Você está na página 1de 9

Title:

A randomized prospective trial of hyperbaric oxygen in a referral burn


center population.
Authors:
Brannen AL; Still J; Haynes M; Orlet H; Rosenblum F; Law E; Thompson
WO
Author's Address:
Augusta Regional Medical Center, Georgia, USA.
Source:
The American Surgeon [Am Surg] 1997 Mar; Vol. 63 (3), pp. 205-8.
Clinical Trial; Journal Article; Randomized Controlled Trial
:
English
:
Country of Publication: UNITED STATES NLM ID: 0370522 Publication
Model: Print Cited Medium: Print ISSN: 0003-1348 (Print) Linking ISSN:
00031348 NLM ISO Abbreviation: Am Surg Subsets: MEDLINE
s:
Hyperbaric Oxygenation*
Burns/*therapy
Adult; Burns/mortality; Burns/surgery; Chi-Square
Distribution; Female; Humans; Length of Stay; Male; Middle
Aged; Regression Analysis; Survival Analysis; Treatment Outcome
As
Various studies of the effect of hyperbaric oxygen (HBO) in a wide
variety of disease entities have been carried out. In the treatment of
burns, animal and human studies have yielded somewhat contradictory
results. Controlled studies in humans are limited. A randomized study on
the effect of HBO was conducted involving 125 burn patients admitted
within 24 hours of injury who were matched by age, burn size, and
presence or absence of inhalation injury. Patients in the treatment arm
received oxygen at two atmospheres of pressure for 90 minutes twice a
day for a minimum of 10 treatments and a maximum of one treatment
per total body surface per cent burn. The control group was treated in a
similar fashion, except for the absence of HBO. There were no
statistically significant differences between the two groups for the
outcome measures of mortality, number of operations, and length of
stay for the survivors. In this large clinical trial, we were unable to
demonstrate any significant benefit to burn patients from the use of
HBO.
Entry Date
Date Created: 19970307 Date Completed: 19970307 Latest Revision:
20041117

http://search.ebscohost.com/login.aspx?
direct=true&db=mnh&AN=9036884&site=ehost-live
C
<A href="http://search.ebscohost.com/login.aspx?
direct=true&db=mnh&AN=9036884&site=ehost-live">A randomized
prospective trial of hyperbaric oxygen in a referral burn center
population.</A>

MEDLINE with Full Text

A RANDOMIZED PROSPECTIVE TRIAL OF HYPERBARIC


OXYGEN IN A REFERRAL BURN CENTER POPULATION
Various studies of the effect of hyperbaric oxygen (HBO) in a wide variety of
disease entities have been carried out. In the treatment of burns, animal and
human studies have yielded somewhat contradictory results. Controlled studies
in humans are limited. A randomized study on the effect of HBO was conducted
involving 125 burn patients admitted within 24 hours of injury who were
matched by age, burn size, and presence or absence of inhalation injury.
Patients in the treatment arm received oxygen at two atmospheres of pressure
for go minutes twice a day for a minimum of 10 treatments and a maximum of
one treatment per total body surface per cent burn. The control group was
treated in a similar fashion, except for the absence of HBO. There were no
statistically significant differences between the two groups for the outcome
measures of mortality, number of operations, and length of stay for the
survivors. In this large clinical trial, we were unable to demonstrate any
significant benefit to burn patients from the use of HBO.

HYPERBARIC OXYGEN (HBO) IS currently employed as an accepted treatment


in a variety of therapeutic situations (Grim et al.( n1)), including
decompression sickness, acute carbon monoxide poisoning, acute cyanide
poisoning, chronic refractory osteomyelitis, acute gas embolism, gas gangrene,
crush injuries, refractory anaerobic infection, radiation necrosis, preparation of
radiated bone for grafting, treatment of skin grafts and flaps with questionable
viability, and enhancement of healing in selected wound problems. A number
of studies have involved the use of HBO in attempts to improve the treatment
of burned individuals, but controlled studies are rare, both animal and human.
We have carried out a randomized study of matched, paired patients at the
Burn Unit of Augusta Regional Medical Center involving 125 acutely burned
patients.

Patients and Methods


Acutely burned patients admitted within 24 hours of injury were evaluated for
inclusion into the study. The patients were randomized in pairs on the basis of
age, burn size, and the presence or absence of inhalation injury and were
assigned to receive either routine burn management or the same routine of
burn care with the addition of HBO therapy. Patients received their first HBO
treatment within 24 hours of burn injury and were treated at two atmospheres
of pressure for 90 minutes twice a day for a minimum of at least 10 treatments
and a maximum of one treatment per per cent total body surface area (TBSA)
burn. All patients received the maximum number of HBO treatments for their
burn size, unless treatments could not be given due to serious illness or some
problem preventing treatment in the chamber. The project was approved by
the Institutional Review Board of Humana Hospital Medical Center, which was
subsequently renamed Augusta Regional Medical Center (Augusta, GA).
Consent forms were signed by the patient or an appropriate family member.

Limitations on the availability of hyperbaric chamber space meant it was not


possible to make HBO therapy available to all patients on admission, so that,
unless chamber space was available, the patient could not be included in the
study. Therefore, the group randomized to receive HBO was treated on a
space-available basis. These patients were compared with a group of patients
treated in similar fashion and matched for age, burn size, square centimeters
of excised area of burn, and presence or absence of inhalation injury selected
from the patient group not receiving hyperbaric therapy.

The primary outcome variable used in this analysis was length of stay (LOS).
Secondary variables included mortality and number of surgeries. The analyses
of LOS and number of surgeries was performed on only the data from the
surviving patients, because the values of those who died were censored.
Statistical analysis of the importance of the HBO treatment on LOS was
performed using a two-stage stepwise regression; the first stage contained a
list of possible covariates known to influence LOS (e.g., number of surgeries,
per cent and depth of burn, and presence of inhalation injury). Following the
determination of covariates in the first stage of the model analysis, an
indicator variable to isolate the effect of HBO treatment was added to the
model as the second stage. Thus, the importance of HBO treatment on LOS
follows after a statistical adjustment of LOS due to covariates known to affect
LOS. This model is a modification of analysis of covariance, where the stepwise
regression algorithm is used to identify the best subset of covariates prior to
testing for the factor of interest (in this case, the effect of HBO treatment). The
number of surgeries was analyzed using a similar approach. Analysis of
mortality was conducted using a Chi2 statistic calculated from the 2 by 2
contingency table. Differences in demographic statistics between the two
groups of patients by survival were assessed using 2 by 2 analysis of variance
for continuous variables and contingency tables for categorical variables.

Results
There were 125 patients enrolled in the study, with 63 receiving HBO therapy
and 62 not receiving HBO (controls). One-third of patients received their first
treatment within 8 hours after injury. Table 1 shows the means of
demographic variables of the sample, split by survival within the two groups
(HBO and control) regarding age, burn size, inhalation injuries, and square
centimeter of excision. Although the number of deaths was small in each group
(11 per cent), the age, per cent full-thickness burn and per cent TBSA burn of
those who died were significantly larger in each group than those who
survived. Inhalation injury was more likely among those who died, but square
centimeters of excised burn wound was smaller among those who died than
among those who survived. There were no significant differences among the
HBO group and the control group means, and there were no significant
interactions of treatment group by survival.

The primary outcome variable, LOS, was analyzed for the 111 surviving
patients, because those patients who died had censored values for their LOS.
Statistical analysis of covariance of LOS showed that LOS covaried most
strongly with number of surgeries, accounting for 70.2 per cent of the variance
(P < 0.0001). Age was the next most important predictor, followed by per cent
TBSA burn and presence of inhalation injury. Taken as a set, these four
variables accounted for 76.6 per cent of the variance. LOS increased with
increasing levels of each of the factors. No other variables, including treatment
with HBO, were statistically significant in the presence of these four predictors.
Forcing a treatment group variable into the model increased the r 2 by only
0.00065 and decreased the adjusted r2 from 0.756 to 0.755. The presence of
inhalation injury added an average of 4.3 additional days to the patient's stay
(P = 0.012). Estimates from the regression model are given in Table 2.

The number of surgeries covaried with per cent full-thickness burn (r 2 = 0.518;
P < 0.001) but no other covariates. Adding the effect of the treatment group
increased the r2 by 0.001 and decreased the adjusted r2 from 0.513 to 0.510.
Thus, the number of surgeries, adjusted for percentage full-thickness burn,
was not different between the two treatment groups.

There is no difference in LOS, mortality, or number of surgeries between the


control and HBO-treated groups once these variables were adjusted for the
patient's condition.

Discussion
HBO therapy involves the inhalation of 100 per cent oxygen at a pressure
greater than that of atmospheric pressure at sea level. The mechanism of
action is felt to be achieved by hyperoxygenation of the patient's blood. Other
concomitant effects include vasoconstriction, fibroblast proliferation,
enhancement of white cell killing of bacteria, and neovascularization of
wounds.

The rationale for use of HBO in burn patients includes the potential benefits of
vasoconstriction, which may offset the vasodilation resulting from the acute
burn. This yields a decrease in the exudation of plasma into the intravascular
space, which may in turn inhibit the development of burn shock and decrease
fluid requirements. HBO also counteracts hypoxia of the tissues, felt to be of
benefit in controlling wound infection and in promoting re-epithelialization of
the burn wound.

Animal experiments involving the use of HBO in burn wounds have been
carried out. Ketchum et al.( n2) noted that burn wound healing time was
increased by 30 per cent in rabbits treated with HBO. Korn et al.( n3) found
faster re-epithelialization of burn wounds and an earlier return of capillary
patency after hyperbaric therapy. Nylander et al.( n4) reported that rat scalds
treated with HBO had less tissue edema, faster repletion of adenosine
diphosphate, reduced lactate levels, and preservation of phosphorylase.

Not all reports are favorable. Perrins( n5) reported no effect of HBO in a pig
scald model. Niccole et al.( n6) reported no advantage in wound healing
achieved by HBO when the modality was compared with topical antibiotics.

Human information is fairly limited. Wada et al.( n7) in 1965 reported


improved healing of second-degree burns in coal miners treated with HBO for
carbon monoxide poisoning. Clinical series have been reported by a number of
authors. Waisbren et al.( n8) showed a 75 per cent reduction in the need for
grafting. Cianci et al.( n9) showed reduced LOS for burns up to 39 per cent. In
another study, Cianci et al.( n10) showed reduced need for surgery and
grafting and decreased cost to patients with 40 to 80 per cent burns.
Hammarlund et al.( n11) showed reduced edema and decreased exudation of
fluid from burn blister wounds. Grossman( n12) showed decreased LOS and
decreased mortality. Grube et al.( n13) found no benefit in his cases and noted
a number of complications due to the hyperbaric therapy. Problems in various
studies have included temporary vision loss, pneumothorax, and perforated ear
drums.

Controlled studies are difficult to design, and relatively few have been carried
out. As indicated, a number of authors have reported in controlled studies
satisfactory results with HBO in patients. It is recognized that the optimal way
to do a study of HBO in burn patients would be to enroll all comers who
consent to the study. Then a sham treatment would be done in which the
patient was actually placed in the hyperbaric chamber and left there for same
period of time as the treated patient but with the HBO not running. This would
be the optimal blinded approach. It would, however, pose the ethical questions
of risk and additional painful manipulation to the untreated patient who had no
possibility of receiving benefit. Also, the lack of availability of chamber space at
our institution precluded this study design. Therefore, the study model that
was selected was based on having hyperbaric chamber time available for
actual use when a patient was admitted. Patients treated were then matched
with untreated controls, as indicated in the protocol, to achieve comparable
patient groups. As can be seen from the data, this matching of patients was
essentially successful, and the two groups were not significantly different with
regard to age, burn size, presence or absence of an inhalation injury, and
areas of burn excision. We found no significant difference in mortality, with an
11 per cent death rate in both groups. The requirement for surgery was the
same in both groups. Length of hospital stay was also the same statistically. It
is recognized that LOS depends on a great many factors, including such things
as income level and availability of home care, which we did not attempt to
evaluate. In our model, benefit as shown by improved mortality, decreased
LOS, and decreased need for surgery could not be demonstrated. The efficacy
of HBO therapy when added to our conventional program of management did
not improve these parameters.
DISCUSSION

DR. ROBERT P. CARRAWAY (Birmingham, Alabama): Dr. Still has covered


the animal and clinical studies indicating the beneficial effects of hyperbaric
oxygen in the treatment of thermal injuries. Those studies have reported
preservation of the microcirculation, more rapid epitheliazation, and a
reduction in the incidence infection, number of surgeries, length of hospital
stay, and mortality. In 1992, there was a report advocating the use of early
HBO therapy for burns, but as Dr. Still pointed out, distance was a problem in
starting early treatment. Dr. Still, were you able to see any change in the
length of stay in your patients receiving early HBO treatments? Did they get an
earlier discharge from your burn center? Second, your treatment schedule of a
minimum of 10 and a maximum of 1 treatment per per cent body surface area
varies from the standard recommendations. In considering your protocol, it
appears the HBO survivors with 21 per cent total body surface area would have
received 21 treatments and the nonsurvivors with 52 per cent total body
surface area burns would have received 52 treatments. I was confused about
your rationale for the difference in the treatment times and how long your
patients received hyperbaric therapy. Could you help me understand that
better? And lastly, hyperbaric therapies have been shown in human studies to
exert a positive beneficial effect on the burn wound by reducing edema and
plasma extravasation. Controlled randomized studies have also showed
reduction in fluid requirements. Did you note similar changes in fluid
requirements in your HBO treated group?

I feel the investigators have made a contribution in the treatment of burns


using HBO. The population that they have studied is such that the per cent
burn is small, and it confirms the fact that individuals with burns less than 20
per cent do not require HBO. Finally, I do not feel that your study addresses
larger burns.

DR. JOSEPH STILL (Closing Discussion): Dr. Carraway, we did not analyze
the 16 patients who were treated within the defined 8-hour period, thinking
that the number was too small to be significant. The number of treatments
that we chose was on a basis of Dr. A. Lynne Brannen's desire to provide
patients with as much hyperbaric exposure as possible, and so we picked the
one treatment per 1 per cent body burn. I am not sure why he didn't use the
three treatments in the first 24 hours, except again to say that most of our
patients didn't arrive in time to receive three treatments within the first 24
hours. We were very impressed when we started using HBO. There was
definitely less fluid loss, the patients were drier, and they appeared to heal
earlier. At the present time, we are investigating fluid requirements in patients
treated with and without HBO and we should have that information available
within a year.

TABLE 1. Means of Demographic Variables, by HBO Treatment and


Survival
Survival
HBO-Treated Group

Variable Yes No

Number of patients 55 7
Females 9 3
Males 46 4

Means

Age 29.5 44.9


Full-thickness burn 7.8 36.4
Partial burn (% TBSA) 13.4 15.7
Total burn (% TBSA) 21.6 52.1
Inhalation injuries (%) 17 (31) 6 (86)
cm2 excised burn 1,447 859

Survival

Control Group

Variable Yes No P Value(*)

Number of patients 56 7
Females 15 4
Males 41 3

Means

Age 32.7 49.4 0.012


Full-thickness burn 6.7 45.0 0.001
Partial burn (% TBSA) 14.2 7.9 ns
Total burn (% TBSA) 20.9 52.9 0.001
Inhalation injuries (%) 17 (30) 5 (71) 0.001
cm2 excised burn 1,436 705 ns

(*) Significance of difference in means between survived group


and nonsurvived group. There were no significant differences
among the HBO group and the Control group means.
TABLE 2. Regression Estimates for Length of Stay
Variable b SE b Beta t Sig t

Age(years) 0.155 0.038 0.194 4.09 0.0001


Total percentage burn 0.166 0.065 0.136 2.54 0.0124
Inhalation injury(*) 4.273 1.851 0.115 2.31 0.0229
Number of operations 6.288 0.453 0.745 13.89 0.0001
Constant -3.018 2.012
R2 0.765
Standard error 8.526

b, regression coefficient; SE b, standard error of b; Beta,


standardized regression coefficient (b/SE b); t, t test of
hypothesis b (or Beta) = 0; Sig t, P value of the t test.

(*) Inhalation injury: 0, no injury; 1, injury.


REFERENCES
(n1.) Grim P. Gottlieb L, Boddie A, et al. Hyperbaric oxygen therapy: State of
the art review. JAMA, 1990;263:2216-20.

(n2.) Ketchum F, Thomas A, Hall AEL. Proceedings of the Fourth International


Congress on Hyperbaric Medicine. Williams & Wilkins Co., 1970, pp 383-394.

(n3.) Korn HN, Wheeler ES, Miller TA. Effect of hyperbaric oxygen on second
degree bum wound healing. Arch Surg, 1977; 112:732.

(n4.) Nylander G, Nordstrum H, Ericksson E. Effect of hyperbaric oxygen on


edema formation after a scald burn. Burns Including Therm Inj 1984;10:1993-
6.

(n5.) Perrins D. Failed attempt to limit tissue distruction in scalds of pig skin
with HBO: Proceedings of the Fourth International Congress on Hyperbaric
Medicine. Tobyo Ijaku Shirn 1970;381.

(n6.) Niccole M, Thornton J, Danet R, et al. Hyperbaric oxygen in burn


management under controlled studies. Surgery 1977;727-33.

(n7.) Wada J, Ikeda K, Kagaya, et al. Igakunoay MI (Japan) Proceeding of


Fourth International Congress on HBO Treatment. 1965;5:53.

(n8.) Waisbren V, Schutz D, Colentine G, et al. Hyperbaric oxygen for severe


burns. Burns Including Therm Inj 1982;8:176-9.

(n9.) Cianci P, Lueders H, Lee H, et al. Adjunctive hyperbaric oxygen therapy


reduces length of hospitalization in thermal burns. J Burn Care & Rehab,
1989;10:432-500.

(n10.) Cianci P, William C, Lueders H, et al. Adjunctive hyperbaric oxygen in


treatment of thermal burns: An economic analysis. J Burn Care & Rehab,
1990;11:140-3.

(n11.) Hammarlund C, Svedman C, Svedman P. Hyperbaric oxygen treatment


of healthy volunteers with UV-radiated blister wounds. Burns 1991;17:296-
300.

(n12.) Grossman A. Hyperbaric oxygen in the treatment of burns. Ann Plast


Surg 1978;1:163-7.

(n13.) Grube BJ, Marvin JA, Heimbach DM. Therapeutic Hyperbaric oxygen:
Help or hindrance in burn patients with carbon monoxide poisoning. J Burn
Care and Rehab 1988;9:249-52.

Presented at the 64th Annual Scientific Meeting and Postgraduate Course


Program, Southeastern Surgical Congress, Tampa, Florida, February 4-7, 1996.

Address correspondence and reprint requests to Edward J. Law, M.D., 1120


George C. Wilson Drive, Augusta, GA 30909.
~~~~~~~~

By A. LYNNE BRANNEN, M.D.,(*) JOSEPH STILL, M.D.,(*) MICHAEL HAYNES,


M.D.,(*) HERMANN ORLET, M.D., (*) FRED ROSENBLUM, M.D.,(*) EDWARD
LAW, M.D.,(*) WILLIAM O. THOMPSON, PH.D.(A), From the (*)Augusta
Regional Medical Center and (A)Department of Biostatistics, Medical College of
Georgia, Augusta, Georgia

Am Surg is copyrighted. Text may not be copied without the express written
permission of the publisher except for the imprint of the video screen content
or via the output options of the EBSCOhostsoftware. Text is intended solely for
the use of the individual user.

Você também pode gostar