503

ORIGINAL ARTICLE

Cooling methods used in the treatment of exertional heat

illness
J E Smith
...............................................................................................................................
Br J Sports Med 2005;39:503507. doi: 10.1136/bjsm.2004.013466

Objective: To review the different methods of reducing body core temperature in patients with exertional
heatstroke.
Methods: The search strategy included articles from 1966 to July 2003 using the databases Medline and
Premedline, Embase, Evidence Based Medicine (EBM) reviews, SPORTDiscus, and cross referencing the
bibliographies of relevant papers. Studies were included if they contained original data on cooling times
.......................
or cooling rates in patients with heat illness or normal subjects who were subjected to heat stress.
Correspondence to: Results: In total, 17 papers were included in the analysis. From the evidence currently available, the most
Dr Smith, Emergency effective method of reducing body core temperature appears to be immersion in iced water, although the
Department, Derriford practicalities of this treatment may limit its use. Other methods include both evaporative and invasive
Hospital, 4 Fort Terrace,
Plymouth PL6 5BU, UK; techniques, and the use of chemical agents such as dantrolene.
jasonesmith@doctors. Conclusions: The main predictor of outcome in exertional heatstroke is the duration and degree of
org.uk hyperthermia. Where possible, patients should be cooled using iced water immersion, but, if this is not
Accepted
possible, a combination of other techniques may be used to facilitate rapid cooling. There is no evidence to
27 September 2004 support the use of dantrolene in these patients. Further work should include a randomised trial comparing
....................... immersion and evaporative therapy in heatstroke patients.

H
eat illness is an unchecked increase in body core has been suggested that the major determinant of outcome in
temperature that occurs when intrinsic or extrinsic heatstroke is the duration of hyperthermia. One case series
heat generation overwhelms homoeostatic thermo- found a trend towards improved survival in patients cooled to
regulation. Consequential dysfunction at cellular and organ a core temperature below 38.9C within 60 minutes,6 and
level results in a spectrum of disease from minor heat cramps another report found improved survival when patients were
through symptoms of heat exhaustion to life threatening cooled to the same level within 30 minutes, although the
heatstroke.1 Heatstroke is characterised by neurological methods of cooling are not explored in detail in this paper.3
disturbance associated with haematological abnormalities, The evidence is limited, but what little there is would suggest
and may result in multiorgan failure and death.2 that outcome may be improved when core temperature is
Classical or environmental heat illness occurs in those reduced as quickly as possible. This remains the cornerstone
whose thermoregulatory control mechanisms are inefficient, of treatment of heatstroke patients, and is emphasised in the
such as the very young or elderly, and those subjected to Inter-Association Task Force on Exertional Heat Illness
extreme temperatures. The main factor in the development of consensus statement.7
this condition is a high environmental temperature, with Several methods of body cooling have been described, as
clusters of cases occurring after heat waves (700 heat related outlined in box 1. The evidence to support one method over
deaths were reported after the Chicago heat wave of 1995)3 the other is limited and appears contradictory, and authors
and the annual pilgrimages in the Middle East. have proposed there is no evidence one way or the other.1 The
In contrast, exertional heat illness typically affects young aim of this review is therefore to examine and appraise this
athletes or military personnel, who are pushed to their evidence, and give a summary and recommendations based
physical limits and suffer a clinical and pathological on its findings.
syndrome caused by an inability to dissipate heat produced
by muscular exercise. This can occur at any time of year, and METHODS
is a reflection of intrinsic heat production rather than A comprehensive search of the literature was carried out,
external heat. There are usually identifiable risk factors such using Medline and Premedline 1966 to July 2003, Embase,
as dehydration, concurrent illness, lack of sleep, obesity, Evidence Based Medicine (EBM) reviews (including the
alcohol ingestion, wearing too much clothing, or poor Cochrane database of systematic reviews and the Cochrane
cardiovascular fitness.1 4 5 However, sometimes no precipitat- central register of controlled trials), and the SPORTDiscus
ing factors are evident, and the reason why one individual is database. Search terms included heat stress disorders, heat
affected by heatstroke while others remain unaffected is not stroke, heatstroke, heat exhaustion, heat illness, and heat
clearly understood. injury. Papers were included if they contained original data
Despite differences in their pathophysiology, the recom- on cooling times or cooling rates in patients with heat illness,
mended treatment of these conditions is similar. An or normal subjects who were subjected to heat stress. Owing
assessment of airway, breathing, and circulation should be to the lack of evidence on the treatment of patients with
the priority, and basic life support instituted if appropriate. exertional heat illness in particular, papers describing the
High flow oxygen should be administered, intravenous access treatment of patients with classical or environmental heat
achieved, and initial observations should include a rectal illness were also examined. The bibliographies of relevant
temperature. Subsequently the emphasis is placed on papers were examined and cross referenced. Papers were
reduction of core temperature as quickly as possible, as it critically appraised for the quality of evidence presented. The

www.bjsportmed.com
504 Smith

Box 1: Methods of body cooling

lost to follow up. Comparison between 2 types of

blinded to patients and clinicians. No mention of

No significant difference between study Better study design. Powered to detect difference
given. Outcomes not clearly defined. 2 patients

methods and different doses of dantrolene. The

Cooling rate 0.20C/min for iced water, Non-randomised, large potential for bias. No
outcomes being blinded. Classical heatstroke
numbers. Not blinded. No exclusion criteria

calculation. Randomisation not defined. Not

Randomisation technique not defined. Small

69.2 (4.8) min (p,0.01). No difference blinded. Inclusion and exclusion criteria not

outcome target given. Flawed comparison

clear. Possible bias from different cooling

randomisation in blocks of 10. Treatment

N Body immersion in iced water

Small numbers not justified by power

conclusion does not reflect the results

group and control group in cooling time in cooling time of 30 min. Computer
N Evaporative cooling: spraying water over the patient
and facilitating evaporation and convection with the
use of fans
N Immersing the hands and forearms in cold water
N Use of ice or cold packs in the neck, groins, and axillae

evaporative cooling
N Invasive methods: iced gastric, bladder, or peritoneal
lavage

Comments
N

patients
Chemically assisted cooling with dantrolene

(67.9 (12.4) v 69.0 (10.5) min), organ

lower in treatment group: 49.7 (4.4) v
difference between groups. No deaths
(BCU) and 64 min (35120) (CM); no

dysfunction, length of hospital stay or

criteria used were relevance to the question asked and study

Mean cooling time 62 min (35120)

Mean (SD) cooling time significantly

in outcome or mortality (all patients

design. Studies were preferred in accordance with the usual
hierarchy of evidence, namely controlled clinical trials,

0.11C/min for wet towels

prospective studies (including case control studies), and case
reports.8 Review articles were examined for their reference

in patients followed up
lists.

Key results
SEARCH RESULTS

survived)

mortality
Two controlled trials comparing cooling methods (one
randomised9 and one non-randomised comparison10), and
two randomised controlled trials11 12 investigating the use of
dantrolene in the treatment of heat illness were found, and

neurological outcome, mortality

Cooling time to Tre,38.5C,

Cooling time to Tre,39.0C,

Cooling time to Tre,39.4C,

these are examined in detail in table 1. Several case series6 1319

organ dysfunction, length of

Cooling rates to unspecified

and experimental models using healthy volunteers2024 were

hospital stay, mortality

also found, and these are summarised in table 2. In total 17
papers were used in the analysis.
Summary of controlled trials comparing different methods of cooling in heatstroke patients
Outcomes

Immersion
mortality

target
In some centres immersion in iced water is the preferred
method of cooling, including the US Marine Corps training
base at Parris Island,25 although this may not be optimal

Non-randomised comparative

CNS, Central nervous system; BCU, body cooling unit; CM, conventional methods; Tre, rectal temperature.
treatment for patients with a reduced level of consciousness,
and, for those who are alert, it is uncomfortable and often
Prospective randomised

Prospective randomised

Prospective randomised
intolerable. However, the best cooling times and rates for
treatment of heat illness patients were achieved using this
controlled trial

controlled trial

controlled trial
technique.
Study type

In a non-randomised study, a comparison was made of

iced water immersion and placing wet towels over the torso,10

trial
showing that iced water immersion cooled patients faster
than the wet towels, with a cooling rate of 0.20C/min for iced cooled either by iced water immersion or
water and 0.11C/min for wet towels. However, the study
9C. Either given dantrolene 24 mg/kg
20 classical heatstroke patients, Tre.41.
16 heatstroke patients (Tre.40C, dry
skin, CNS symptoms). Randomised to

design allowed clinicians to choose which patients received

52 classical heatstroke patients with

which treatment, a source of considerable selection bias. It

21 exertional heatstroke patients,
Patient group and interventions

Tre.40.1C. Given dantrolene

also compared two similar forms of cooling, rather than

comparing immersion with evaporative techniques.
In a case series of 28 classical heatstroke patients,
BCU or CM (see text)

2 mg/kg or placebo

immersion in iced water enabled cooling to less than 102F

(38.9C) in less than 45 minutes in all patients.14 Of note,
some of the patients in this study could not tolerate being
wet towels

immersed and were massaged with ice instead. These

or not

patients were not analysed separately, and the relative

importance of the ice massage cannot therefore be elucidated.
Armstrong et al,10 1996, USA

In another case series of 27 patients with exertional

Reference, date, and country

Al-Harthi et al, 1986, Saudi

Channa et al, 1990, Saudi

heatstroke or heat exhaustion, all were cooled by iced water

Bouchama et al,12 1991,

immersion, achieving a mean cooling time (to Tre,39C) of

19.2 minutes, with a cooling rate of 0.15C/min.18
9

11

Supporters of other cooling methods suggest that iced

Saudi Arabia

water immersion may cause peripheral vasoconstriction and

Table 1

therefore slow cooling, although this has never been proven

Arabia

Arabia

experimentally. Indeed, a recent study using normal volun-

teers has shown that cooling rates were fastest when the
coldest water was used.25a

www.bjsportmed.com
 Table 2 Summary of papers describing case series and experimental cooling models
Author, date, and country Patient group and interventions Study type Outcomes Key results Comments
20
Wyndham et al, 1959, 6 healthy volunteers exercised until Within subject crossover trial Cooling time to Tre 38.3C, cooling Fastest mean cooling time 50 min, Small numbers. Applicable to heat stroke
South Africa Tre 40C, cooled by immersion or rate (fall in Tc) over 60 min rate 0.07C/min, with evaporative patients?
evaporation cooling
21
Weiner & Khogali, 1980, UK 6 healthy volunteers exercised until Tty Within subject crossover trial Reduction in Tty of 2.0C (to 37.5C) Cooling time 6.5 min (rate Small numbers. Methodology not
39.5C, then cooled by immersion in 0.31C/min) with BCU, 18.4 min explained in detail. Tympanic
15C water, cold air spray or warm air (rate 0.11C/min) with immersion measurement used. Applicable to heat
spray (BCU) stroke patients?
Kielblock et al,22 1986, 5 healthy volunteers exercised until Tre Within subject crossover trial Reduction in Tre of 2.0C (to baseline) Mean cooling time 73.6 min Small numbers. Methodology not
South Africa (cold packs), 59.8 min (evaporative explained in detail. Applicable to heat
Cooling methods for exertional heat illness

reached 2C above baseline, then

subjected to different methods of cooling); p,0.01. Combined: stroke patients?
cooling (see results) 53.6 min or 0.04C/min (no
difference)
Clapp et al,23 2001, USA 5 healthy volunteers exercised until Tre Within subject crossover trial Cooling rate (fall in Tc) over 30 min Mean cooling rates 0.25C/min Small numbers. Not compared with
38.8C, then subjected to different (torso immersion), 0.16C/min standard of care. Low target
methods of cooling (see results) (hand and feet immersion) temperature. Applicable to heat stroke
and 0.11C/min (fan) patients?
24
Mitchell et al, 2001, USA 10 healthy volunteers exercised to Within subject crossover trial Cooling rate (fall in Tc) during two Mean cooling rates 0.020.05C/min, Study not designed to look at heatstroke
preset limits, then cooled using 4 12 min cooling periods no difference between passive cooling model. Inadequate temperature rise and
methods (see results) and fan or spray assisted cooling cooling times to assess treatments
13
Khogali & Weiner, 1980, 18 heatstroke patients cooled in a Case series Cooling time to Tre,38C. Mortality Cooling times ranged from 30 to Data on cooling rates not given
Kuwait BCU 300 min. 2 patients died. Calculated
mean cooling rate 0.05C/min
14
Hart et al, 1982, USA 28 classical heatstroke patients. All Case series Cooling time to Tre,102F (38.89C) 26 out of 28 cooled within 30 min, all Different intervention (immersion v
cooled by iced water immersion or within 45 min massage) not defined in study group. No
ice massage if immersion not possible numerical data given. Study designed to
primarily look at other parameters
Vicario et al,6 1986, USA 39 heatstroke patients, cooled by Case series Temperature reduction over first hour, 69% cooled to T,38.9C within Case series looking at prognosis
various means depending on physician mortality 60 min. Mortality 21%, lower when depending on cooling times, rather than
preference cooled faster comparing methods
15
Graham et al, 1986, USA 14 classical heatstroke patients, Case series Cooling time to Tre,39.4C, mortality Cooling times ranged from 34 to Cooling methods not well defined: mixed
cooled by evaporative methods. Ice also 89 min, median 60 min. 1 patient methods. Cooling rates not calculable
used around torso died
16
Al-Aska et al, 1987, 25 classical heatstroke patients cooled Case series Cooling time to Tre,39C Mean cooling time 40 min (rate Methods not well described: brief letter
Saudi Arabia with simplified cooling bed 0.09C/min) with simplified cooling
bed
17
Poulton & Walker, 1987, USA 3 heatstroke patients (2 exertional, 1 Case series Cooling time to T,38.5C Mean cooling time 24 min, rate Methods not adequately described. Only
classical), cooled by helicopter 0.10C/min mean data given
downdraft
18
Costrini, 1990, USA 27 patients with exertional heat Case series Cooling time to Tre,39C, mortality Mean cooling time 19.2 min or rate Methods not adequately described. Only
illness, all cooled by iced water 0.15C/min. No deaths mean data given
immersion
Horowitz,19 1989, USA Heatstroke patient cooled by iced Case report Cooling rate. No end point given Cooling rate 0.11C/min during Case report. Evaporative cooling also
peritoneal lavage lavage cycle. Patient survived used

BCU, Body cooling unit; Tre, rectal temperature; Tty, tympanic temperature; Tc, core temperature.

www.bjsportmed.com
505
506 Smith

Evaporative methods defined. The results showed a significantly shorter cooling

Evaporative cooling involves the removal of clothing, spray- time (Tre,39.0C) in the study group (49.7 (4.4) v 69.2
ing tepid water over the patient, and facilitating evaporation (4.8) minutes, p,0.01), but no difference in outcome
and convection with the use of a fan. This was shown in the between the two groups. Table 1 outlines the methodological
first published study of its kind to be better than other flaws in this study, which should induce caution in the
methods of cooling in normal volunteers.20 The basic idea was interpretation of its results. In the second trial, 52 classical
modified into a more sophisticated model termed a body heatstroke patients were randomised to receive either
cooling unit, involving water sprayed at 15C being mixed dantrolene 2 mg/kg or placebo (normal saline).12 All patients
with warm air flow to maintain skin temperature at 32 were externally cooled using a body cooling unit, and rectal
33C.13 21 Theoretical advantages of this include reducing temperature was continuously recorded to determine cooling
peripheral vasoconstriction and thereby allowing heat to be time (time to reach Tre,39.0C). The results showed no
lost more effectively. In a randomised controlled trial of difference in cooling times between groups (treatment group
classical heatstroke patients, the body cooling unit was 67.9 (12.4) minutes v control group 69.0 (10.5) minutes),
compared with conventional methods of cooling (covering and no difference in complications, length of stay, or
the patient in a wet sheet, spraying tap water on to the sheet, mortality (one patient died in the control group). From this
and using fans to blow air over the patient), but no difference well designed trial, the authors conclude that, at the dose
in cooling time was found.9 This comparison was also flawed, given, dantrolene has no additional benefit over traditional
in that it compared two very similar methods of cooling with external cooling methods.
no reference made to iced water immersion therapy.
In another study, the body cooling unit was shown to be
DISCUSSION
better than other methods of cooling in normal volunteers,21
As is evident from this collection of the available data, there
achieving rapid cooling with a mean rate of 0.31C/min, but in a
is conflicting evidence as to the best form of cooling in
case series of 18 heatstroke patients achieved a mean cooling
heatstroke patients. Treatment should start with an assess-
rate of only 0.05C/min, with cooling times (to Tre,39C)
ment of airway, breathing, and circulation, and measurement
ranging from 30 to 300 minutes.13 In another case series, a
of rectal temperature. According to the best evidence
simplified cooling bed achieved slightly better results, with a
currently available, it would appear that immersion in iced
mean cooling time of 40 minutes (rate 0.09C/min).16
water is the most effective method of whole body cooling,
and should be used where possible. However, it may not be
Other cooling methods practical from a logistic or clinical perspective, and treatment
Immersing the hands and forearms in cold water results in may have to be tailored to the clinical situation. For patients
reduction of heat stress in normal subjects,2628 but has not with a reduced level of consciousness, it may prove hazardous
been evaluated in the treatment of heatstroke patients. The and is not recommended unless specific facilities allowing
placement of ice packs in the axillae, groin, and neck has continued intensive medical care are available.
been recommended as an easy method to use in the field,29 30 If immersion is unavailable or inappropriate, cooling may
but when compared with evaporative cooling in five healthy necessarily involve a combination of evaporative cooling
subjects, cooling times were longest when the ice packs were techniques and other methods such as immersion of the
used alone and shortest when both methods were used extremities in cold water. At the roadside of an endurance
simultaneously.22 Another method of evaporative cooling has event, this may be the most appropriate method of cooling.
been described using the downdraft of a helicopter, which Future work should include a prospective randomised trial
was reported to have better results than using the body comparing iced water immersion with evaporative techniques
cooling unit, although this case series only involved three in exertional heat illness patients to clarify the optimal method.
patients (mean cooling time to Tre,38.5C 24 minutes, rate There is no clear evidence to support the use of dantrolene
0.10C/min), and the method is clearly not generally in the treatment of exertional heatstroke, and the priority,
available.17 Invasive methods of cooling include gastric, after an assessment of airway, breathing, and circulation,
peritoneal, and bladder lavage with cold water. These have should be to institute external cooling methods to reduce core
been investigated in animal models,3133 but the only data in temperature as quickly as possible. Other suggested treat-
human subjects come from a case series where gastric lavage ments such as the use of antipyretics and immunomodula-
is described,6 and a single case report, when cooling was tory agents have not yet been evaluated in human subjects.
achieved in a patient resistant to evaporative techniques.19 This review was essentially limited to human studies, and
The role of invasive cooling methods has therefore not been data from animal studies have not been included. It may be
fully established. subject to publication bias, and has included papers from
predominantly English Language journals. The distinction
Dantrolene has been drawn between those papers describing cooling in
Dantrolene is an agent that impairs calcium release from the heatstroke patients and those involving normal volunteers
sarcoplasmic reticulum and by doing so reduces muscle exposed to experimental conditions. It should be stressed
excitation and contraction. It is used in the treatment of that experiments in which cooling rates are measured in
malignant hyperthermia and neuroleptic malignant syn- normal subjects who have been exercising have limited
drome, reducing heat production that occurs as a result of applicability to treating heatstroke patients whose thermo-
muscle rigidity or hypertonicity typical of these conditions.34 regulation mechanisms have failed. The treatment of classical
It has been suggested that similar pathophysiological heat illness patients has also been included, although
mechanisms underlie heat illness, and dantrolene has been whether this can be extrapolated to patients with exertional
proposed as a possible means to reduce core temperature in heat illness is not clear.
this condition.3540 The search strategy for this review yielded Comparison of papers was made more difficult by the
two randomised trials11 12 as shown in table 1. different outcome measures used by different groups, but
In the first trial, 20 heatstroke patients were randomised where possible cooling times and rates have been extracted
either to receive dantrolene 24 mg/kg (study group) or not and compared. The lack of robust evidence highlights the
(control group).11 Patients were externally cooled using either requirement for a large, well designed, randomised controlled
a body cooling unit or conventional evaporative methods, trial comparing iced water immersion with evaporative
although which patients received which method is not techniques for the treatment of heatstroke.

www.bjsportmed.com
Cooling methods for exertional heat illness 507

14 Hart GR, Anderson RJ, Crumpler CP, et al. Epidemic classical heatstroke:
What is already known on this topic clinical characteristics and course of 28 patients. Medicine 1981;61:189.
15 Graham BS, Lichtenstein MJ, Hinson JM, et al. Nonexertional heatstroke. Arch
Intern Med 1986;146:8790.
N Classical or environmental heat illness results from 16 Al-Aska A, Abu-Aisha H, Yaqub B, et al. Simplified cooling bed for heat
stroke. Lancet 1987;i:381.
inefficient thermoregulatory control mechanisms in 17 Poulton TJ, Walker RA. Helicopter cooling of heatstroke victims. Aviat Space
those subjected to extreme temperatures. Environ Med 1987;58:35861.
18 Costrini A. Emergency treatment of exertional heatstroke and comparison of
N Exertional heat illness results from an inability to whole body cooling techniques. Med Sci Sport Exerc 1990;22:1518.
dissipate heat produced by muscular exercise. 19 Horowitz BZ. The golden hour in heat stroke: use of iced peritoneal lavage.
Am J Emerg Med 1989;7:61619.
N The recommended treatment of these conditions is 20 Wyndham CH, Strydom NB, Cooke HM, et al. Methods of cooling subjects
similar: assessment of airway, breathing, and circula- with hyperpyrexia. J Appl Physiol 1959;14:7716.
21 Weiner JS, Khogali M. A physiological body cooling unit for treatment of heat
tion and basic life support, followed by reduction of stroke. Lancet 1980;i:507.
core temperature as quickly as possible. 22 Kielblock AJ, Van Rensburg JP, Franz RM. Body cooling as a method for
N Several methods of body cooling have been recom- reducing hyperthermia. An evaluation of techniques. S Afr Med J
1986;9:37880.
mended. 23 Clapp AJ, Bishop PA, Muir I, et al. Rapid cooling techniques in joggers
experiencing heat strain. J Sci Med Sport 2001;4:1607.
24 Mitchell JB, Schiller ER, Miller JR, et al. The influence of different external
cooling methods on thermoregulatory responses before and after intense
intermittent exercise in the heat. J Strength Cond Res 2001;15:24754.
25 Gaffin SL, Gardner JW, Flinn SD. Cooling methods for heatstroke victims. Ann
What this study adds Intern Med 2000;132:678.
25a Proulx CI, Ducharme MB, Kenny GP. Effect of water temperature on cooling
efficiency during hyperthermia in humans. J Appl Physiol 2003;94:131723.
N The main predictor of outcome in exertional heatstroke 26 Livingstone SD, Nolan RW, Cattroll SW. Heat loss caused by immersing the
hands in water. Aviat Space Environ Med 1989;60:116671.
is the duration and degree of hyperthermia. 27 Allsopp AJ, Poole K. The effect of hand immersion on body temperature when
N Patients should be cooled using iced water immersion, wearing impermeable clothing. J R Nav Med Serv 1991;77:417.
28 House JR, Holmes C, Allsopp AJ. Prevention of heat strain by immersing the
but, if this is not possible, a combination of other
hands and forearms in water. J R Nav Med Serv 1997;83:2630.
techniques may be used. 29 Richards D, Richards R, Schofield PJ, et al. Management of heat exhaustion in
N There is no evidence to support the use of dantrolene in Sydneys the sun city to surf fun runners. Med J Aust 1979;2:45761.
30 Roberts WO. Managing heatstroke: on-site cooling. Phys Sportsmed
these patients. 1992;20:1728.
N Further work should include a randomised trial 31 Bynum G, Patton J, Bowers W, et al. Peritoneal lavage cooling in an
anaesthetized dog heatstroke model. Aviat Space Environ Med
comparing immersion and evaporative therapy in 1978;49:77984.
heatstroke patients. 32 Syverud SA, Barker WJ, Amsterdam JT, et al. Iced gastric lavage for treatment
of heatstroke: efficacy in a canine model. Ann Emerg Med 1985;14:42432.
33 White JD, Riccobene E, Nucci R, et al. Evaporation versus iced gastric lavage
treatment of heatstroke: comparative efficacy in a canine model. Crit Care
Med 1987;15:74850.
CONCLUSIONS 34 Ward A, Chaffman MO, Sorkin EM. Dantrolene. A review of its
pharmacodynamic and pharmacokinetic properties and therapeutic use in
The treatment of exertional heatstroke should begin with an malignant hyperthermia, the neuroleptic malignant syndrome and an update
assessment of airway, breathing, and circulation, and of its use in muscle spasticity. Drugs 1986;32:13068.
initiation of resuscitation if necessary. Subsequently whole 35 Denborough MA. Heat stroke and malignant hyperpyrexia. Med J Aust
1982;1:2045.
body cooling should be the priority. Where possible, patients 36 Denborough MA. Fatal thermal injury. Med J Aust 1989;150:608.
should be cooled using iced water immersion, although this 37 Larner AJ. Dantrolene for exertional heatstroke. Lancet 1992;339:182.
may not always be practical and a combination of other 38 Lydiatt JS, Hill GE. Treatment of heat stroke with dantrolene. JAMA
1981;246:412.
techniques may be needed to facilitate rapid cooling. There is 39 Paasuke RT. Drugs, heatstroke and dantrolene. Can Med Assoc J
no evidence to support the use of dantrolene in these 1984;130:3412.
patients. 40 Knochel JP. Treatment of heat stroke. JAMA 1983;249:10067.

Competing interests: none declared

REFERENCES
1 Bouchama A, Knochel JP. Heat stroke. New Engl J Med 2002;346:197888.
.............. COMMENTARY ..............
2 Epstein Y, Moran DS, Shapiro Y, et al. Exertional heat stroke: a case series.
Med Sci Sports Exerc 1999;31:2248.
The authors review of the management of exertional heat
3 Dematte JE, OMara K, Buescher J, et al. Near-fatal heat stroke during the
1995 heat wave in Chicago. Ann Intern Med 1998;129:17381. illness clearly highlights the problem in deciding the best
4 Chung NK, Pin CH. Obesity and the occurrence of heat disorders. Mil Med treatment for this group of patients, who are typically young
1996;161:73942. and healthy but who have a high mortality and morbidity,
5 Gardner JW, Kark JA, Karnel K, et al. Risk factors predicting exertional heat
illness in male Marine Corps recruits. Med Sci Sport Exerc 1996;28:93944.
related directly to the duration of the temperature rise.
6 Vicario SJ, Okabajue R, Haltom T. Rapid cooling in classic heatstroke: effect With regard to the best form of treatment, there are two
on mortality rates. Am J Emerg Med 1986;4:394. camps, equally split between immersion and evaporative
7 Inter-Association Task Force on Exertional Heat Illness. Consensus statement. therapy. The evidence presented here appears to support the
2002. http://www.acsm.org/publications/pdf/HeatIllnessConsensusStmt.pdf
(accessed 28 Jul 2004).
use of immersion in ice water, while acknowledging the
8 Oxford Centre for Evidence-based Medicine. Levels of evidence. http:// limitations both in terms of the quality of evidence available
www.cebm.net/levels_of_evidence.asp (accessed 28 Jul 2004). and the practicalities of using this technique.
9 Al-Harthi SS, Yaqub BA, Al-Nozha MM, et al. Management of heat stroke This is yet another area of sports medicine in which the
patients by rapid cooling at Mecca pilgrimage. Saudi Med J 1986;7:369.
10 Armstrong LE, Crago AE, Adams R, et al. Whole-body cooling of answer to a seemingly simple question is still unknown, and
hyperthermic runners: comparison of two field therapies. Am J Emerg Med presents an area in need of a well conducted randomised
1996;14:3558. trial.
11 Channa AB, Seraj MA, Saddique AA, et al. Is dantrolene effective in heat
stroke patients? Crit Care Med 1990;18:2902.
12 Bouchama A, Cafege A, Devol EB, et al. Ineffectiveness of dantrolene sodium L Wallis
in the treatment of heatstroke. Crit Care Med 1991;19:17680. University of Cape Town, PO Box 901, Wellington 7654, South Africa;
13 Khogali M, Weiner JS. Heat stroke: report of 18 cases. Lancet 1980;ii:2768. leewallis@bvr.co.za

www.bjsportmed.com