.)
Introduction: We investigated risk factors associated with ascending rapidly and/or losing buoyancy control among
Methods: Dive and diver information were collected and depth/time loggers attached to recreational divers. Case dives
recording an ascent > 18 m min-1 were compared with control dives made at the same dive site and time by divers recording
ascents ≤ 18 m min-1. In a second analysis, case dives with reported buoyancy problems were compared with control dives
during which no problems were reported. Conditional logistic regression identified factors significantly associated with
ascending faster than 18 m min-1 or reporting a buoyancy problem.
Results: In total, 1,032 dive profiles were collected. Case dives (n = 71) recording an ascent > 18 m min-1 were compared
with 282 control dives. The main risk factor for making a rapid ascent was a loss of buoyancy control. Case dives were also
shorter. Dives resulting in reported buoyancy problems (n = 68 cases) were compared with 320 control dives. The three main
risk factors for buoyancy problems were an inability to describe how to check for neutral buoyancy, reportedly not being in
control during the final ascent and maximum ascent rates that were a mean of 20% faster than during control dives.
Conclusions: Further research is necessary to identify if ascending rapidly is the result of a loss of buoyancy control, a lack
of ascent rate reference or a failure to appreciate the potential consequences of ascending rapidly. The inability of many
divers to describe how to check for neutral buoyancy also deserves attention.
Ascent, buoyancy, risk factors, recreational diving, scuba diving
Recreational scuba diving is enjoyed by tens of thousands
in Western Australia (WA).1 Each year in WA, on average,
40 divers are treated for decompression illness (DCI) in
the Fremantle Hospital hyperbaric facility and two divers
die.2,3 In addition, it is likely hundreds of people suffer
minor diving-related morbidity such as marine stings,
ruptured tympanic membranes and pain-only bends for
which treatment is not sought.4 The most serious forms of
diving morbidity are severe DCI and near drowning, and the
most common cause of death among recreational divers is
drowning.5 Loss of buoyancy control and/or rapid ascent
are known diving problems that may lead to drowning and/
or DCI.6,7 Experienced together they are far more likely to
result in injury than either problem alone.8
Rapid ascent was among the top ten contributory factors
reported in 286 American diving fatalities.9 Among 34
breath-hold embolisms, 13 involved rapid ascents and an
analysis concluded “rapid ascent is the most frequently
reported contributory cause of incident”.10 These problems
are just as prevalent among WA divers as they are among
other diving populations.4 Information on the reasons why
divers lose buoyancy control and/or ascend rapidly (i.e.,
faster than 18 m min-1) is limited.11 A Delphi survey of
diving experts suggested the most likely reasons recreational
divers experience these problems. They are shown in order
of likelihood in Table 1.12
Despite the similarity of reasons suggested for each of these
dive problems a recent cross-sectional analysis of 46,801
recreational open-circuit scuba dives made by 4,711 adult
divers found that divers ascending faster than 18 m min-1
(n = 235 divers) were more likely to be younger, male and
have a higher diver certification level, while divers who
reported losing buoyancy control (n = 223 divers) were more
likely to be older, female and have basic diver certification.13
Controlling for age and sex by comparing dives involving a
Potential reasons for
Fail to release gas
Run out of breathing gas Incorrect weighting
Incorrect use of BCD
Ignorance of safe ascent
Incorrect body position
Fail to monitor depth
Loss of weight system
Fail to release gas
Incorrect body position
Incorrect use of BCD
Loss of weight system
Potential reasons for ascending rapidly and losing buoyancy
control in order of suspected likelihood;
(BCD – buoyancy control device)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 31
reported rapid ascent (n = 296) with dives made by the same
divers with no reported rapid ascent (n = 2,598), rapid ascent
dives were shallower, shorter, more likely made from a boat
and were perceived as strenuous.13 Comparing 362 dives
with reported buoyancy problems to 3,174 dives without
buoyancy problems made by the same group of divers, the
study found that buoyancy problem dives were more likely
to have been shorter, made from a live-aboard or day-boat
and to have involved a higher perceived workload.13
By controlling for environmental factors associated with the
dive site and type of dive platform this study aims to further
explore potential factors that increase the risk of losing
buoyancy control and/or ascending rapidly. The maximum
safe rate of ascent recommended by the Professional
Association of Diving Instructors is 18 m min-1.14
Adult certified divers attending organised recreational group
dives were recruited as previously described.3,15 Briefly, dive
businesses and dive clubs in WA were invited to participate.
A researcher (PB) met the divers at popular dive sites around
the coast of WA. The study was approved by the Human
Research Ethics Committee of the University of Western
Dive and diver information were collected using a modified
Divers Alert Network (DAN) Project Dive Exploration
(PDE) questionnaire and Sensus Ultra™ data-loggers
(ReefNet, Mississauga, Ontario) were attached to the front
of each diver’s buoyancy control device (BCD). Depths, (to
+/- 0.01 m resolution and 0.3 m accuracy16), were recorded
every 10 seconds and downloaded from each logger. Diver
data collected included sex, age, weight, dive experience,
certification level and problems experienced during the
dive. Self-reported starting and finishing gas pressures and
stamped cylinder volumes were recorded on the dive record.
Consumed volume of gas was calculated by multiplying
cylinder volume by the difference between starting and
ending cylinder pressures, expressed as surface-equivalent
air consumption (SAC) per kilogram of body,weight, (L
Mean depth was calculated by dividing the total of recorded
depths from each dive by the number of samples recorded
between the time the diver left the surface (depth >1 metre
sea water, msw) and the time returned to the surface (depth
= 0). This included divers swimming back to the boat
underwater but excluded time spent at the surface. For
example, when taking a bearing back to the boat near the
end of a dive it is assumed that divers at the surface would
have temporarily discontinued using scuba and breathed
air from the atmosphere. Surface air consumption was
calculated by dividing the gas volume used by the number of
minutes spent underwater and by the mean ambient pressure
in bar at the mean depth, (excluding time at the surface, as
described above). Divers were asked “What is the maximum
recommended safe rate of ascent?” The maximum recorded
rate of ascent (m min-1) during each dive was calculated by
multiplying the maximum negative difference in depth in
msw during any single 10-second sampling period by six.
To control for environmental conditions two case-control
analyses were performed. In the first analysis, dives in which
a diver recorded an ascent rate > 18 m min-1 were classed
as rapid ascent ‘case’ dives and dives made at the same dive
site and at the same time without ascending faster than 18 m
min-1 were classed as ‘control’ dives. In the second analysis,
dives in which a diver reported a buoyancy problem were
classed as ‘case’ dives and dives made at the same dive site
and at the same time by at least one other diver without
reporting buoyancy problems were classed as ‘control’
dives. Data were imported into the Statistical Analysis
System (SAS) version 9.2 (Cary, North Carolina) and the
distribution of variables tested for normality. Bivariate
% with buoyancy problem
Cases (n = 71) Controls (n = 282)
2.27 to 11.13
% with low certification
Mean dive time
(per 5 mins)
No of dives in BCD worn
(per 100 dives)
Years of diving
(per 10 years)
Dives made in last 5 years
(per 100 dives)
76.0 54.0 2.58 1.26 to 5.30 0.03
40.8 48.3 1.33 1.15 to 1.54 <0.01
44.0 100.0 1.22 0.90 to 1.49 0.14
6.0 11.5 1.18 0.87 to 1.67 0.26
75.0 140.0 1.11 0.90 to 1.35 0.47
Bivariate associations with ascending faster than 18 m min-1
(* each risk factor modelled as per units indicated in parentheses)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201232
analyses were conducted for each factor. Variables with
expected cell counts of less than five were excluded from
further analysis. Remaining factors were fitted to conditional
logistic regression models for reporting buoyancy problems
and ascending rapidly. This was achieved by numbering each
organised dive consecutively and stratifying the regression
by dive number. Non-significant associations (P > 0.05)
were removed by backwards elimination.
A description of the participants and the range of diving
conditions has been reported previously.3,15 A total of 1,032
dives were recorded. Of these, 71 dives were made with
recorded ascents faster than 18 m min-1 (‘case dives’) at the
same time as 282 dives were recorded with ascents no faster
than 18 m min-1 (‘control dives’). In a second analytical
sub-set from the 1,032 dives recorded, 68 dives were made
by divers reporting buoyancy problems (‘case dives’) at the
same time as 320 dives during which no buoyancy problems
were reported (‘control dives’).
RAPID ASCENT SUB-SET
Case dives (n = 71) recorded a mean maximum depth of
21.0 (SD 10.0) msw whilst the mean maximum depth during
control dives (n = 282) was 19.7 (9.4) msw (P = 0.30). Case
dives ascended at a median maximum rate of 20.1 m min-1
(range 18.3 to 39.6) whilst the median maximum ascent rate
during control dives was 11.0 m min-1 (range 5.5 to 16.5).
During any 10-second period only one dive recorded an
ascent faster than 30 m min-1. In the thirty-fifth minute of
a dive with a median depth till then of 4.9 msw (maximum
17.9 msw), the diver ascended from 9.0 msw to 2.4 msw,
(a difference of 6.6 msw), recording a mean ascent rate
over 10 seconds of 39.6 m min-1. The dive was the first in
a three-dive series over two days, and the diver reported no
Divers making case dives more often than divers making
control dives reported their final ascent to have been
uncontrolled (24% versus 10%, P < 0.01). Table 2 presents
bivariate comparisons between case and control dives.
Divers self-reported their perceived workload for each dive
as ‘resting/light’, ‘moderate’ or ‘severe’. Case dives had a
higher SAC rate (0.30 L min-1 kg-1 versus 0.23 L min-1 kg-1,
P < 0.01). Based on mean values for the sample as a whole
(n = 1,032) this equates to SAC for control dives being
classed as ‘resting/light’ while case dives were classed as
‘severe’ (Table 3).
When asked “What is the maximum recommended safe rate
of ascent?” divers who did not know were more likely to
ascend faster than 18 m min-1 (35/135, 26%) than divers
who provided a numerical rate (36/208, 17%) (P = 0.05).
Figure 1 plots the recorded maximum ascent rate versus the
estimated maximum safe rate of ascent given by divers (n =
208 dives). In total, 80 dives (38%) exceeded the maximum
safe rate of ascent offered by the diver making the dive.
As Figure 1 shows, there was no correlation between the
stated maximum safe rate of ascent and the actual maximum
ascent rate (r = 0.006). The median recorded maximum rate
of ascent among the 208 dives made by divers able to offer
a numerical maximum safe rate was 11.9 m min-1 (range
5.5 to 39.6).
Multivariate analysis for rapid ascent
Fifteen dives (4%) were not considered because of missing
variables, leaving 338 of 353 dives (96%) in the analysis.
The main risk factor for making a rapid ascent (Table 4)
was a loss of buoyancy control. Shorter dives were also
significantly associated with recording a rapid ascent.
Factors removed by backwards elimination included years of
diving, number of dives made during the previous five years,
SAC mean (SD) 0.22 (0.07) 0.24 (0.08) 0.28 (0.05)
(L min-1 kg-1)
Resting/light Moderate Severe
Surface-equivalent air consumption (SAC) by perceived
workload overall (n = 1,032)
Actual maximum rate of ascent versus estimated maximum
safe rate of ascent during 208 dives; bubble size (area)
represents the number of data points (range 1 to 24)
(Yes versus No)
(per 5 mins)
1.84 to 9.70
1.29 1.12 to 1.50 <0.01
Multivariate risk factors for recording a rapid ascent
(following backwards elimination)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201233
level of certification (low, medium or high) and number of
dives conducted wearing the BCD used on those dives.
BUOYANCY PROBLEMS SUB-SET
Of 1,030 dives where the presence of any dive problem was
recorded (two were left blank), 68 (6.6%) reported buoyancy
problems (cases) during dives made at the same time and
place as 320 (31.0%) control dives during which divers did
not report a buoyancy problem when asked. Characteristics
of case dives and control dives are presented in Table 5.
Case dives had a higher mean SAC rate than control dives
(0.27 L min-1 kg-1 vs 0.22 L min-1 kg-1, P < 0.01). As found
in the rapid ascent case-control analysis, this equates to
control dives being classed as ‘resting/light’ and case dives
being classed as ‘moderate’ or ‘severe’ (Table 3). Among
case dives 24% exceeded the maximum recommended safe
rate of ascent of 18 m min-1 compared with 7% of control
dives (P < 0.01). Case dives were also made by divers who
had fewer dives’ experience with the BCD worn (55.0 versus
125.0, P < 0.01), and when asked, were more likely to state
they did not know what rate a maximum safe rate of ascent
might be (50% versus 35%, P < 0.01).
Multivariate analysis for buoyancy problems
Twenty-nine dives (7%) were not considered because of
missing variables leaving 359 of 388 (93%) in the analysis.
The three main risk factors for reporting a buoyancy problem
(Table 6) were divers who were unable to describe how
to check for neutral buoyancy, who reported not being
in control during the final ascent and dives that included
maximum ascent rates that were a mean of 20% faster than
control dives. Factors removed by backwards elimination
included the age of the diver, number of years of experience
and certification level.
This study explored potential factors that may increase the
risk of losing buoyancy control and/or ascending rapidly,
based on suggestions from an ‘expert’ panel.12 While many
of the potential reasons were supported, several were not.
Ascending rapidly was significantly associated with reporting
a buoyancy problem. However, the wide confidence interval
suggests an imprecise estimate (Table 4). Whether a rapid
ascent followed a buoyancy problem or if rapid ascent was
interpreted as a buoyancy problem was not investigated in
this study. Ascending faster than 18 m min-1 was associated
with dives ending sooner (Table 4) though it cannot be stated
with certainty whether dives ended prematurely because
of unintentional ascents. Also, we found that 38% of the
208 recorded dives exceeded the rate of ascent given by
the diver as a maximum safe limit. However, there was no
correlation between stated maximum safe ascent rate and
actual maximum ascent rate (Figure 1). Faster ascent rates
have been found to generate higher Doppler-detected venous
bubble counts.17 Bubbles are, however, present in otherwise
uneventful dives and do not necessarily result in DCS.11
% not in control
(Low vs High)
Unable to check for
(per 10 years)
Faster max. ascent rate
(per m min-1)
Fewer years’ diving
(median; per year)
(n = 68)
(n = 320)
Unadjusted OR 95% CI
26.75 10.10 to 70.81 <0.01
74:16 52:36 4.36 1.96 to 9.68 <0.01
80 48 4.28 2.18 to 8.43 <0.01
45.2 41.6 2.16 1.48 to 3.19 <0.01
1.17 1.09 to 1.25 <0.01
6.0 12.0 1.03 1.00 to 1.07 0.07
Bivariate associations with reporting a buoyancy problem
(* each risk factor modelled as per units indicated in parentheses)
In control during
ascent (No vs. Yes)
Able to check for
(No vs. Yes)
Faster max. ascent
rate (per m min-1)
9.93 to 91.88
7.76 2.95 to 20.41 <0.01
1.10 1.00 to 1.21 0.04
Multivariate risk factors for reporting a buoyancy problem
(following backwards elimination)
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201234
Therefore, for reasons that remain unclear and warrant
further research, educating recreational divers about a
numerical recommended safe ascent rate limit appears to be
ineffective among a substantial proportion of them. Almost
one quarter of the divers in the current study commented
that they relied upon the speed of their exhaled bubbles as a
marker for ascending safely. However, there is no published
guideline specifying what size of bubble ascends slower
than 18 m min-1 and bubble ascent rate may be affected by
salinity and water temperature. Coupled with the difficulty
associated with magnification of bubbles due to the differing
refractive indices of water-to-glass and glass-to-air, bubbles
are likely to be an unreliable gauge of ascent rate.11
Self-reported buoyancy problems were found in this study
to be significantly associated with being unable to describe
how to check for neutral buoyancy, though once again, the
wide confidence intervals suggest an imprecise estimate of
the added risk. In the Delphi study (Table 1), poor training/
skill level was considered the third most likely cause of
divers losing buoyancy control.12 Insufficient knowledge
or training was identified as early as 1964 as a risk factor
in 50% (n = 83) of British diving fatalities.18 Explanations
for why dives made by divers who were unable to describe
how to check for neutral buoyancy were more likely to
involve buoyancy problems include that they may have
begun the dive incorrectly weighted, as also suggested in
the Delphi study, or that they may not have known how
to establish neutral buoyancy during the dive. However,
the exact reasons why divers who were unable to describe
how to check for neutral buoyancy were also more likely to
self-report a buoyancy problem remain undetermined and
require further research.
At the bivariate level, case dives were also made by divers
with less dive experience with the BCD worn, as suggested
in the Delphi study (Table 1), where unfamiliar equipment
was ranked the sixth most likely reason divers lose buoyancy
control.12 Case dives recorded a higher mean SAC rate.
Referring back to Table 3, this equates to control dives
being classed as ‘resting/light’ and case dives classed as
‘moderate’ or ‘severe’, suggesting that buoyancy problems
were associated with the workload of a dive, as has been
reported elsewhere.13 After adjusting for potential risk
factors, reporting a buoyancy problem was associated with
reporting being out of control during the final ascent and
recording a faster maximum mean ascent rate over at least
10 seconds. In the Delphi study, failing to release air during
ascent was listed as the second most likely cause of divers
losing buoyancy control.12 However, while failing to release
air during ascent may explain reporting of both a buoyancy
problem and an out-of-control ascent in the current study,
the exact causes of these problems were not identified nor
the volume of air released during ascent measured.
Limitations of this study include that it remains uncertain
how non-participants may have differed to participants.
How self-organised dives may differ to professionally
organised dives was also not explored. Therefore, caution is
needed in generalising these findings beyond the population
The 10-second sampling rate was selected for data-loggers
to capture sustained ascents whilst ignoring lesser vertical
fluctuations, for example, caused by overhead swell or a
diver’s breathing. No physiological consequences were
measured following each ascent, and this study does not
establish a clear link between risk factors for rapid ascent
over ten seconds and actual diving morbidity. It remains
possible, likely even, that diving morbidity is more strongly
associated with ascents sustained beyond 10 seconds’
duration. It is also possible that rapid ascent for at least 10
seconds carries greater risk of injury in the shallows than
ascent from deeper depths and when it occurs at the end of
a dive rather than earlier. In this study, however, any ascent
over 10 seconds was included regardless of when it occurred
during the dive. In short, it is likely that not all ascents carry
equal risk but all were treated equally in this study, in keeping
with the advice of diver training agencies to not exceed a
linear ascent rate of 18 m min-1.14
Despite the widespread availability and use of personal dive
computers with in-built audible and/or visual ascent-rate
alarms, (and despite many divers stating a maximum safe
rate of ascent of 18 m min-1 or less), many divers in this
study ascended faster than 18 m min-1. Additional research
is necessary to explore why divers ascend so rapidly. Key
issues that need identifying include whether ascending
rapidly is linked to a loss of buoyancy control, a lack of
ascent-rate reference or a failure to appreciate the potential
consequences of ascending rapidly. The inability of many
divers to describe how to check for neutral buoyancy at
the start of the dive is concerning and deserving of further
We are grateful to Dr Petar Denoble and the DAN for
permission to use the PDE survey forms and for adapting
the PDE database to suit this project. We also thank database
managers Lisa Li of the DAN and Robin Mina of the School
of Population Health, the University of Western Australia.
1 Australian Bureau of Statistics. Participation in sport and
physical activities – 4177.0 – 1999–2000. Perth, WA:
Australian Bureau of Statistics; 2001.
Buzzacott P, Rosenberg M, Pikora T. Western Australian
recreational scuba diving fatalities, 1992-2005. Aust N Z J
Public Health. 2009;33:212-4.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201235
3 Buzzacott P, Pikora T, Heyworth J, Rosenberg J. Exceeding the
limits - estimating tissue pressures among Western Australian
recreational divers. Diving Hyperb Med. 2010;40:204-8.
Buzzacott P. Diving injuries amongst Western Australian scuba
course graduates [Masters dissertation]. Perth: University of
Western Australia; 2006.
Lippmann J. Review of scuba diving fatalities and
decompression illness in Australia. Diving Hyperb Med.
Vann R, Freiberger J, Caruso J, Denoble P, Pollock N,
Uguccioni D, et al. Report on decompression illness, diving
fatalities and project dive exploration. 2006 ed (based on 2004
data). Durham, NC: Divers Alert Network; 2006.
Cumming B. NDC diving incidents report. South Wirral,
Cheshire: British Sub-Aqua Club; 2008. Available from: www.
Acott C. Diving incidents – errors divers make. In: Safe Limits:
An international dive symposium. Cairns, QLD: Division of
Workplace Health and Safety; 1994. p. 25-38.
Scuba diving. A quantitative risk assessment. Report No:140.
Newport, Isle of Wight: Health and Safety Executive; 1997.
10 Tetlow S. Formal risk identification in professional SCUBA
(FRIPS). Cranfield: Cranfield University; 2006.
11 Oxer H. Safe limits – Assessing the risks. In: Safe limits:
an international dive symposium. Cairns, QLD: Division of
Workplace Health and Safety; 1994. p. 86-92.
12 Buzzacott P, Rosenberg M, Pikora T. Using a Delphi technique
to rank potential causes of scuba diving incidents. Diving
Hyperb Med. 2009;39:29-32.
13 Buzzacott P, Denoble P, Dunford R, Vann R. Dive problems
and risk factors for diving morbidity. Diving Hyperb Med.
14 Richardson D, editor. Go dive. Rancho Santa Margarita, CA:
International PADI Inc; 1999.
15 Buzzacott P, Rosenberg M, Heyworth J, Pikora T. Risk factors
for running low on gas in recreational divers in Western
Australia. Diving Hyperb Med. 2011;41:85-9.
16 Wilk K. Sensus Ultra developer’s guide. Mississauga, Ontario:
17 Carturan D, Boussuges A, Vanuxem P, Bar-Hen A, Burnet
H, Gardette B. Ascent rate, age, maximal oxygen uptake,
adiposity, and circulating venous bubbles after diving. J Appl
18 Miles S. One hundred and sixty-five diving fatalities. J R Nav
Med Serv. 1964;50:129.
Submitted: 18 July 2011
Accepted: 07 January 2012
Peter Buzzacott, BA, MPH, PhD, is a research associate
at the School of Sports Science, Exercise and Health, The
University of Western Australia.
Terri Pikora, BHSc, MPH, PhD, is an adjunct Research
Associate Professor at the School of Population Health, The
University of Western Australia.
Michael Rosenberg, BAppSc, DipEd, MPH, PhD, is
Associate Professor and the Director of the Health
Promotion Evaluation Unit at the School of Sport Science,
Exercise and Health, The University of Western Australia.
Jane Heyworth, BappSc, PGDipHlthSc, MPH, PhD, is
Professor and the Sub Dean of Health Science at the School
of Population Health, The University of Western Australia.
Address for correspondence:
Dr Peter Buzzacott
17 College Row,
Bunbury, WA 6230, Australia.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 36
Rapidly advancing technology has enabled ultrasound
machines to become more affordable and compact, and to
provide higher-quality imaging. Ultrasound provides a safe
and effective, dynamic and repeatable form of imaging that
can be performed at the patient bedside, and is free from
the harmful effects of ionising radiation. The combination
of these factors has led to ultrasound becoming increasingly
popular across nearly every speciality of medicine.
Point-of-care ultrasound is defined as ultrasound performed
and interpreted at the bedside and has led to the concept of
the ‘ultrasound stethoscope’.1 Ultrasound education for
non-imaging specialties is now relatively advanced, with
guidelines established by many specialty colleges.2 It is
now being included in the syllabus for many speciality
registrar training schemes and is being considered for
inclusion in undergraduate training in many centres in the
United States, the United Kingdom and Australia.3 Some
American medical schools are even beginning to provide
their students with hand-held ultrasound machines for use
during clinical rotations.4
A formal role for the use of point-of-care ultrasound in the
field of hyperbaric medicine has yet to be clearly established;
however, we see many possibilities for both clinical and
research purposes. Within hyperbaric chambers, ultrasound
transducers have been passed through access ports to study
physiological parameters.5–7 To our knowledge, ultrasound
scanning with a machine inside the chamber has not been
Potential applications of ultrasound in hyperbaric
Ready and immediate access to an ultrasound machine
within a recompression chamber could benefit patients in
a number of ways.
The role of ultrasound in the detection of pneumothoracies
is well established in emergency medicine.8 Divers with
cerebral arterial gas embolism (CAGE) have pulmonary
barotrauma by definition and may have an increased
risk of developing a pneumothorax. If this occurs during
hyperbaric treatment and remains undetected during
ascent, the consequences are potentially catastrophic.
Routine treatment of CAGE involves keeping the patient
supine. For pneumothorax detection, a supine chest X-ray
has a sensitivity ranging from 28% to 75%, whereas lung
ultrasound has a sensitivity ranging from 86% to 98% even
with minimal training.9,10 The absence of the lung sliding
sign, comet tail artefacts and the presence of a contact
point confirms the diagnosis. The study can be successfully
completed within 2–3 minutes.11
The clinical challenge of pneumothorax detection relies on
identifying increased resonance to percussion and reduced
breath sounds on the affected side. Early detection inside a
noisy chamber can be very difficult and the decision to needle
the chest without convincing evidence of pneumothorax is
often difficult. The ability to image at depth with in-chamber
ultrasound would allow detection of supine pneumothoracies
before compression, and, if one developed at depth, would
allow thoracocentesis to be performed when indicated. It
would also allow clinicians to entertain other diagnoses
when pneumothorax had been excluded as a cause for
deterioration at depth.
CRITICAL CARE PATIENTS
Critical care patients inside the chamber pose unique
problems to the hyperbaric physician. Some hyperbaric
facilities run daily hyperbaric oxygen treatments for
intensive care patients. In-chamber ultrasound provides a
useful tool for a wide range of critical care applications.
Pulmonary ultrasonography has been shown to be more
Ultrasound in diving and hyperbaric medicine
Ian C Gawthrope
(Gawthorpe IC. Ultrasound in diving and hyperbaric medicine. Diving and Hyperb Med. 2012;42(1):36-39.)
Ultrasound is a safe and effective imaging modality, the use of which is increasing exponentially in many areas of clinical
medicine. In this article, we present what is, to our knowledge, the first in-chamber use of an ultrasound machine. We discuss
the challenges this presented and how they were addressed, and explore the possible clinical applications that in-chamber
ultrasound may deliver in hyperbaric medicine.
Ultrasound, hyperbaric medicine, equipment, review article
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201237
accurate than auscultation or chest radiography for the
detection of pneumothorax, pleural effusion, consolidation
and alveolar interstitial syndrome in the critical care
setting.12 Cardiac function can easily be assessed with
bedside echocardiography (cardiac ultrasound), and its
use has ‘boomed’ within intensive care.13 The adequacy of
intravascular filling can be accurately assessed by visualising
inferior vena cava (IVC) diameter and determining
respiratory variation.14 Also, as a patient receives fluids, the
changes in IVC parameters can be used to gauge response.
Ultrasound has become the standard of care for procedural
guidance and to confirm intravascular line placement.
The use of ultrasound is well documented in the measurement
of intravascular bubbles.15–18 Echocardiography has been
confirmed as a viable alternative to the traditional aural
Doppler for the assessment of decompression stress.15–17
Equivalent bubble scoring scales between aural bubble
assessment and visual echocardiographic assessment have
been developed and continue to be revised.18 Limited
ultrasound is a simpler skill to learn and more easily
reproducible than aural Doppler.15,16 In-chamber use could
provide us with further understanding of bubble formation
and resolution during treatment.
In-chamber ultrasound provides us with an excellent
research tool to gain further information on diverse
physiological parameters within the hyperbaric environment.
With expertise on hand within the chamber, it alleviates
the difficulties of second-hand image acquisition when
transducers are passed through ports in the chamber.6,7
Selection and testing of an ultrasound device
Our requirements were for a portable ultrasound machine
with good image quality that was suitable for chamber use
at depth, with a range of ultrasound transducers suitable for
echocardiography, abdominal imaging and vascular imaging.
With the assistance of our Biomedical Services, Fremantle
Hospital, we determined what were likely to be the major
issues facing us in our quest to perform ultrasound under
pressure. Key issues identified were:
• Electrical/power supply issues;
• Fire risk;
• Pressure/mechanical damage risk.
With our biomedical colleagues we approached various
ultrasound distributors to discuss the possibility of testing
their machines at depth.
ELECTRICAL/POWER SUPPLY ISSUES
There is little guidance on the testing and modification of
electrical equipment for hyperbaric use. Review articles
report on the use of medical devices under increased
pressure, and basic safety principles and guidelines exist.19–22
However, there are no Australian standards for equipment use
in a high-pressure, oxygen-rich environment. The American
National Fire Protection Association document NFPA 53
contains a recommended practice on materials, equipment
and systems used in oxygen-enriched atmospheres and
there are general recommendations from the European
Committee for Standardisation.23,24 In the absence of
Australian standards, Fremantle Biomedical Services took
these guidelines as a suitable standard for testing.
All the laptop-sized ultrasound machines on the market
currently have a lithium-based battery system in tandem
with a 240-volt mains supply. Lithium batteries have been
shown to overheat under increased pressure and the increased
risk of fire has deemed them unsuitable for chamber use
at depth. Our in-chamber power supply is a filtered direct
current (DC) power of 12 or 24 volts. Of the machines we
tested only one, the Logiq e™, made by GE Healthcare,
was able to function on a 24-volt DC supply; this markedly
narrowed the field.
It was determined that for in-chamber use we would remove
the internal batteries and connect to the 24-volt DC supply.
In changing from the factory supplied alternating current
(AC)/DC power converter to the straight 24-volt DC supply
line, the grounding is lost. This was considered a hazard that
may cause both electric shock and possible sparking and fire
risk. A quick-blow ceramic fuse was therefore installed in
the active line to prevent any such occurrence.
Fire and sparking risk is the most dangerous and likely
hazard in a hyperbaric chamber. To minimise this risk,
temperature of all components needs to be kept low, and
equipment clean, dust free and well ventilated. The NFPA
guidelines specify that the maximum surface temperature of
any component within the chamber is to be limited to 85OC.
Temperature recordings from the service diagnostic tools,
which took around 100 samples during testing, demonstrated
that the central processing unit heated up the fastest. The
maximum temperature recorded was 64OC.
At 24 volts DC, the peak current being drawn was shown
to be 2.13 amps without the probe and 2.5 amps with the
probe. The NFPA guideline recommends that the maximum
power of in-chamber devices is limited to 48 Watts. The
peak power draw from the Logiq e™ is 60 Watts, 12 Watts
greater than that recommended. After due consideration and
with spark proof connectors in place, Biomedical Services
were confident that, with the peak surface temperatures only
reaching 64OC, the unit would run safely at pressure.
Dust can act as a flammable agent and it is important that
potentially hazardous equipment within the chamber stay
dust free. A maintenance plan was drawn up to ensure the
ultrasound console was kept clean and free of dust.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 2012 38
PRESSURE/MECHANICAL DAMAGE RISK
The Logiq e™ contains no sealed regions susceptible to
a pressure difference and the main chassis has two main
airflow paths leading out to vents on either side of the device.
The ultrasonic transducers are completely sealed, which
could lead to problems with pressure difference although it
was noted that transducers had previously been successfully
used when passed through ports into chambers.5–7
THE PROCESS OF INTRODUCTION TO THE
Having addressed all the various concerns outside of the
chamber, we proceeded to introduce the ultrasound machine
to operation at increased pressure in sequential steps.
The ultrasound transducers: The ultrasound transducers,
which contain piezoelectric crystals, were initially tested
alone in the chamber. Image quality and integrity of the
crystals were checked on the surface after the probes had
been sent to increasing pressures up to 405 kPa.
The ultrasound machine: After this assessment and the
required modifications, the laptop ultrasound machine was
certified safe to trial alone in the chamber. The internal
batteries were removed, the unit connected to the 24-volt
DC supply in the chamber, and the transducer held onto a
phantom to provide a visible image through the chamber
porthole (Figure 1). Temperature recordings were further
checked during the unmanned trials within the chamber. The
maximum temperatures did not exceed the 64OC previously
recorded. No new or unexpected issues were encountered.
Maintenance: The machine is to be tested monthly for
preventative maintenance, primarily for removal of dust, a
check of system logs, an electrical safety test and hard disk
Introduction to clinical use: The Biomedical Services
completed a modification report and a user’s instruction
guide. The first manned use of the entire ultrasound machine
was carried out in April 2010. A group of consenting dual-
qualified hyperbaric and emergency physicians went with
the ultrasound machine to 405 kPa. One of the group was
trained in ultrasound and carried out limited examinations as
would be performed clinically within a hyperbaric chamber.
Images were stored for review after the dive. The GE Logiq
e™ ultrasound machine, after modification, provided images
safely to depths up to 405 kPa, with no impairment of image
Since testing, and with no alternatives available, the Logiq
e™ ultrasound machine was purchased and modified for
hyperbaric use. Biomedical Services certified it safe for
manned use within the chamber and further testing on
consenting volunteers was performed without problems.
Ultrasound has now been introduced to clinical work and
a number of the hyperbaric staff trained in its use. As well
as those involved in our research projects, consent is now
sought from all critical care patients to have the ultrasound
in the chamber if required, and we have imaged over 30
patients without problems. All patients or their immediate
family are required to give informed consent to have the
ultrasound machine in the chamber. We have a safe working
procedure and the machine use is carefully monitored by
Biomedical Services as per the agreed protocol. We have
received ethical approval for a number of research studies,
including a formal echocardiography study at depth.
As we have experienced in emergency medicine, the
potential indications for ultrasound in hyperbaric medicine
are expanding rapidly, particularly now we are able to
perform ultrasound at depth. Having said this, it is important
that users understand its limitations and the added safety
aspects of in-chamber use.
In our unit, it has become standard-of-care to ultrasound
the chest of all potential CAGE patients to exclude
pneumothorax prior to treatment. Within the chamber
under pressure, we have found ultrasound to be invaluable
in assessing the fluid resuscitation status of septic patients.
We have witnessed nitrogen bubble resolution inside
the chamber with commercial divers undergoing surface
decompression and now routinely monitor staff for bubble
counts following patient treatments. We have picked up
occult wound collections needing drainage in two patients
undergoing treatment for non-healing wounds, facilitating
The Fremantle Hyperbaric and Diving Medicine Unit, is
currently in the process of finalising plans to move to a new
site and construct a new chamber. At significant extra cost
plain radiography facilities could be provided within the
The Logiq e™ ultrasound set up for
the pressurised chamber trials
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201239
chamber. Whilst this may occasionally be useful, with the
successful advent of in-chamber ultrasound we feel this is
unlikely to add significantly to the point-of-care imaging
we can now perform.
If ultrasound is perceived as a useful addition to our field
and a potential market exists, we may have an opportunity
to work with the manufacturers to produce equipment that
is compatible to our unique environment.
We believe ultrasound will have an important role to play
in hyperbaric medicine and have shown that it can be used
safely and successfully in the hyperbaric environment.
Thanks to Fremantle Hospital Biomedical Services, GE
Healthcare for their support.
1 Moore CL, Copel JA. Point of care ultrasonography. New Engl
J Med. 2011;364:749-57.
Ultrasound training recommendations for medical and surgical
specialities. London: The Royal College of Radiologists, UK;
2005. Available from: http://www.rcr.ac.uk/docs/radiology/
Wong I, Jayatilleke T, Kendall R, Atkinson P. Feasability
of a focused ultrasound training programme for medical
undergraduate students. Clin Teach. 2011;8:3-7.
Rao S, van Holsbeeck L, Musial JL, Parker A, Bouffard JA,
Bridge P, et al. A pilot study of comprehensive ultrasound
education at the Wayne State University School of medicine:
a pioneer year review. J Ultrasound Med. 2008;27:745-9.
Marabotti C, Belardinelli A, L’Abbate A, Scalzini A, Chiesa
F, Cialoni D, et al. Cardiac function during breath-hold diving
in humans: An echocardiographic study. Undersea Hyperb
Molenat F, Boussuges A, Grandfond A, Rostain JC, Sainty
JM, Robinet C, et al. Haemodynamic effects of hyperbaric
hyperoxia in healthy volunteers: an echocardiographic and
Doppler study. Clin Sci. 2004;106:389-95.
Lafay V, Boussuges A, Ambrosi P, Barthelemy P, Frances Y,
Gardette B, et al. Doppler- echocardiography study of cardiac
function during a 36 atm (3,650 kPa) human dive. Undersea
Hyperb Med. 1997;24:67-71.
Blaivas M, Lyon M, Duggal S. A prospective comparison
of supine chest radiography and bedside ultrasound for the
diagnosis of traumatic pneumothorax. Acad Emerg Med.
Wilkerson R, Stone M. Sensitivity of bedside ultrasound and
supine anteroposterior chest radiographs for the identification
of pneumothorax after blunt trauma. Acad Emerg Med.
10 Wernecke K, Galanski M, Peters PE, Hansen J. Pneumothorax:
evaluation by ultrasound. J Thoracic Imaging. 1987;2:76-8.
11 Dulchavsky SA, Schwarz KL, Kirkpatrick AW, Billica RD,
Williams RD, Diebel LN, et al. Prospective evaluation of
thoracic ultrasound in the detection of pneumothorax. J
12 Lichtenstein D, Goldstein I, Mourgeon E, Cluzel P, Greiner
P, Rouby JJ. Comparative diagnostic performances of
auscultation, chest radiography, and lung ultrasonography
in acute respiratory distress syndrome. Anaesthesiology.
13 Schmidt GA. ICU Ultrasound: The coming boom. Chest.
14 Barbier C, Loubieres Y, Schmit C, Haydon J, Ricome JL, Jardin
F, et al. Respiratory changes in inferior vena cava diameter are
helpful in predicting fluid responsiveness in ventilated septic
patients. Intensive Care Med. 2004;30:1740-6.
15 Eftedal O, Brubakk AO. Agreement between trained and
untrained observers in grading intravascular bubble signals in
ultrasonic images. Undersea Hyperb Med. 1997;24:293-9.
16 Brubakk AO, Eftedal O. Comparison of three different
ultrasonic methods for quantification of intravascular gas
bubbles. Undersea Hyperb Med. 2001;28:131-6.
17 Eftedal, OS, Lydersen, S, Brubakk AO. The relationship
between venous gas bubbles and adverse effects of
decompression after air dives. Undersea Hyperb Med.
18 Pollock NW. Use of ultrasound in decompression research.
Diving Hyperb Med. 2007;37:68-72.
19 Kot J. Medical equipment for multiplace hyperbaric
chambers. Part 1: Devices for monitoring and cardiac support.
European Journal of Underwater and Hyperbaric Medicine.
20 Kot J. Medical equipment for multiplace hyperbaric chambers.
Part 2: Ventilators. European Journal of Underwater and
Hyperbaric Medicine. 2006;7(1):9-12.
21 Kot J. Medical equipment for multiplace hyperbaric chambers.
Part 3: Infusion pumps and syringes. European Journal of
Underwater and Hyperbaric Medicine. 2006;7(2):29-31.
22 Burman F, Sheffield R, Posey K. Decision process to assess
medical equipment for hyperbaric use. Undersea Hyperb
23 NFPA 53: Recommended practice on materials, equipment,
and systems used in oxygen-enriched atmospheres. Quincy,
MA: National Fire Protection Association; 2004.
24 Pressure vessels for human occupancy (PVHO) - Multi-place
pressure chamber systems for hyperbaric therapy- Performance,
safety requirements and testing. prEN14931:2004. Brussels:
European Committee for Standardisation (CEN); 2004.
Submitted: 10 April 2011
Accepted: 04 November 2011
Ian C Gawthrope, BM, DCH, FACEM, is a specialist in
Emergency and Hyperbaric Medicine at Fremantle Hospital,
Address for correspondence:
Dr Ian Gawthrope
Fremantle Hyperbaric Unit
PO BOX 480, WA 6160, Australia
Phone: +61-(0) 8-9431-2233
Fax: +61-(0) 8-9431-2235
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201240
Scuba divers’ pulmonary oedema: recurrences and fatalities
Carl Edmonds, John Lippmann, Sarah Lockley and Darren Wolfers
(Edmonds C, Lippmann J, Lockley S, Wolfers D. Scuba divers’ pulmonary oedema: recurrences and fatalities. Diving
Hyperb Med. 2012;42(1):40-44.)
Scuba divers’ pulmonary oedema (SDPE) is an increasingly recognised disorder in divers. We report three fatal cases
of SDPE, demonstrating its potentially serious nature even in the absence of underlying cardiac disease demonstrable
clinically or at autopsy. This, together with the frequency of recurrences, has implications on assessing fitness for
subsequent diving, snorkelling and swimming. The differential diagnosis of this disorder is also considered, as is its
possible inducement by salt water aspiration and its relationship to drowning.
Scuba diving, pulmonary oedema, salt water aspiration, deaths, case reports
Scuba divers’ pulmonary oedema (SDPE) was first recorded
in 1981.1 Comprehensive reviews have been prepared since
by various authors.2–6 In these reviews, the physiological
bases of the disorder have been canvassed; it is one type of
immersion pulmonary oedema (IPE). SDPE presents with
scuba divers developing fast shallow respirations, dyspnoea,
fatigue, cough and white or sometimes blood-stained
frothy expectoration. The signs include hypoxia and the
auscultatory evidence of pulmonary oedema. Investigations
reveal impaired spirometry and reduced lung compliance,
hypoxaemia and characteristic radiological (plain chest
X-ray or CT scan) abnormalities.
SDPE is said to be more frequent in older divers and those
with cardiovascular pathology.1–6 It tends to recur in at least
30% of cases.5 Exertion during the dive is often not excessive
and frequently the condition becomes more evident during
ascent or surface swimming. Spontaneous resolution is often
prompt after leaving the water. Only one death has been
reported in the traditional medical literature and this was
based on significant pre-existing cardiac pathology.5 The
latter is characteristic of some of these SDPE cases and is one
aetiological feature that may be amenable to correction.
These three case histories illustrate the difficulty in
predicting the development of non-cardiac based SDPE, the
significance of recurrences and the possibility of death from
this disorder. They have implications regarding appropriate
advice that is given to affected divers.
Case history 1
Incident 1: A 51-year-old female nurse had no significant
past medical history other than a mild allergic diathesis
in early life, presenting with eczema and hay fever. She
was an experienced scuba diver, logging over 900 dives
without incident and possessing open-water and deep-
diving qualifications. She was considered a conservative but
enthusiastic club diver. The day before the incident she had
completed two non-decompression, computer-assisted dives
in an area well known to her. The first was to 24 metres’ sea
water (msw) for 50 minutes, followed by a surface interval of
three hours; the second to 7 msw for 10 minutes, aborted due
to currents and poor visibility. That afternoon and night she
consumed 70 grams of alcohol, together with other fluids.
The following day, she felt well, although a little fatigued.
At 0800 h she commenced a dive profile that she had
undertaken on other occasions without difficulty. This
involved a 30-metre surface swim, fully equipped but finning
on her back and with the regulator out. The conditions
were described as perfect, and the current was considered
“moderate at the worst”. Although she reported that she
did not experience any aspiration, she did state that the
wash from a boat splashed over her head once, causing her
to cough and swallow some sea water. Later, during this
four-minute swim, she became dyspnoeic. Her companion
observed that it was a “tough swim” and that her lips
appeared cyanotic and her breathing rate was rapid during the
minute she spent resting on the marker buoy. In subsequent
interrogations, she denied any salt water aspiration, chest
discomfort, palpitations or syncopal sensation at that time.
Because they thought there could be less current at depth,
they commenced the dive but only reached about 12 msw
in one to two minutes. They aborted the dive after three
minutes, due to her progressive dyspnoea and feeling
fatigued. They ascended slowly, over about five minutes,
before surfacing near the shore. She was then assisted in
walking and removing her equipment.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201241
Her coughing was frequent with expectoration initially
whitish but becoming pink and frothy and she was aware of
fluid rattling in her chest. She was dyspnoeic and cyanotic,
with a grey appearance. She improved somewhat over the
next quarter of an hour and was then able to walk unassisted.
Ambulance paramedics administered high-concentration
oxygen, until the medevac helicopter arrived. In telephone
discussion with the DAN diving emergency service (DES),
the clinician heard her wheezing and noted her complaints of
dyspnoea and a “rattling” in her chest. She was transferred
to the metropolitan hospital, breathing oxygen administered
via a simple face mask.
Her vital signs on admission at 1045 h were not grossly
abnormal, with a heart rate of 100 beats per minute and
a respiratory rate of 24 breaths per minute, but she still
had a persistent, non-productive cough with wheezing
and crepitations at both lung bases. She was continued on
oxygen and bronchodilators were administered. The chest
X-ray showed minor linear basal densities, more on the right,
consistent with interstitial oedema. All other investigations
(ECG, lung function, electrolytes, biochemistry, liver
function, oxygen saturation) were normal. The respiratory
difficulty settled by 1500 hours and she was discharged
the following day, for later review. Then, her lung function
tests showed improvements of 18% in forced vital
capacity, increasing to 26% following administration of a
bronchodilator. The original impairment was considered to
be consistent with increased airway reactivity associated
with lung damage. A mild neutrophil leucocytosis was
similarly explained. There were no other symptoms or
signs suggestive of decompression sickness or pulmonary
barotrauma and the dive profile was not indicative of these
A month later, a specialist cardiologist consultation included
clinical assessment, ECG, stress testing and transthoracic
echocardiograms, without any abnormality being detected.
He concluded that the episode of pulmonary oedema was
non-cardiogenic and that the patient had normal cardiac
function. Repeat lung function testing at the same laboratory
showed normal lung values and an asthma provocation test
was negative. There was an improvement in lung volumes
compared to the previous tests.
Her enthusiasm to return to diving and to re-establish her
DAN diving insurance for future overseas diving trips led
to consultations with at least six diving medical specialists.
The diagnoses were divided between SDPE and the salt
water aspiration syndrome (SWAS), and advice varied from
unfitness for any diving (snorkel or scuba) to approval for
unrestricted diving. She considered the conflicting advice
available and also attempted her own research on this subject,
and then resumed diving.
Incident 2: Almost a year later, now with another 54 logged
dives, and with no further medical history apart from the
incident above, she died whilst diving. She was participating
in a night dive from shore. There was a moderate wind and
the surface was choppy. Surface water temperature was
about 22OC reducing to 19OC at depth and was described as
comfortable. She was wearing a semi-dry suit.
The victim was with three others, in two buddy pairs. They
swam on the surface for about 30 metres before descending
and working along the sloping bottom to a maximum depth
of 18 msw. For most of the dive the victim appeared to be fine
and responded affirmatively to the buddy’s regular ‘OK?’
signals. However, after about 25 minutes, at a depth of 14
msw, she signalled that she was ‘not OK’. They decided to
return and they swam underwater up the slope and towards
the shore. Each time the buddy enquired if she was OK she
responded in the negative. On reaching a depth of 7 msw, the
buddy held her hand and they slowly ascended and surfaced
in a sheltered area, with a dive time of 37 minutes.
At the surface, she vomited a brown, lumpy liquid. She was
trying to cough and had an audible wheeze. She stated faintly
that she could not breathe and she continued to vomit. Her
BCD was inflated and she rolled over onto her back as the
buddy towed her towards the shore. The buddy could hear her
wheezing and struggling to breathe. She was still conscious
and complained that she could not breathe, but tried to kick
her legs to assist the buddy towing her. The buddy towed
her approximately 100 metres to thigh-deep water beside
rocks. She was assisted onto the rocks. It was believed that
she did not inhale any water during the rescue.
She then became unconscious and apnoeic, and her buddy
commenced basic life support. This produced regurgitation
of stomach contents and some bloody sputum. Others
assisted until the paramedics arrived about 15 minutes later.
They implemented advanced life support but she failed to
At autopsy the lungs were oedematous, weighing over 1.4
kg, and did not appear unduly hyperexpanded. There was
no pathological evidence to indicate other causes of death,
including previous or recent cardiovascular disease. The heart
weighed 310 g. Toxicology was negative. The pathological
diagnosis of acute pulmonary oedema was made.
Case history 2
This 45-year-old woman was apparently healthy and had
become certified as an Open Water Diver one week earlier,
having completed four open-water training dives. She was
then participating in an Advanced Open Water course and
had completed three uneventful dives on the previous day
to a maximum depth of 7 msw, with a surface interval of 8
hours between the last two dives.
On the day, the water was calm and clear with visibility of
10–15 msw, and the dive was at slack water. The victim was
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201242
with a group of six students, accompanied by an instructor
and a divemaster. They descended to a depth of 26 msw
and knelt on the sea bed while answering questions on a
slate. The duration of the dive was 16 minutes and she had
completed other narcosis tasks. The victim then gave a low-
on-air hand signal. The instructor noted that her contents
gauge read 120 bar and gave her his ‘octopus’ regulator to
breathe on briefly while he breathed on her demand valve,
to check that it was ‘OK’; it appeared to be functioning
normally. She then took back her own regulator. However, a
short time later, she again signalled she was low on air before
starting to ascend. The instructor indicated to the others to
remain on the sea bed with the divemaster and caught hold
of the victim by her buoyancy compensator. They then
ascended together while using his buoyancy to control their
ascent rate. Soon after departure he noticed she seemed to
be having some difficulty with her breathing, taking rapid,
short, shallow breaths. However, she refused the offer of
his secondary regulator. She then ceased to respond to his
signals. The ascent was described as controlled and at a rate
of around 15 msw per minute. On surfacing, the instructor
asked if she was ‘OK’ to which she replied “No, I don’t feel
good” before rolling onto her side, unconscious. Shortly
afterwards, white froth began to flow from her mouth.
The instructor then towed the victim some 30 metres to
shore, intermittently providing rescue breaths, despite the
continued flow of frothy sputum. Another diver assisted the
victim onto the shore where she was assessed as unconscious
and apnoeic. Basic life support was commenced and was
complicated by vomitus, water, bile and froth obstructing
her airway. After about ten minutes, another diver arrived
with an automated external defibrillator which indicated
that no shock be given. At this time, the victim had fixed,
Paramedics arrived soon after and commenced advanced
life support. A shockable cardiac rhythm was briefly created
although subsequent defibrillation failed to restore sinus
rhythm. There was continued difficulty ventilating the victim
as the airway appeared to be obstructed by fluid.
An equipment check on the beach showed the remaining air
at 90 bar. Examination of her equipment by the police diving
branch subsequently showed no abnormality in equipment
or gas, except for the hose to her primary regulator. This was
kinked (longstanding) and this kink may have restricted the
air flow. However, a subsequent test dive with the equipment
failed to elicit this restriction, despite using various activities,
positions and depths up to 29 msw.
The victim had passed a fit-to-dive medical but had omitted
to mention that she had taken dexamphetamine (25–30 mg
daily) for adult onset attention deficit hyperactivity disorder
and also suffered from migraine, though rarely. She may
have discontinued this medication before diving as no drugs
were detected by toxicology at autopsy.
Autopsy X-ray two days after death showed generalized
air distribution throughout the body and all the vascular
system. This was attributed to post-mortem decompression
artifact possibly aggravated by the resuscitative attempts.
She was slightly overweight (height 176 cm; weight 84 kg;
BMI 27). The heart weighed 360 g and was normal with
minor degrees of atheroma and up to 20% narrowing of the
coronary arteries. No evidence of infarction or fibrosis was
seen, but there was fine patchy replacement fibrosis in the
heart on histology, which is not explained. The right and left
lungs weighed 915 g and 740 g respectively and were well-
expanded and the parenchyma showed extensive pulmonary
oedema but no congestion. There were gastric contents in
the upper airways.
The pathological diagnosis of acute pulmonary oedema
was made. As the symptoms commenced and progressed
at maximum depth and as there was no preceding ascent,
both decompression sickness and pulmonary barotrauma
diagnoses were dismissed.
Case history 3
Another death was mentioned as an unreferenced addition in
a previous review of SDPE.3 This case probably originated
from a DAN report of a fatality in 1996.7 This was followed
up with the original source and the following information
A 51-year-old experienced, female diver undertook an
uneventful, short, shallow dive with her husband. On
surfacing she became dyspnoeic. She was towed with her
buoyancy compensator inflated and allegedly with her head
above water. She was then brought on board the diving boat
where she lost consciousness and died despite resuscitation
efforts. Autopsy revealed no evidence of pulmonary
barotrauma, air embolism or decompression sickness. The
lungs were extremely oedematous and frothy pink fluid filled
the airways. There was some evidence of arteriosclerosis –
the left anterior descending coronary artery had a stenosis of
over 50% – but the coronaries were still patent. There was
no evidence of previous or recent cardiac disease.
The pathological diagnosis of acute pulmonary oedema
Pons et al described SDPE as a rare event in healthy
individuals.8 The actual incidence is unknown, but it is likely
to be under-diagnosed.3–6,8 Deaths from SDPE are probably
under-reported because the disease is not a high profile one
(even amongst diving clinicians) and pathological findings
are similar to those of drowning.9,10 The latter diagnosis
is often the default one for those who die in the ocean and
have heavy, fluid-filled lungs. Differentiating drowning from
SDPE pathology is a complex and questionable procedure,
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201243
not achieved at most autopsies. Also, a diver incapacitated by
acute pulmonary oedema is then susceptible to superimposed
water aspiration, with drowning obliterating the original
pathology. The identification/distribution of diatoms is
unlikely to be of value, as both can occur with immersion
deaths. There is no single pathognomonic discriminator. It
is possible that emphysema aquosum may be more typical
of drowning pathology, but its aetiology is presumed to be
associated with bronchoconstriction and this occurs also
Recurrences of SDPE have been reported in up to 30% of
cases. This is likely to be a considerable underestimate of the
actual risk, as treating clinicians usually do not perform long-
term reviews on successfully treated cases. Also, contact may
not be possible with this itinerant group and some divers
affected by SDPE may avoid the risk of a recurrence by
avoiding exposure to the cause – scuba diving or snorkelling.
Recurrences may occur in both surface swimming and diving
activities; the real recurrence rate is unknown.
The one death from SDPE that has been reported in
the traditional medical literature was associated with
significant cardiac pathology – in a diver with hypertension,
hyperlipidaemia and arteriopathy and who sustained a
cardiac arrest whilst swimming back to shore. He died
72 hours later from cerebral oedema.5 He had suffered a
SDPE episode that had been well documented, eight months
previously. The problems of cardiac-based SDPE have
already been canvassed and warnings given regarding the
risk of subsequent immersion and diving.6
Other causes of pulmonary oedema that may occur with
scuba diving should be considered in the differential
diagnosis of SDPE. These include existing cardiac disease
and diving- or immersion-induced diseases, e.g., salt water
aspiration and the drowning syndromes, gas-induced
pulmonary toxicity, dysbaric lung disease and pulmonary
decompression sickness. Certain marine envenomations,
especially the Irukandji syndrome, cold urticaria, asthma and
other medical disorders may produce or simulate pulmonary
oedema and be aggravated by the diving environment and
Most differential diagnoses to explain the initial incident
in Case 1 had been excluded by the dive profile or by
subsequent medical assessments and investigations. The
remaining differential diagnosis is what has been termed the
salt water aspiration syndrome (SWAS), which is described
in detail elsewhere.11 Distinguishing between SDPE and
SWAS is a difficult diagnostic conundrum. It is possible
that sea water aspiration may precede or even induce the
development of SDPE in some cases (as may be so in Case 1)
by damaging pulmonary capillaries and then exposing them
to the increased negative inspiratory pressures experienced
with scuba diving, snorkelling and immersion.
SWAS has many clinical features similar to SDPE.12,13 The
dyspnoea, cough and expectoration are common to both,
as are reduced lung volumes, arterial hypoxia and rapidly
changing radiological signs in the lungs. The clinical
manifestations of SWAS, such as fever and rigors, nausea,
headache, muscular pain and mild leucocytosis are probably
due to the combination of the lung pathology of aspiration
and associated cold exposure, in the original series. The main
differentiation, clinically, is that SWAS tends to develop soon
after the dive whereas SDPE develops during the immersion,
and is aggravated with the ascent.
Cases of both SDPE and SWAS have a rapid improvement
with oxygen supplementation, and so the initial rescue from
the water and conventional diver first aid treatments are
applicable to both.
Subsequent management of the SDPE cases is hampered by
the relatively few case histories documented. The medical
advice to be given to victims of SDPE, even those without
cardiac pathology, should probably be based on the high
risk of recurrences, the possibility of death and our failure
to clarify what environmental conditions, apart from
immersion, precipitate the event.
SDPE is a serious illness amongst scuba divers. It tends to
recur, even without known predisposing factors (other than
age and immersion). Cardiac pathology may be influential in
some cases and salt water aspiration in others. However, it is
potentially lethal even in those without pre-existing clinical
or demonstrable cardiac disease and without significant
cardiac pathology, as detected at autopsy.
We present, for the first time to our knowledge, evidence
of fatal consequences of SDPE without any significant
demonstrable cardiovascular pathology.
Advice against further immersion (e.g., snorkelling, scuba
diving) exposure in those victims who survive the first
episode, is probably warranted. The illness and fatality rates
are not known, but are probably underestimated in the diving
We thank the Divers Alert Network, DAN Asia-Pacific and
DAN America and Drs Stephen Wills and Chris Lawrence,
1 Wilmshurst P, Nuri M, Crowther A, Betts I, Webb-Peploe MM.
Forearm vascular response in subjects who develop recurrent
pulmonary oedema when scuba diving: a new syndrome. Br
Heart J. 1981;45:349.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201244
2 Lundgren CE, Miller JN. The lung at depth. Lung biology
in health and disease. Vol 132. New York: Marcel Dekker
Slade JB, Hattori T, Ray CS, Bove AA, Cianci P. Pulmonary
oedema associated with scuba diving. Chest. 2001;120:1686-
Koehle MS, Lepawsky M, McKenzie DC. Pulmonary oedema
of immersion. Sports Med. 2005;35:183-90.
Cochard G, Arvieux J, Lacour J-M, Madouas G, Mongredien
H, Arvieux CC. Pulmonary edema in scuba divers. Recurrence
and fatal outcome. Undersea Hyperb Med. 2005;32:39-44.
Edmonds C. Scuba divers’ pulmonary oedema – a review.
Diving Hyperb Med. 2009;39:226-31.
Diver Alert Network. Report on decompression illness and
diving fatalities. 1998 Edition. Durham, NC: Divers Alert
Network; 2000. Record No: 3696. p. 101.
Pons M, Blickenstorfer D, Oechslin E, Hold G, Greminger
P. Pulmonary oedema in healthy persons during scuba diving
and swimming. Eur Respir. 1995;8:762-7.
Christe A, Aghayev E, Jackowski C, Thali MJ, Vock P.
Drowning – post–mortem imaging findings by computed
tomography. Eur Radiol. 2008;18:283-90.
10 Lorin de la Grandmaison G, Paraire F. Place of pathology
in the forensic diagnosis of drowning. Ann Pathol. 2003;
11 Edmonds C, Lowry C, Pennefather J, Walker R. Diving and
subaquatic medicine. 4th ed. London: Arnold; 2002.
12 Coulange M, Rossi P, Gargne O, Gole Y, Bessereau J, Regnard
J, et al. Pulmonary oedema in healthy SCUBA divers: new
physiopathological pathways. Clin Physiol Funct Imaging.
13 Edmonds C. A salt water aspiration syndrome. Mil Med.
Submitted: 13 November 2011
Accepted: 11 January 2012
Carl Edmonds, MB, BS, MRCP(Lond), DPM, MRCPsych,
FRANZCP, FRACP, DipDHM, FAFOM, is a consultant in
diving medicine, Sydney, Australia.
John Lippmann, BSc, Dip Ed, MAppSc, is Executive Director,
Divers Alert Network (DAN) Asia-Pacific.
Sarah Lockley, BMedSc, BMed(Hons), FRACGP, is a Royal
Australian Naval Reserve Medical Officer, MLC Medical
Darren Wolfers, LM, MBBS(Hons), FANZCA, Cert DHM
(ANZCA) is a consultant in the Department of Diving and
Hyperbaric Medicine, Prince of Wales Hospital, Randwick,
Address for correspondence:
11/69-74 North Steyne
Manly, NSW 2095
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201245
Hyperbaric oxygen therapy did not improve arm volume or functional
scores in post-radiation lymphoedema
• No evidence of a clinically significant reduction in arm
volume or functional scores in lymphoedema following
radiotherapy one year after hyperbaric oxygen therapy
• Some non-significant indication of improved function
scores at six months after HBOT.
Gothard L, Haviland J, Bryson P, Laden G, Glover M,
Harrison S, et al. Randomised phase II trial of hyperbaric
oxygen therapy in patients with chronic arm lymphoedema
after radiotherapy for cancer. Radiother Oncol. 2010;
Lead author’s name and e-mail: <john.yarnold@icr.
Three-part clinical question:
For patients with lymphoedema in the upper limb following
axillary or supraclavicular radiotherapy, does the application
of HBOT improve arm volume?
Hyperbaric oxygen, lymphoedema, radiotherapy, breast
Non-blinded randomised controlled trial without intention-
to-treat. 2:1 randomisation schedule.
The study patients:
Women previously irradiated in the axilla or supraclavicalur
area and who have developed lymphoedema in the arm
resistant to standard therapy and with increased arm volume
of at least 15%.
Control group (n = 20; 16 analysed):
Best standard lymphoedema care according to a 2006
international consensus; no sham hyperbaric therapy.
Experimental group (n = 38; 30 analysed):
Best care as above plus daily HBOT at 243 kPa for 90
minutes to 30 treatments over six weeks.
See Tables 1 and 2
1 High dropout rate reduces our confidence in these figures.
Authors were unable to enroll sufficient patients
to satisfy their power calculations and this study is
2 There was some indication of benefit in functional
lymphoedema scores at six months but no significance
testing was reported.
3 Average interval from onset to therapy was 12 years;
this may have biased against a treatment effect.
Michael Bennett, 18 November 2010
Outcome at 1 year
>8% change in volume
Control group HBO group
-80% to 196%
-0.12 to 0. 29
0.150 8 to 3 to
Arm volume at one year
Functional lymphoedema scores at six months and one year (significance values not given)
Self-assessment lymphoedema quesionnaire
(0 best to 100 worst) median and IQR
Time to outcome
47.9 (18.7 to 64.1)
32.3 (17.7 to 53.6)
12 months 45.8 (13.0 to 62.5) 37.5 (20.8 to 52.1) ?
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201246
Continuing professional development
Hyperbaric oxygen therapy for delayed
The intended audience consists of all physicians subscribing
to Diving and Hyperbaric Medicine (DHM), including
anaesthetists and other specialists who are members of
the Australia and New Zealand College of Anaesthetists
(ANZCA) Diving and Hyperbaric Medicine Special Interest
Group (DHM SIG). However, all subscribers to DHM may
apply to their respective CPD programme coordinator or
specialty college for approval of participation. This activity,
published in association with DHM, is accredited by the
ANZCA Continuing Professional Development Programme
for members of the ANZCA DHM SIG under Learning
Projects: Category 2/Level 2: 2 credits per hour. ANZCA
Fellows may only claim for this provided they submit their
answers to the CPD coordinator.
The questions are designed to affirm the takers’ knowledge
of the topics covered, and participants should be able to
evaluate the appropriateness of the clinical information as
it applies to the provision of patient care.
Authors of these activities are required to disclose activities
and relationships that, if known to others, might be viewed
as a conflict of interest. Any such author disclosures will be
published with each relevant CPD activity.
Do I have to pay?
All activities are free to subscribers.
Practitioners are referred to the following background
references and reading:
Frequency of radiated patients:
2 Staffurth J. Radiotherapy Development Board: a review of the
clinical evidence for intensity-modulated radiotherapy. Clin
Oncol (R Coll Radiol). 2010;22:643-57.
3 Stone HB, Coleman CN, Anscher MS, McBride WH. Effects
of radiation on normal tissue: consequences and mechanisms.
Lancet Oncol. 2003;4:529-36. [Review]
Hyperbaric oxygen for the treatment of radiation injury:
4 Feldmeier JJ, Hampson NB. A systematic review of the
literature reporting the application of hyperbaric oxygen
prevention and treatment of delayed radiation injuries:
an evidence based approach. Undersea Hyperb Med.
Clarke RE, Tenorio LMC, Hussey JR, Toklu AS, Cone DL,
Hinoiosa JG, et al. Hyperbaric oxygen treatment of chronic
refractory radiation proctitis: a randomized and controlled
double-blind crossover trial with long-term follow-up. Int J
Radiat Oncol Biol Phys. 2008;72:134-43.
Craighead P, Shea-Budgell MA, Nation J, Esmail R, Evans
AW, Parliament M, et al. Hyperbaric oxygen therapy for late
radiation tissue injury in gynecologic malignancies. Curr
Sidik S, Hardjodisastro D, Setiabudy R, Gondowiardjo S.
Does hyperbaric oxygen administration decrease side effects
and improve quality of life after pelvic radiation? Acta Med
Cancer and hyperbaric oxygen:
8 Granowitz EV, Tonomura N, Benson RM, Katz DM, Band V,
Makari-Judson GP, et al. Hyperbaric oxygen inhibits benign
and malignant human mammary epithelial cell proliferation.
Anticancer Res. 2005;25(6B):3833-42.
Cell production in relation to hyperbaric oxygenation:
9 Thom SR, Bhopale VM, Velazquez OC, Goldstein LJ, Thom
LH, Buerk DG.Stem cell mobilization by hyperbaric oxygen.
Am J Physiol Heart Circ Physiol. 2006;290:H1378-86.
How to answer the questions
Please answer all responses (A to E or F) as True or False.
Answers should be posted by e–mail to the nominated CPD
For EUBS members for this CPD issue this will be Dr Erik
Jansen, e-mail: <email@example.com>.
For ANZCA DHM SIG members and SPUMS members,
this will be Dr David Cooper, e-mail: <david.cooper@
On submission of your answers, you will receive a set
of correct answers with a brief explanation of why each
response is correct or incorrect.
Successfully undertaking the activity will require a correct
response rate of 80% or more. Each task will expire within
24 months of its publication to ensure that additional, more
recent data has not superceded the activity.
MOPS (maintenance of professional standards), radiotherapy,
hyperbaric oxygen therapy, hyperbaric medicine, bone
necrosis, soft-tissue radionecrosis
CME ACTIVITY 2011/1
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201247
Question 1. Radiation therapy of malignant tissue has the
A. In industrialised countries, about half of all cancer
patients receive some type of radiation therapy sometime
during the course of their treatment.
B. Delayed radiation injury is estimated to affect 5–15%
of long-term survivors.
C. The risk of getting delayed radiation injury is limited
to the first year after radiation therapy.
D. The acute reaction to radiation is always a strong
indicator for development of delayed radiation injury.
E. Only the malignant tissue will react with delayed
F. Intensity-modulated radiotherapy ( IMRT )increases
delayed radiation injuries.
Question 2. The radiation injury:
A. Vascular injury is not an effect of radiation injury.
B. Cellular depletion and tissue fibrosis are important
components of delayed radiation injury.
C. Hypoxia may occur due to delayed radiation injury.
D. Reliable assays are available to accurately identify
patients who are at high risk of developing delayed
E. The delayed radiation injury may be precipitated by
further injury or trauma.
F. Delayed radiation injury resolves over time without
Question 3. Osteoradionecrosis, delayed radiation injury
of the jaw:
A. In the treatment of osteoradionecrosis of the jaw, it
is essential to includes surgical extirpation of necrotic
B. When surgery is required the majority of hyperbaric
oxygen treatments should be given after the surgical
C. The risk of osteoradionecrosis is increased due to
impaired saliva production and the increased frequency
of dental caries.
D. Osteoradionecrosis of the jaw does not affect the
nutrition of the patient.
E. It is generally agreed that hyperbaric oxygen treatment
is always indicated as a prophylaxis before extractions
and other oral surgical procedures in radiation patients.
Question 4. Radiation injury of abdominal organs:
A. It is possible to find a recent randomised controlled
study which demonstrates the effect of hyperbaric oxygen
treatment on radiation proctitis.
B. Common symptoms of delayed radiation injury of the
pelvic region include increased stool frequency, urgency,
spotting of blood and partial incontinence.
C. Hyperbaric oxygen administration decreases delayed
radiation injury effect and improve quality of life after
D. Hyperbaric oxygen treatment is indicated in the
treatment of radiation cystitis when conventional
treatment of instillation of alum or formalin is not
E. If all studies on radiation injury of abdominal organs
are pooled, 87% of the patients have a successful or partly
successful result of hyperbaric oxygen treatment.
Question 5. Hyperbaric oxygen and cell proliferation:
A. Hyperbaric oxygen enhances cancer growth or
recurrence in humans.
B. Hyperbaric oxygen inhibits in vitro growth of human
mammary transplanted tumor.
C. Hyperbaric environment increases the number of stem
cells in humans.
D. Hyperbaric oxygen environment increases the number
of stem cells in humans.
E. Hyperbaric oxygen increases angiogenesis by release
of vascular endothelial growth factor.
The database of randomised controlled trials in hyperbaric medicine maintained by
Michael Bennett and his colleagues at the Prince of Wales Hospital
Diving and Hyperbaric Medicine Unit, Sydney is at:
Professor Bennett advises that this site is being reconstructed and brought up to date currently. For the past two years, the
Prince of Wales team have been heavily involved in the installation of their new three-compartment rectangular chamber,
and the complete rebuild of the department, and apologise that this website has not been in their focus during this time.
Diving and Hyperbaric Medicine Volume 42 No. 1 March 201248 Download full-text
Mastering rebreathers, 2nd
Softcover, full colour, 704 pages
Best Publishing Company, Flagstaff AZ, 2010
Available from: <www.bestpub.com>
There is little doubt that one of the most significant
developments in recreational diving over the last 15 years
has been ‘technical diving’ methods. These diving methods,
previously the province of military and some occupational
diving groups, extend both the depths and durations of
dives. They have opened up an entirely new world of
exploration and underwater experience, and the appeal is
No single technique epitomises technical diving like the use
of a rebreather. These devices recycle exhaled gas through a
carbon dioxide (CO2) absorbent, and maintain safe oxygen
levels in the recycled gas by various means, depending on
type. This minimises consumption of expensive inert gases
such as helium, which is used during deep diving to minimise
nitrogen narcosis. To all intents and purposes (depending
on the type of rebreather), gas consumption may be limited
to as little as the oxygen that the diver metabolises, giving
extremely long underwater durations from a small gas
supply. Unlike open-circuit scuba, this duration is unaffected
by depth. This makes rebreathers the ultimate deep-diving
However, rebreathers are complex and they have many
failure modes. Perhaps not surprisingly, their use appears
to be significantly more hazardous than use of open-circuit
scuba. There is probably no single answer to mitigating
this risk, but one strategy believed by many (including
this writer) to be crucial is high standards of training and
education for rebreather divers.
Enter Mastering rebreathers (2nd edition). Jeff Bozanic is a
well-known and respected member of the diving community
who published the first edition of this book in 2002. The
range of rebreather technologies and models has moved
on considerably in the 10 years since then, and it was
appropriate that the publication be updated or risk becoming
irrelevant. Bozanic has a PhD in education and is a very
experienced rebreather diver; an ideal combination of skills
for an undertaking of this nature. Most importantly, he is
an experienced instructor on multiple models of rebreather,
and thus has considerable insight into those areas of relevant
knowledge that are difficult to impart to students. This is
reflected in the style of the book, which is fundamentally a
textbook for the novice rebreather diver.
Though substantial at 700 pages, virtually half of it is
given to appendices. It is organised into 14 chapters and
the aforementioned appendices. The first seven chapters
could be described as providing background information.
These are entitled ‘Introduction to rebreathers’, ‘History
of rebreathers’, ‘Types of rebreathers’, ‘Diving physics’,
‘Physiology’, ‘Theory’ and ‘Rebreather design’. Chapters
8–10 detail the approach to an actual rebreather dive and
are entitled ‘Preparing for the dive’, ‘Diving techniques’
and ‘Post-dive procedures’. Chapter 12, which covers
‘Emergency procedures’ belongs in this group also. The
chaptered section is rounded out by Chapter 11 ‘Long-term
maintenance’, and Chapters 13 and 14 entitled ‘Travel’ and
‘Where do you go from here?’ respectively. The second half
of the book consists of 20 appendices which, other than one
covering ‘Dive tables’, a glossary and an index, are given
over to aspects of the procedures (including checklists and
maintenance schedules in some cases) for using a wide range
of different brands of rebreather. Some of these sections are
quite comprehensive and others less so because the material
reflects the content that the respective manufacturers were
inclined to provide. There is a wealth of information in these
appendices, and those who enjoy possessing knowledge
about a variety of rebreathers they do not own will find
The first thing that should be made clear is that this is not
a manual for advanced rebreather diving. It does not cover
deep diving, mixed-gas diving, decompression diving, and
diving in special environments (such as caves). These are
subjects that Bozanic intends to address in a second volume
whose release is apparently not far off. This is not a criticism.
Indeed, this reviewer applauds the stated aim of the book
which is to “discuss introductory rebreather principles and
introduce readers to rebreathers, basic physiology and
physics, and their use in recreational environments”. Nor do
I mean that experienced rebreather divers will not find the
book useful. As an experienced and well-informed rebreather
diver, I still found the material highly interesting and a
great resource, especially as a repository for the numerous
slightly arcane rebreather-related equations that I do not
carry around in my head.
The book is well written in a clear didactic style and will
serve rebreather novices extremely well. It is presented
for the most part in both Imperial and metric units, so it
has utility in this regard beyond the USA. Indeed, I would
have no hesitation in recommending it as a textbook for
any entry-level rebreather course as a supplement to unit-
specific and training-agency material. Each chapter ends
with a series of multi-choice and occasional short answer
questions designed to test understanding of key concepts,
further increasing utility as a training textbook. In general,
I found these questions to be well thought out and pitched
at the right level for a new rebreather diver.