Intravenous injection of methylprednisolone reduces the incidence of postextubation stridor in intensive care unit patients
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Clinical Investigations
Intravenous injection of methylprednisolone reduces the incidence
of postextubation stridor in intensive care unit patients*
Kuo-Chen Cheng, MD; Ching-Cheng Hou, MD; Heng-Ching Huang, MD; Shu-Chih Lin, RRT;
Haibo Zhang, MD, PhD
Objective: To determine whether treatment with corticoste- tion. Postextubation stridor was confirmed by examination using
roids decreases the incidence of postextubation airway obstruc- bronchoscopy or laryngoscopy.
tion in an adult intensive care unit. Measurements and Main Results: The incidences of postextu-
Design: Clinical experiment. bation stridor were lower both in the 1INJ and the 4INJ groups
Setting: Adult medical and surgical intensive care unit of a than in the control group (11.6% and 7.1% vs. 30.2%, both p <
teaching hospital. .05), whereas there was no difference between the two treated
Patients: One hundred twenty-eight patients who were intu- groups (p ⴝ .46). The cuff leak volume increased after the second
bated for >24 hrs with a cuff leak volumetional Research Ethic Board. Informed con- nate (Pharmacia Upjohn, Kalamazoo, MI) at confidence interval. To find an optimal cutoff
sent was obtained from the patients or their 40 mg in 2 mL of normal saline every 6 hrs value of CLV predictive of postextubation stri-
relatives before entering the study. over 24 hrs; the 1INJ group received one in- dor, a receiver operating characteristic curve
All patients were ⬎18 yrs of age, had been fusion of methylprednisolone sodium succi- was used. We considered p ⬍ .05 to be statis-
intubated ⬎24 hrs, and met the weaning cri- nate followed by three injections of normal tically significant.
teria defined as respiratory rate ⬍30 breaths/ saline every 6 hrs over 24 hrs; and a placebo
min, negative inspiratory force ⬎25 cm H2O, group received intravenous infusion of normal RESULTS
tidal volume ⬎5 mL/kg of ideal body weight, saline in 2 mL every 6 hrs over 24 hrs before
and rapid shallow index ⬍105 breaths/min/L. extubation. Nonintervention Arm. A total of 334
Exclusion criteria were any treatment with Extubation was performed 1 hr after the patients were included for the cuff-leak test;
corticosteroids 1 wk before extubation (n ⫽ last injection of intervention fluid. Cuff leak 13 patients were withdrawn due to deteri-
164), nasal or throat disease/surgery (n ⫽ 9), volume was not used as a criterion for extu- oration of clinical conditions resulting in a
gastrointestinal bleeding (n ⫽ 33), hypergly- bation but was used afterward for data analy- delayed extubation (n ⫽ 12) or self-
cemia (blood sugar ⬎250 mg%, n ⫽ 19), acute sis. Patients were monitored for 48 hrs after
extubation (n ⫽ 1). Thus, 321 patients were
cardiac attack and cardiac surgery (n ⫽ 36), or extubation. The clinical stridor symptom was
history of extubation during the same hospi- defined by the presence of an audible high- enrolled in the study protocol (Fig. 1). Of
talization course (n ⫽ 78). The reason to ex- pitched wheeze associated with respiratory 321 patients, 193 patients who had a CLV
clude patients with gastrointestinal bleeding distress requiring medical intervention (10, ⱖ24% of tidal volume during inflation
or hyperglycemia was to better distinguish 14, 15). Epinephrine inhalation and, in case of served as a nonintervention arm. The cutoff
between underlying diseases and steroids- failure in response to three doses of epineph- of 24% was based on the study by Sandhu
induced complications if any. rine inhalation, bilevel positive airway pres- and coworkers (9). The other 128 patients
Auscultation Cuff-Leak Test. Auscultation sure (Respironics, Murrysville, PA) were given with a CLV ⬍24% served as an intervention
cuff-leak and cuff-leak volume were examined to treat stridor, and reintubation was per- arm and were randomized into a placebo
and classified into three categories: a) no leak, formed in patients who were not improved group (n ⫽ 43), a 4INJ group (n ⫽ 42), or
where no sound of leak was heard by using under epinephrine inhalation or/and bilevel
a 1INJ group (n ⫽ 43). The incidence of
stethoscope detection; b) mild leak, where a positive airway pressure treatment. In the in-
leak was heard using a stethoscope; and c) tervention arm, only patients with stridor un- postextubation stridor was 2.6% in the
significant leak, where the sound of a leak was derwent either bronchoscopy or laryngoscopy nonintervention arm and 30.2% in the pla-
heard without using a stethoscope. to further identify the occurrence of laryngeal cebo group of the intervention arm. This
Cuff Leak Volume. The actual tidal volume edema by an in-charge physician during rein- suggests that the cutoff point of 24% of
at expiration was measured before and after tubation. CLV reliably predicts the development of
deflation of the endotracheal tube cuff as pre- The patients in the nonintervention arm postextubation stridor in this popula-
viously described (9). The difference in the underwent extubation immediately after the tion of ICU patients.
actual tidal volume before and after cuff defla- cuff-leak test. In patients of the intervention To identify the critical point of CLV in
tion was defined as the cuff leak volume (CLV). arm, extubation was performed 1 hr after the predicting the development of postextu-
Mechanical Ventilation. The patients were last injection of drug or placebo saline depend-
bation stridor, we retrospectively pre-
mechanically ventilated in a volume assist- ing on the study protocol enrolled.
Statistics. Patient characteristics were an- pared and analyzed a receiver operating
control mode (Puritan-Bennett 7200 AE,
Carlsbad, CA, or Bird 8400, Palm Springs, CA). alyzed using the unpaired Student’s t-test for characteristic curve and found that 18%
The tidal volume was set at 8 mL/kg ideal body continuous variables. Fisher’s exact test was was the best predictor in our patient pop-
weight with respiratory rate 20 breaths/min used for analysis of categorical variables. To ulation in the present study (Fig. 2).
and zero positive end-expiratory pressure dur- estimate the risk of developing postextubation There was an excellent agreement be-
ing CLV measurement. stridor, logistic regression analysis was ap- tween the auscultation cuff-leak test and
Protocol. Patients with a CLV ⱖ24% of plied. In univariate analysis, crude odds ratio the CLV (Fig. 3).
tidal volume during inflation were the time and a 95% confidence interval of odds ratio Since the robustness of a multivariate
matching control arm for 48 hrs after extuba- was estimated for each risk factor we studied. logistic model depends on the inclusion of
tion (nonintervention arm). This cutoff point A forward stepwise method was then used to
all relevant variables, we particularly exam-
was set based on a previous report where 13 of construct the best final model, which includes
the risk factors significantly correlated to de- ined if the level of training of the operator
110 patients who had CLVs ranging from 0%
to 24% developed postextubation stridor (9). veloping stridor and leaves out those do not to intubate patients was a significant risk
Patients who had a CLV ⬍24% of tidal improve the model. In multivariate analysis, factor to cause postextubation stridor. It
volume during inflation were considered at an independent effect of each risk factor was appeared that residents were the most fre-
high risk to develop postextubation stridor (9, estimated with adjusted odds ratio and 95% quent operators of intratracheal intubation
10) and were included in the randomized,
double-blinded, and placebo-controlled clini-
cal trial (intervention arm). The randomiza-
tion procedure was carried out with random
numbers. The placebo and methylpred-
nisolone had an identical volume and appear-
ance and were packaged in an identical man-
ner by a respiratory therapist who was not
involved in the trial. Neither the physician
who prescribed the drug nor the staff who
administered the infusion were aware of the
intervention treatment of infusion fluid.
These patients were divided into three
groups: The 4INJ group received intravenous Figure 1. Distribution of groups of patients. CLV, cuff leak volume; 1INJ, one injection of methyl-
infusion of methylprednisolone sodium succi- prednisolone; 4INJ, four injections.
1346 Crit Care Med 2006 Vol. 34, No. 5Table 1. Patients’ characteristics and intratracheal intubation operators in the intervention arm
Control (n ⫽ 43) 1INJ (n ⫽ 42) 4INJ (n ⫽ 42) p Value
Gender, M/Fa 15/28 15/28 19/23 .523
Age, yrsb 68 ⫾ 16.1 63.1 ⫾ 16 67.3 ⫾ 17.6 .333
APACHE IIb 18.3 ⫾ 14.2 17.3 ⫾ 7 17.0 ⫾ 5.3 .638
Glasgow Coma Scaleb 10.2 ⫾ 3.9 9.0 ⫾ 3.9 9.6 ⫾ 3.5 .394
Duration of intubation, hrsb 169.5 ⫾ 98.8 175.7 ⫾ 93.5 149.8 ⫾ 91.8 .425
Endotracheal tube depth, cmb 21.3 ⫾ 1.7 21.4 ⫾ 1.6 21.4 ⫾ 1.5 .934
Patients sourcea .183
Medical ICU 25 17 18
Surgical ICU 18 27 24
Intubation operatora .102
Attending physician 10 3 5
Figure 2. Receiver operating characteristics curve Resident 22 28 25
to predict the occurrence of postextubation stridor Anesthesiologist 5 8 11
by measurement of cuff leak volume. A value of Unknown prior to admission 6 4 1
18% was considered as a cutoff point, with sensi-
tivity of 85.3% and specificity of 72.2%. Control, patients received placebo; 1INJ, patients received one injection of sodium succinate; 4INJ,
patients received four injections of the drug.
a
Fisher’s exact test; banalysis of variance.
Table 2. Risk factors of developing postextubation stridor in untreated patients
Crude OR (95% CI) p Value
Sedation
Yes 1.00
No 3.40 (1.10, 10.54) .034
Gender
Male 1.00
Female 6.42 (2.30, 17.96) .0004
GCS
13–15 1.00
9–12 4.04 (0.71, 22.87) .114
3–8 6.91 (1.42, 33.70) .017
Figure 3. Relationship between auscultation cuff- Age, yrs
leak test and cuff leak volume (%). No leak, no ⬍60 1.00
sound of leak was heard by using stethoscope 60–69 0.41 (0.05, 3.61) .422
detection; mild leak, leak was heard by using 70–79 1.96 (0.57, 6.72) .287
stethoscope; large leak, sound of leak was heard ⱖ80 3.11 (0.89, 10.85) .076
without using stethoscope. *p ⫽ .007, small leak Intubation time, hrs 1.00 (1.00, 1.00) .797
Endotracheal tube size, Fr
vs. no leak; p ⬍ .001, large leak vs. small leak and
ⱕ7.0 1.00
no leak. 7.5–8.0 0.86 (0.32, 2.30) .762
Intubation unit
Operating room 1.00
Emergency room 3.84 (0.77, 19.18) .101
compared with attending physicians and ICU 3.15 (0.59, 16.86) .18
anesthesiologists, but there was no differ- General ward 8.5 (1.43, 50.54) .019
ence with respect to the background train- Prior to admission 0.00 (0.00) .804
ing of operators before patient randomiza- APACHE II 1.06 (0.95, 1.17) .299
tion into treated and nontreated groups OR, odds ratio; CI, confidence interval; GCS, Glasgow Coma Scale; ICU, intensive care unit;
(Table 1). APACHE, Acute Physiology and Chronic Health Evaluation.
A number of other variables including Data from the untreated patients of the nonintervention arm (n ⫽ 193) and the control group of
the use of sedation, Glasgow Coma Scale the intervention arm (n ⫽ 43).
(GCS), Acute Physiology and Chronic
Health Evaluation II score, gender, and age Intervention Arm. The patients were after the second and the fourth injection
were analyzed to identify the most sensitive similar in age, gender, Acute Physiology (both p ⬍ .05) in the 4INJ group and after
risk factors contributing to the develop- and Chronic Health Evaluation II score, the second injection in the 1INJ group
ment of postextubation stridor. It is sug- duration of intubation, endotracheal tube (p ⬍ .05, Fig. 4).
gested that nonsedation treatment, low size, and types of ICU (medical vs. surgical) We found that 18% of CLV was the
GCS score, and being female were risk fac- where the patients were treated (Table 1). optimal predictor for stridor develop-
tors to develop postextubation stridor in In the placebo group, approximately a ment; we then analyzed the frequency of
the untreated patients (n ⫽ 236, including 4% increase from baseline in CLV was increased CLV ⬎18% from baseline after
patients in the nonintervention arm [n ⫽ observed at 24 hrs. In the groups treated each injection of methylprednisolone, al-
193] and the control group [n ⫽ 43] in the with methylprednisolone, a significant though the decision of extubation was
intervention arm) (Table 2). increase from baseline in CLV was seen not based on the values of measured CLV.
Crit Care Med 2006 Vol. 34, No. 5 1347Table 4. Intubation duration and intensive care unit (ICU) stay in relation to Glasgow Coma Scale
(GCS) score in the intervention arm
GCS 3–8 (n ⫽ 51) 9–15 (n ⫽ 55) p Value
Intubation duration, hrs 178.5 ⫾ 85.0 149.2 ⫾ 88.81 .086
Stridor 10 4 .061
Control 6 3
1INJ 2 0
4INJ 2 1
ICU stay, days 12.4 ⫾ 6.5 11.4 ⫾ 7.7 .485
Hospital stay, days 30.3 ⫾ 12.8 31.6 ⫾ 21.4 .701
Control, patients received placebo; 1INJ, patients received one injection of sodium succinate; 4INJ,
patients received four injections of the drug.
Figure 4. Differences in cuff leak volume (%)
changes with frequency of injections. Control
group received four injections of normal saline As many as 19% patients had increased postextubation stridor by suppressing
every 6 hrs; the one injection of methylpred-
CLV ⬎18% 12 hrs after treatment in the mucosal inflammation including inhibi-
nisolone (1INJ) group received an initial injec-
tion of methylprednisolone followed by three in-
1INJ group, and 40 – 60% of patients tion of leukocyte migration, maintenance
jections of normal saline; and the 4INJ group showed increased CLV ⬎18% during a of cell membrane integrity, attenuation
received four injections of methylprednisolone. prolonged treatment in the 4INJ group of lysosome release, and reduction of fi-
*p ⬍ .05 vs. control group; †p ⬍ .05 4INJ group (Table 3). broblast proliferation (19, 20) and tissue
vs. control group. The changes in CLV were consistent swelling (21). However, intravenous infu-
with the incidence of postextubation stri- sion of -methasone had no protective
dor. As many as 30.2% of patients devel- effect in attenuation of postextubation in-
oped postextubation stridor in the pla- flammation in pediatric ICU (22). In ex-
Table 3. Frequency of increased cuff leak volume
cebo group compared with an incidence perimental settings, dexamethasone ef-
⬎18% after each injection from baseline as low as 7.1% and 11.6% in the 4INJ fectively suppressed postintubation laryn-
group and the 1INJ group, respectively geal edema in monkeys (23).
Number of Injections (both p ⬍ .05, Fig. 5). The rate of rein- There are few data in adult patients
tubation was 18.6% (eight of 43) in the compared with pediatric patients with re-
1st 2nd 3rd 4th placebo group compared with 7.1% spect to the use of corticosteroids. A bo-
(three of 42) in the 4INJ group and 4.7% lus intravenous injection of dexametha-
Control 11 8 13 14
1INJ 13 19 16 17 (two of 43) in the 1INJ group (both p ⬍ sone or hydrocortisone 1 hr before
4INJ 16 40 53 60 .05). Neither the CLV nor the incidence of extubation failed to prevent laryngeal
postextubation stridor was different be- edema in adults (4, 12). The discrepancies
Control, patients received placebo; 1INJ, pa- tween the two treated groups (Fig. 1).
tients received one injection of the drug; 4INJ,
observed in these previous studies could
Since a low GCS was associated with an be due to several factors including age,
patients received four injections of the drug. increased risk of stridor, one may be con-
Data are reported as percent patients with inclusion criteria, duration of intubation,
cerned that exposure of this population of dose, timing and length of treatment, and
increase in cuff leak volume ⬎18% from baseline
patients to the present protocol could place
(prior to injection of methylprednisolone sodium risk levels of developing stridor.
succinate). them at a risk of inhalation. Importantly,
Cuff leak volume has been considered
our data showed no difference in the
as a good indicator to predict the occur-
lengths of ICU stay and hospital stay be-
rence of postextubation stridor. Sandhu
tween the small size groups of patients with
et al. (9) found that of 110 patients stud-
low and high GCS on completion of the
ied, all 13 patients (12%) who developed
study protocol (Table 4).
postextubation stridor had a CLV ranging
from 0% to 24%. Their data are sup-
DISCUSSION ported by the present study, where the
Endotracheal intubation is a useful incidence of postextubation stridor was
approach to manage critically ill patients 2.6% (five of 193 patients) if CLV was
who may or may not require mechanical ⱖ24% and increased to 30.2% (13 of 43
ventilation. However, this intervention patients) with CLV ⬍24%. The difference
may induce laryngeal mucosa trauma as a (12% vs. 30%) in the incidence of post-
result of mucosa abrasion and necrosis extubation stridor seen at an identical
secondary to compression, which leads to cuff leak cutoff point of ⬍24% between
postextubation stridor (2, 3, 16, 17). His- the Sandhu et al. and the present study
Figure 5. Incidence of postextubation stridor
tologic findings of postextubation stridor may be explained by a population differ-
(PES) in the intervention arm (cuff leak volume
⬍24%) of the study. *p ⫽ .015, 1INJ group vs. are characterized with mucosal inflam- ence, because only trauma patients were
control group; †p ⫽ .005, 4INJ group vs. control mation, mucosal ulceration, and edema included in the former study and all kinds
group. 1INJ, one injection of methylpred- (18). Corticosteroids have been used in of ICU patients were included in the
nisolone; 4INJ, four injections. infants and children to prevent or reduce present study. Thus, the different inci-
1348 Crit Care Med 2006 Vol. 34, No. 5dences of postextubation stridor seen in leak tests after each injection of methyl- REFERENCES
the studies may reflect the severity of prednisolone, since the conditions of the
illness. Taken together, these observa- patients were relatively stable, where all 1. Natanson C, Shelhamer JH, Parrillo JE: In-
tions suggest that examination of CLV is patients met the extubation criteria be- tubation of the trachea in the critical care
a useful approach to identify patients at fore being included in the current proto- setting. JAMA 1985; 253:1160 –1165
2. Stauffer JL, Olson DE, Petty TL: Complica-
high risk of developing postextubation col. We credit the protective effect of
tions and consequences of endotracheal in-
stridor. methylprednisolone demonstrated in the tubation and tracheotomy. A prospective
In the present study, a CLV of 18% was present study to two major factors: the study of 150 critically ill adult patients.
the best cutoff value to predict postextu- homogeneous selection of patient popu- Am J Med 1981; 70:65–76
bation stridor based on its 85.3% of sen- lation by using the cuff-leak test, and a 3. Kastanos N, Estopa Miro R, Marin Perez A, et
sitivity at 18% CLV compared with 79.7% prolonged therapy with methylpred- al: Laryngotracheal injury due to endotra-
of sensitivity at 24% CLV. However, there nisolone compared with previous studies cheal intubation: Incidence, evolution, and
is no significant difference in the speci- by other investigators (4, 12). predisposing factors. A prospective long-
ficity (72.2% vs. 77.8%) between the A few risk factors were identified in term study. Crit Care Med 1983; 11:362–367
CLVs at 18% and 24%. Furthermore, al- the present study including the use of 4. Ho LI, Harn HJ, Lien TC, et al: Postextuba-
tion laryngeal edema in adults. Risk factor
though the CLV ⬎24% is very useful in sedation before intubation, GCS score,
evaluation and prevention by hydrocorti-
predicting the absence of stridor and CLV and gender. First, the use of sedation in sone. Intensive Care Med 1996; 22:933–936
⬍24% identifies a high-risk group, 70% advance of intubation may help reduce 5. Jaber S, Chanques G, Matecki S, et al: Post-
of control participants did not develop patients’ anxiety that would have caused extubation stridor in intensive care unit pa-
stridor. We speculate that this gap seen them to fight against intubation, result- tients. Risk factors evaluation and impor-
between the predicted high incidence and ing in vocal cord damage (24), as the tance of the cuff-leak test. Intensive Care
the actual low rate might be due to a quality of endotracheal intubation can Med 2003; 29:69 –74
number of potential mechanisms includ- contribute to laryngeal hoarseness and 6. Epstein SK, Ciubotaru RL: Independent ef-
ing secretions adherent to the outside of vocal cord squeal (25). Second, a low GCS fects of etiology of failure and time to rein-
the endotracheal tube and differences in score (3– 8 points) was associated with a tubation on outcome for patients failing ex-
tubation. Am J Respir Crit Care Med 1998;
measurements between inspired tidal vol- high incidence of postextubation stridor,
158:489 – 493
ume and inhaled tidal volume around the which may result from a reduced ability 7. Demling RH, Read T, Lind LJ, et al: Inci-
endotracheal tube. to cough and a more frequent require- dence and morbidity of extubation failure in
Other studies reported CLVs of 10% ment of intratracheal suction. Third, con- surgical intensive care patients. Crit Care
and 12% as cutoff points to predict the sistent with our results, several other in- Med 1988; 16:573–577
development of postextubation stridor (5, vestigators also reported that female 8. Marik P: The cuff-leak test as a predictor of
8 –10). The difference in cutoff points patients had a higher risk to develop stri- postextubation stridor: A prospective study.
among the studies may be due to differ- dor compared with male patients (4, 21, Respir Care 1996; 41:509 –511
ences in tidal volume used during mea- 26). This may be related to anatomy, as 9. Sandhu RS, Pasquale MD, Miller K, et al:
surement of cuff leak, as we used a tidal Measurement of endotracheal tube cuff leak
the larynx is smaller, the mucosal cover-
to predict postextubation stridor and need
volume of 8 mL/kg compared with a tidal age is thinner, and the mucosal mem- for reintubation. J Am Coll Surg 2000; 190:
volume of 10 –12 mL/kg in other studies brane is less resistant to trauma in female 682– 687
(5, 8 –10). We demonstrated an excellent than in male subjects (12, 27). 10. Miller RL, Cole RP: Association between re-
correlation between the measurements of The side effects with steroid therapy duced cuff leak volume and postextubation
auscultation cuff leak test and CLV. The over 24 hrs are minimal (21). In the stridor. Chest 1996; 110:1035–1040
former technique appears to be more present study there were no obvious com- 11. Fisher MM, Raper RF: The “cuff-leak” test for
convenient to perform at bedside and is plications such as gastrointestinal bleed- extubation. Anaesthesia 1992; 47:10 –12
without risk of developing aspiration ing, high blood sugar, or increased risk of 12. Darmon JY, Rauss A, Dreyfuss D, et al: Eval-
pneumonia. infection. uation of risk factors for laryngeal edema
Treatment with intravenous injection after tracheal extubation in adults and its
prevention by dexamethasone. A placebo-
of methylprednisolone sodium succinate
CONCLUSIONS controlled, double-blind, multicenter study.
resulted in a significant increase in CLV
Anesthesiology 1992; 77:245–251
after the second injection both in the 13. Anene O, Meert KL, Uy H, et al: Dexametha-
4INJ group and in the 1INJ group where Our study reveals that the use of a
sone for the prevention of postextubation
normal saline was used as a second injec- cutoff point of CLV can help identify pa-
airway obstruction: A prospective, random-
tion. This observation suggests that a) a tients who have a high risk of developing ized, double-blind, placebo-controlled trial.
time frame of 6 –7 hrs after a single in- postextubation stridor. Either a single or Crit Care Med 1996; 24:1666 –1669
jection is required to exert the protective multiple injections of methylpred- 14. Daley BJ, Garcia-Perez F, Ross SE: Reintu-
effect of methylprednisolone in reducing nisolone 6 hrs before extubation can ef- bation as an outcome predictor in trauma
the incidence of postextubation stridor; fectively reduce the incidence of postex- patients. Chest 1996; 110:1577–1580
and b) injections with either four sepa- tubation stridor in critically ill adult 15. Chevron V, Menard JF, Richard JC, et al:
patients. Unplanned extubation: Risk factors of devel-
rate doses or a single initial dose of meth-
opment and predictive criteria for reintuba-
ylprednisolone do not make any signifi- tion. Crit Care Med 1998; 26:1049 –1053
cant difference in the incidence of ACKNOWLEDGMENT 16. Bishop MJ, Weymuller EA Jr, Fink BR: La-
postextubation stridor. We do not antici- ryngeal effects of prolonged intubation.
pate any significant changes in respira- We thank Chin-Li Lu for statistical Anesth Analg 1984; 63:335–342
tory mechanics over the period of cuff assistance. 17. Bishop MJ: Mechanisms of laryngotracheal
Crit Care Med 2006 Vol. 34, No. 5 1349injury following prolonged tracheal intuba- AG, Rall TW, Nies AS (Eds). New York, Per- Does the sedative agent facilitate emergency
tion. Chest 1989; 96:185–186 gamon Press, 1990, pp 1431–1462 rapid sequence intubation? Acad Emerg Med
18. Way WL, Sooy FA: Histologic changes pro- 21. Hawkins DB, Crockett DM, Shum TK: Corti- 2003; 10:612– 620
duced by endotracheal intubation. Ann Otol costeroids in airway management. Otolaryn- 25. Mencke T, Echternach M, Kleinschmidt S, et
1965; 74:799 – 813 gol Head Neck Surg 1983; 91:593–596 al: Laryngeal morbidity and quality of tra-
19. Demling W, Oech SR: Steroid and antihista- 22. Goddard JE Jr, Phillips OC, Marcy JH: Beta- cheal intubation: A randomized controlled
minic therapy for postintubation subglottic methasone for prophylaxis of postintubation trial. Anesthesiology 2003; 98:1049 –1056
edema. Anesthesiology 1961; 22:933–936 inflammation. A double-blind study. Anesth 26. Gaynor EB, Greenberg SB: Untoward se-
20. Haynes RC: Adrenocorticotropic hormones, Analg 1967; 46:348 –353 quelae of prolonged intubation. Laryngo-
adrenocortical steroids and their synthetic 23. Biller HF, Bone RC, Harvey JE, et al: Laryn- scope 1985; 95:1461–1467
analogues; inhibitors of the synthesis and geal edema: An experimental study. Ann Otol 27. Whited RE: A prospective study of laryngo-
actions of adrenocortical hormones. In: The Rhinol 1970; 174:1084 –1087 tracheal sequelae in long-term intubation.
Pharmacological Basis of Therapeutics. Gilman 24. Sivilotti ML, Filbin MR, Murray HE, et al: Laryngoscope 1984; 94:367–377
1350 Crit Care Med 2006 Vol. 34, No. 5You can also read
