Keywords
elevated intraocular pressure - orbital compartment syndrome - orbital decompression
- fluid resuscitation - parkland formula - burn
Intraorbital pressures, reflected by intraocular pressure (IOP) measurements, rapidly
rise regardless of etiology if swelling of the orbital soft tissues takes place. This
can lead to complications including orbital compartment syndrome (OCS), which is similar
to the condition described in abdominal compartment syndrome and compartment syndrome
of the extremities in orthopaedic cases.[1]
[2]
[3]
[4]
[5] Orbital compartment syndrome is a sight-threatening emergency and can rapidly lead
to irreversible vision loss. Several mechanisms have been proposed, though no clear
etiology has been proven.[1] The orbit is a nonexpandable space, surrounded by four bony walls superiorly, inferiorly,
medially, and laterally and a thick, fibrous orbital septum anteriorly. Vision loss
due to OCS is thought to be caused by elevated orbital pressure leading to occlusion
of posterior ciliary arteries, decreasing perfusion to the retina and optic nerve.[6]
Previous literature has suggested an association between orbital compartment syndrome
and severe burn patients who require large volume fluid resuscitation, which we define
as greater than two times the fluid volume recommended by the standard fluid resuscitation
formula.[7]
[8]
[9]
[10] In one study, nearly one-third of patients, with ≥ 25% total body surface area (TBSA)
burnt, required emergent surgical orbital decompression.[11] In several studies, justification of such treatment was supported by findings of
significant vision loss following severe burn and trauma.[1]
[2]
[12]
Fluid resuscitation is a cornerstone of acute burn injury management. Since its conception
in 1968, the Parkland formula, also known as the Baxter formula, has helped guide
clinicians in delivering acute resuscitative fluids. This formula was based on the
observation in both canine models and a pilot human population, that one could restore
cardiac output, extracellular fluid, and plasma volumes to near normal by delivering
intravenous (IV) fluids at a volume of 4.3 mL/kg/% TBSA over the first 24 hours after
thermal injury.[7] While many variations have been described, the Baxter formula remains the most commonly
cited fluid resuscitation guideline and has markedly decreased rates of mortality
previously attributed to burn shock, poor perfusion, and acute renal injury.
Recently, however, several studies have cited an increased incidence of resuscitation
with fluids that exceed the recommended volume calculated from the Baxter formula,
in a phenomenon termed “fluid creep.”[13] There are several hypothesized reasons described by Saffle including overresuscitation
by inexperienced health providers, overestimates of % TBSA involved, and unrestricted
administration of IV fluids in triage prior to admission to a burn center.[13]
[14] Furthermore, modern iterations of the Baxter formula have omitted use of colloid,
which was an original component of Baxter's formula in 1968 when he found that the
use of a plasma bolus at 24 hours postinjury was helpful in restoring extracellular
fluid balance.[13] This is potentially problematic because increased absolute volumes of IV fluid resuscitation
have been correlated with a rising incidence of edema-related complications including
pericardial effusions, compartmental compression in unburned extremities, abdominal
compartment syndrome, and elevated IOPs.[4]
[11]
[15]
In our experience at a major regional burn center, we have taken care of burn patients
who developed elevated IOPs but found that surgical intervention is rarely indicated.
The goal of this study is to better characterize the risk factors associated with
elevated IOP, including fluid resuscitation, and to discuss possible reasons why none
of our patients developed orbital compartment syndrome.
Methods
We retrospectively reviewed the medical records of burn patients from July 2008 to
June 2017 after receiving approval from the University of California San Diego (UCSD)
Human Research Protections Program Institutional Review Board. All patients were treated
at the UCSD Burn Intensive Care Unit (ICU). Patients ≥ 18 years of age with ≥ 25%
TBSA burn were included. All patients included in the review underwent fluid resuscitation
as calculated by the Baxter formula. Any patients with orbital trauma, hyphema, glaucoma,
third-degree periorbital burns, or chronic intraocular-lowering medications were excluded
from the study. Thus, 47 patients met study criteria.
Evaluation of pertinent medical data included eye exam findings, past history, age,
sex, race, past medications, concurrent medications and health problems, radiographic
images, treatment, and subsequent clinical course as documented in the patient chart.
We also recorded readmission, death, and posttreatment status of all patients at the
3-day end point.
Additionally, the UCSD Burn ICU meticulously kept track of fluid resuscitation for
each burn patient. In every case, the stated goal was to follow the Baxter formula
for fluid resuscitation. The recommended fluid volume from the Baxter formula was
calculated for every patient based on their admission weights and TBSA percentages.
The actual amount of fluids delivered to the patient were then compared with the amount
prescribed by the Baxter formula.
Periocular burns were defined as burns involving the eyelids. Mortality was defined
as failure to survive to hospital discharge. Peak IOP was defined as the single highest
IOP measured with a commercially available Tonopen (Reichert XL, Depew, NY). All IOP
measurements were recorded in mm Hg per eye per patient. The highest single IOP value
was recorded for a patient's right and left eyes, respectively, during the course
of hospital stay. These values were not necessarily obtained on the same day for a
given patient, but rather represent the single highest IOP value obtained over the
course of the patient's hospitalization.
Peak IOP and fluid levels were log transformed to reduce skewness. Linear regression
was used to evaluate whether there was an association between log-peak IOP and fluid
volume as measured by both total fluid (mL) and mL/kg/% TBSA burned. Logistic regression
was used to control for other characteristics when evaluating the association between
total fluid and treatment of high IOP. Continuous variables were compared using a
two-sided t-test and categorical variables were compared using a two-sided Fisher's exact test.
Tests were conducted at the α = 0.05 significance level. Patient data were analyzed using the R programming language
(v. 3.5.1).
Results
[Table 1] compares the demographics and clinical characteristics of the 47 patients included
in this study. The % TBSA ranged from 25 to 85% (mean, 45%). In every case, the Baxter
formula guided fluid resuscitation. No patient developed a relative afferent pupillary
defect, and none had additional evidence of optic nerve compromise such as impaired
color vision.
Table 1
Comparison of demographics and clinical characteristics of patients who were treated
for high-intraocular pressure versus those who were not
|
Characteristic
|
No treatment (n = 41)
|
Treated for high intraocular pressure (n = 6)
|
p[a]
|
|
Range or %
|
Mean ± SE
|
Range or %
|
Mean ± SE
|
|
|
Age (y)
|
20 − 76
|
46.0 ± 2.46
|
21 − 62
|
46.7 ± 6.63
|
0.925
|
|
Weight (kg)
|
50 − 136
|
86.1 ± 3.35
|
63.4 − 110
|
78.1 ± 6.89
|
0.3282
|
|
Peak IOP (mm Hg)
|
8 − 29
|
18.6 ± 0.709
|
21.5 − 37
|
30.2 ± 2.30
|
0.003
|
|
% female
|
34.1
|
N/A
|
33.3
|
N/A
|
1[b]
|
|
% periocular burn
|
26.8
|
N/A
|
83.3
|
N/A
|
0.01336[b]
|
|
% mortality
|
4.9
|
N/A
|
16.7
|
N/A
|
0.3426[b]
|
|
Baxter formula, predicted fluid (mL), 24 h
|
6,026 − 30,696
|
15,785 ± 1,011
|
1,886 − 29,240
|
15,442 ± 3,766
|
0.933
|
|
Total fluid, 24 h (mL)
|
1,738 − 37,254
|
11,597 ± 1,210
|
4,964 − 35,883
|
19,763 ± 4,569
|
0.1372
|
|
Total fluid, 48 h (mL)
|
3,000 − 31,706
|
16,369 ± 1,138
|
7,206 − 57,196
|
26,909 ± 6,974
|
0.1931
|
|
Total fluid, 72 h
|
3,812 − 39,137
|
13,912 ± 1,469
|
7,834 − 81,758
|
32,928 ± 11,314
|
0.1545
|
|
Fluid, 24 h, mL/kg/% TBSA burned
|
54.6 − 1,067
|
339 ± 35.9
|
196 − 1,071
|
489 ± 132
|
0.3155
|
|
Fluid, 48 h, mL/kg/% TBSA burned
|
127 − 1,643
|
520 ± 52.8
|
106 − 1,177
|
711 ± 168
|
0.3202
|
|
Fluid, 72 h, mL/kg/% TBSA burned
|
62.9 − 1,276
|
427 ± 52.1
|
124 − 1,775
|
889 ± 291
|
0.1745
|
Abbreviations: IOP, intraocular pressure; N/A, not available; TBSA, total body surface
area; SE, standard error.
a
p from t-test.
b
p from Fisher's exact test.
No patients developed orbital compartment syndrome or required surgical intervention
during or after their course of hospitalization. Therefore, we assessed clinical differences
among burn patients who required treatment for elevated IOP, defined as ≥ 30 mm Hg,
and those who did not. The peak IOP ranged from 8 to 37 mm Hg (mean, 20 mm Hg).
Of the 47 total patients, only six patients had IOPs high enough to warrant treatment
with IOP-lowering medications. The peak IOP of the treatment group was 21.5 to 37 mm
Hg (mean, 30.2 mm Hg) versus 8 to 29 mm Hg (mean, 18.6 mm Hg). The difference in peak
IOP between the two groups was statistically significant (p = 0.003).
In addition, presence of periocular burns was a significant risk factor in patients
who required treatment for high IOP (p = 0.001). Periocular burns were categorized into presence or absence of periocular
burns for analysis. Concerning the extent of injury, 16 patients (34%) had second-degree
eyelid burns, with the vast majority being bilateral; 15 (32%) had singed lashes,
but no second-degree burns. The remaining patients who sustained periocular burns
had varying degrees of facial involvement and/or eyelid edema from third spacing.
No patients sustained third-degree burns of the periorbita.
The patients who required treatment, five out of six were treated with one drop of
combination dorzolamide/timolol (Cosopt) in both eyes twice a day. One of these five
patients had IOPs on admission of 36 and 38 mm Hg and was started on brimonidine tartrate
three times a day in both eyes in addition to combination dorzolamide/timolol. Within
12 hours, the IOP had decreased to <30 mm Hg in both eyes, at which point brimonidine
tartrate was stopped. The patient was continued on combination dorzolamide/timolol
twice a day for a total of 48 hours until the IOPs came down to 7 and 9 mm Hg. The
sixth patient was prescribed one drop of latanoprost in both eyes at bedtime for one
day.
The relative fluid resuscitation volumes were assessed for differences between the
patient group treated for high IOP and the group that was not treated and are included
in [Table 1]. The amount of fluid recommended by the Baxter formula at 24 hours was not statistically
different between the two groups (p = 0.933). Total volume of fluid administered at 24, 48, and 72 hours were also not
statistically significant between the two groups. The total amount of fluid delivered
at 24 hours was higher in the group that required treatment, 19,763 ± 4,569 versus
11,597 ± 1,210 mL. This was not statistically significant (p = 0.1372), even after adjusting for other key covariates (p = 0.07; [Table 2]). However, logistic regression analysis found a statistically significant association
between risk of treatment for high IOP and periocular burns (p = 0.04).
Table 2
Association between total fluid at 24 hours (mL, natural log) and treatment of high-intraocular
pressure, before and after adjusting for key covariates
|
Covariate
|
p
[a]
|
|
Total fluid, 24 h (mL), unadjusted
|
0.1372[b]
|
|
Total fluid, 24 h (mL), adjusted
|
0.0731
|
|
Age
|
0.8459
|
|
Sex
|
0.4145
|
|
Weight (kg)
|
0.1679
|
|
Sex and weight (kg)
|
0.3285
|
|
Periocular burn
|
0.0356
|
a
p from logistic model for ln(fluid) and covariates.
b
p from t test.
Linear regression analysis found no significant correlation between peak IOP in the
first 48 hours and the fluid volume administered at 24 hours, both in total ml (r
2 = 0.026, p = 0.282) and mL/kg/% TBSA (r
2 = 0.041, p = 0.177). There were no statistically significant differences between the groups
in terms of age, sex, weight, or mortality.
The mean predicted volume of fluid resuscitation delivered, derived from the Baxter
equation (4.3 mL/kg/% TBSA burned), was 15,741 ± 986 mL. The mean total volume delivered
in the first 24 hours was 12,662 ± 1,254 or 356 ± 35.8 mL/kg/% TBSA burned (0.8 times
the level prescribed by the Baxter formula).
Roughly two-thirds of both groups received fluid resuscitation in accordance with
the Baxter formula. Only one patient in the study exceeded double the amount prescribed
by the Baxter formula at 24 hours but did not require treatment. Two others exceeded
double the amount within the first 48 hours. One underwent hemodialysis to correct
for overresuscitation, and the other died during hospitalization.
Discussion
Our results may offer insights on how to prevent the development of orbital compartment
syndrome in severely burn patients with proper management of fluid resuscitation volumes.
As a priority in our burn unit, the stated goal is to minimize deviations from the
Baxter formula for fluid resuscitation. The mean total volume delivered in the first
24 hours was 0.8 times the level predicted by the Baxter formula. No patients developed
orbital compartment syndrome during or after the course of hospitalization in our
study. Few even required topical pressure lowering agents, and none required surgical
intervention. Stricter adherence to the Baxter formula could potentially explain the
absence of orbital compartment syndrome in our burn patients, though this hypothesis
is suggestive and not conclusive. A previous study found that patients receiving more
than twice the fluids recommended by the Baxter formula were 4.4 times as likely to
develop severe orbital compartment syndrome.[11] However, our results detected that fluid volume was not a significant risk factor
for patients requiring topical IOP-lowering intervention. We believe that this could
be due to the small sample size of our cohort and that few patients required such
intervention which could have led to lower statistical power to replicate this link.
Only one patient received more than double the amount prescribed by the Baxter formula
in the first 24 hours and did not require IOP-lowering treatment. In another case
where the fluid resuscitation exceeded twice the Baxter formula within the first 48 hours,
hemodialysis was initiated to correct for overresuscitation. The last patient who
exceeded twice the Baxter formula in 48 hours expired during hospitalization. One
possible limitation of this study was that data collected reflected total fluid volume
replenishment and not the rate of fluid bolus in these patients. The rate of fluid
administration could be a factor in the development of orbital compartment syndrome,
but our study focused on total volume replacement.
The management of OCS has been well described.[1]
[16]
[17]
[18]
[19] IOP and vision may be fully restored to baselines following prompt diagnosis and
treatment. Medical management includes use of steroids, carbonic anhydrase inhibitors,
and mannitol. In emergent cases, lateral canthotomy and cantholysis and/or orbital
decompression can surgically reduce intraorbital pressure. Return of vision has been
reported after decompression suggesting return of perfusion.[1]
[16]
[19]
[20] Studies have shown immediate decrease in intraorbital pressure after canthotomy
and cantholysis, and case reports have demonstrated the efficacy of this procedure
to restore vision.[21]
[22]
[23]
[24] If the patient fails to show improvement, the orbit may be further decompressed
by opening of the orbital septum and/or bony expansion of the orbit by removing one
of the orbital walls.
In this study, we did not find the need for surgical options such as canthotomy and
cantholysis and orbital decompression. The presence of periocular burns, however,
was a significant risk factor associated with elevated IOP requiring pressure lowering
medication (p = 0.003). A total of 83.3% of patients who required treatment to lower IOP had periocular
burns compared with 26.8% of those who did not ([Table 1]). Logistic regression analyses also found a statistically significant association
between treatment of high IOP and presence of periocular burns (p = 0.04; [Table 2]). This suggests that periocular burns could play a significant role in increasing
IOP, which may be a reflection of underlying high intraorbital pressure or a primary
globe side effect of the thermal injury. Five out of six patients who required treatment
received one drop of combination dorzolamide/timolol (Cosopt) in both eyes twice a
day. Within their hospital stay, all patients had IOPs less than 20 mm Hg with topical
treatment. In our opinion, topical treatment of high IOPs could be considered part
of standard ophthalmic management in these patients and may be related to globe trauma
and anterior lid involvement, rather than reflecting deeper impending orbital compartment
syndrome.
Conclusion
It is important to note the relative geographic and population size differences between
regional burn centers. The UCSD Regional Burn Center may have a different population
than centers with higher percentages of orbital compartment syndrome. These burn centers
may have larger geographic catchment areas and rural locales from which they accept
patients. Since the first 24 hours of fluid resuscitation is of paramount importance,
it is possible that a higher relative reliance on triage providers and emergency medical
personnel unfamiliar with the Baxter protocol may account for some of the observed
differences. Future studies may explore whether this association may hold true. The
results of this study show promise that orbital compartment syndrome can be prevented
with strict fluid replenishment, close surveillance, and treatment of ocular issues
early in the treatment course.
