A repeat literature search showed no new trials eligible for inclusion and there have been no substantive changes to the review.
Dexamethasone may help babies at high risk of complications when being taken off mechanical breathing support.
The tube in the baby's airway that enables mechanical ventilation (machine-assisted breathing) can cause injury. This can lead to complications when the tube is removed (extubation). Giving dexamethasone (a corticosteroid drug) around the time of extubation can sometimes help prevent complications. The review found that as there are adverse effects of dexamethasone, the benefits only outweigh the risks for babies at high risk of complication (such as those who have received several, or prolonged, intubations).
Endotracheal tubes are foreign bodies that may injure the upper airway causing laryngeal edema. This in turn may result in failure of extubation in preterm infants. Corticosteroids have been used prophylactically to reduce upper airway obstruction and facilitate extubation.
In newborn infants having their endotracheal tube removed following a period of intermittent positive pressure ventilation (IPPV), what are the effects of intravenous corticosteroids on the incidence of endotracheal reintubation, stridor, atelectasis and adverse side effects?
Searches were made of the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library) (dexamethasone and extub*), MEDLINE (MeSH search terms "dexamethasone", "extubat*" and "exp infant, newborn"), previous reviews including cross references, abstracts of conferences and symposia proceedings, expert informants and journal handsearching mainly in the English language. These searches were updated in October 2004.
Trials were included which used random or quasi-random patient allocation, and which compared intravenous steroids given immediately prior to a planned extubation with placebo.
Data were extracted independently by the two authors and analysed in Revman for all trials. Prespecified subgroup analyses were performed to examine differences in response between infants at high risk for upper airway edema and those receiving routine prophylaxis prior to extubation.
Administration of dexamethasone prior to extubation significantly reduced the need for reintubation of the trachea. This result applies to both the high risk group and to the total population of infants enrolled. However, the incidence of extubation failure was zero in the trial that attempted to exclude infants at high risk of airway edema. The side effects of higher blood sugar levels and glycosuria were found in the 2 trials where these were sought.
Implications for practice
Dexamethasone reduces the need for endotracheal reintubation of neonates
after a period of IPPV. In view of the lack of effect in low risk infants
and the documented and potential side effects, it appears reasonable to restrict
its use to infants at increased risk for airway edema and obstruction, such
as those who have received repeated or prolonged intubations.
Implications for research
Issues of dosage and applicability to the extremely low birthweight population
could be addressed in future trials. Longer term outcomes such as chronic
lung disease, duration of assisted ventilation and length of hospital stay
should also be examined.
Endotracheal intubation is used to provide intermittent positive pressure ventilation (IPPV) for a number of neonatal conditions. The presence of a foreign body in contact with delicate upper airway mucosa can lead to injury. This may take the form of laryngeal oedema, vocal cord injury or subglottic stenosis, all of which may present clinically as upper airway obstruction after extubation (Arensman, 1988). This may in turn lead to increasing respiratory distress requiring reintubation of the trachea. Factors that may increase the likelihood of damage include repeated passage of an endotracheal tube, prolonged intubation and the presence of a large tube relative to the size of the glottis. Various agents including systemic corticosteroids have been used both prophylactically and as treatment to reduce upper airway obstruction and facilitate extubation (Carlo 1992).
In addition, corticosteroids may also have a beneficial effect on the lower airways by decreasing oedema and secretions due to the toxic effects of oxygen and IPPV. Increased compliance of the lung has been observed following corticosteroid treatment (Durand 1995). A reduction of respiratory distress and consequent pressure gradients across the upper airway may decrease the tendency towards upper airway narrowing during extubation.
This review updates the existing review of intravenous dexamethasone in neonates being extubated which was published in the Cochrane Library (Davis 1999).
To determine whether the use of intravenous corticosteroids improves the post-extubation course of newborn infants as evidenced by effects on the incidence of endotracheal reintubation, stridor, atelectasis and adverse side effects. Subgroup analysis was performed to determine whether there was a difference in response between the group of newborn infants as a whole and those thought to be at high risk of upper airway obstruction because of prolonged intubation or traumatic or multiple intubations. Sensitivity analysis was planned to investigate the role of potential confounding variables such as the use of nasal CPAP post-extubation.
All trials utilising random or quasi-random patient allocation were included.
Newborn infants who received Intermittent Positive Pressure Ventilation (IPPV) and were about to have their endotracheal tube removed.
Intravenous administration of corticosteroid.
The primary outcome of interest was the need for reintubation of the trachea. Secondary outcomes included
· the presence of clinical signs suggestive of upper airway obstruction (stridor, respiratory distress, atelectasis) and
· adverse effects such as hyperglycemia and hypertension.
Searches were made of the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 1, 2003) dexamethasone and extub*), MEDLINE (1966 - April 2003, MeSH search terms "dexamethasone", "extubat*" and "exp infant, newborn"), previous reviews including cross references, abstracts of conferences and symposia proceedings, expert informants, journal handsearching mainly in the English language and expert informant searches in the Japanese language by Prof. Y. Ogawa.
Criteria and methods used to assess the methodological quality of the included trials:
The standard method of the Cochrane Collaboration and its Neonatal Review
Group were used. The methodological quality of each trial was reviewed by
the second author blinded to trial authors and institutions.
Methods used to collect data from the included trials:
Each author extracted data separately before comparison and resolution of differences.
Methods used to synthesize data:
Standard method of the Neonatal Review Group with the use of relative
risk (RR), risk difference (RD) and number needed to treat (NNT).
Subgroup analysis was performed to determine whether there was a difference
in response between the group of newborn infants as a whole and those thought
to be at high risk of upper airway obstruction because of prolonged intubation
or traumatic or multiple intubations. Sensitivity analysis was planned to
investigate the role of potential confounding variables such as the use of
nasal CPAP post-extubation.
Three randomized trials were identified which addressed the effects of treatment with dexamethasone to facilitate extubation of newborn infants following a period of IPPV. A full description of each is included in the Table, Characteristics of Included Studies. No trials were excluded from the review and no ongoing trials were identified.
The participants in the three trials differed with respect to risk of airway edema. Couser 1992 selected a population considered at high risk because of either traumatic or multiple intubations or intubation for more than 14 days. This group had a mean weight at entry of around 1200g. Ferrara 1989 specifically excluded patients intubated more than once and enrolled infants with a mean duration of intubation of less than a week. His patients comprised a heavier (mean > 2kg) and more mature (mean > 33 weeks) population. The trial of Courtney 1992 also studied a heavier (mean extubation weight 2200-2400g), more mature population (mean 32-34 weeks gestational age). There was no effort to include or exclude infants on the basis of perceived risk of airway edema although an upper limit of 30 days of intubation was placed to minimise the effect of subglottic stenosis.
All three trials used dexamethasone intravenously. Courtney 1992 used 3 doses of 0.5 mg/kg, Couser 1992 3 doses of 0.25mg/kg and Ferrara 1989 used a single dose of 0.25 mg.
The primary outcome, need for endotracheal intubation, was assessed in all trials. The presence of stridor was noted by Couser 1992 and Ferrara 1990 and atelectasis on CXR reported by Ferrara 1989 and Courtney 1992. Side effects of dexamethasone including hypertension and hyperglycemia were evaluated by Couser 1992 and Courtney 1992.
All the trials are of high quality and full details are provided in the Table, Characteristics of Included Studies.
Method of subject allocation:
All three included studies were randomized controlled trials. Only Courtney 1992 specified the use of a random number table and this trial and Couser 1992 used off site randomization.
Masking of caregivers
The use of a normal saline placebo ensured caregivers were unaware of
group of assignment. In view of the systemic side effects it is possible
that the actual treatment group could have been suspected in some cases.
Completeness of outcome assessment
Couser 1992 and Ferrara 1989 had complete followup of all randomized infants for all outcomes. Courtney 1992
excluded nine randomized infants from studies of pulmonary mechanics. However
all were included for the clinically important outcomes of endotracheal reintubation
and atelectasis.
Masking of outcome assessors
This was achieved in all trials.
There was no disagreement between reviewers with respect to quality assessment done independently. After discussion, there was also no disagreement regarding data extraction.
For the population as a whole, dexamethasone prior to extubation reduced the use of endotracheal reintubation [RR 0.18 (0.04,0.97), RD -0.086 (-0.162,-0.010)]. Since number needed to treat (NNT) is the inverse of RD, on average twelve infants would need to be treated to prevent 1 reintubation, although confidence intervals are wide [NNT 12 (6,100)]. In the subset of those thought to be at high risk of upper airway edema, a trend favouring dexamethasone did not reach statistical significance (confidence intervals surrounding RR include 1). However the RD does reach significance [RD -0.174 (-0.0344,-0.004)] and on average six high risk infants need to be treated to prevent one reintubation, with very wide confidence intervals [NNT 6 (3,250)].
The drug is not without side-effects and defining the group of infants in whom it is most likely to be effective is desirable. The trial in which infants with one risk factor for upper airway edema (multiple intubations) were excluded (Ferrara 1989) had a reintubation rate in the total study population of 0 of 59 infants. The overall treatment effect is therefore derived from the high risk population of Couser 1992 and the population including all intubated infants of Courtney 1992.
Dexamethasone was also effective in reducing the incidence of postextubation stridor [RR 0.39 (0.16,0.93), RD -0.167 (-0.307, -0.027), NNT 6 (3,37)]. This result is heavily influenced by the trial of Couser 1992. Post hoc analysis on the basis of the dose of dexamethasone administered shows that while the multiple dose strategy of Couser reduced the incidence of stridor, the single dose regime of Ferrara 1989 did not. The long term outcome of subglottic stenosis was reported only by Couser and the low incidence of this complication means caution should be exercised in interpreting the trend favouring dexamethasone.
A trend exists towards a reduction in the rate of postextubation atelectasis with dexamethasone but once again the low risk population of Ferrara 1989 recorded no cases of this adverse outcome.
Glycosuria occurred only in infants treated with dexamethasone (7 of 27) in the only trial evaluating this outcome (Couser 1992). Using the measure of RD, this finding is statistically significant [RD 0.259 (0.079,0.440)] and the number needed to treat to produce this harm is 4 (2,13). Bedside monitoring also showed a statistically significantly higher glucose level in the dexamethasone group (Courtney 1992). The clinical importance of these findings is not fully examined. In each of the two trials recording blood pressure as an outcome (Couser 1992 and Courtney 1992), the authors reported no differences between groups. No details are presented to allow meta-analysis of this outcome.
Two studies recorded pulmonary function tests before and after extubation. Couser 1992 demonstrated significant differences between dexamethasone and placebo groups with respect to total respiratory resistance, dynamic compliance and tidal volumes, all favouring dexamethasone. In the lower risk population of Courtney 1992, no such differences were found. Couser 1992 also reported pCO2 values postextubation that were statistically lower in the dexamethasone group. The absolute difference of 5 mm Hg was of limited clinical importance.
An a priori subgroup analysis was planned to examine the confounding variable, use of postextubation NCPAP. This was unable to be performed as the two trials in which this cointervention was used (Couser 1992 and Ferrara 1989) did so on an ad hoc basis and did not report outcomes related to NCPAP administration.
The outcome, requirement for endotracheal reintubation, is of modest clinical importance in the spectrum of those relating to neonatal intensive care. In addition to its implications of instability for the infant, reintubation has economic consequences, the presence of an endotracheal tube significantly increasing the cost of intensive care. Dexamethasone is a potent glucocorticoid with many effects beyond reducing airway edema. The disturbance to glucose metabolism is well demonstrated by two of the included studies. In combination, these facts suggest that it is important to define both a group at low risk of requiring reintubation, in whom dexamethasone may not be justified and a high risk group that will benefit from administration of the drug before extubation.
Description of the study populations of Courtney 1992 and Ferrara 1989 does not allow definition of their risk factors as outlined by Couser 1992. However it is likely that the latter study included a population at substantially higher risk of requiring reintubation compared with the other two. In a high risk population, the equation of treating six infants with dexamethasone in order to prevent one reintubation seems to favour treatment. The absence of infants requiring reintubation in the Ferrara 1989 trial in larger infants having had only one endotracheal tube suggests treatment is not indicated in this group. Interestingly, in a randomized trial of the routine use of dexamethasone prior to extubation in older children (Tellez 1991), no benefit could be demonstrated.
Since we regard the avoidance of intubation and other pulmonary benefits
as having greater clinical importance than glycosuria, we believe the use
of dexamethasone can be justified for infants at high risk of upper-airway
edema and obstruction after extubation.
Some uncertainty exists about the clinical importance of the benefit
attributed to dexamethasone by the limited time of followup reported by these
studies. The need for reintubation beyond 24 hours after ETT removal is not
reported. It is possible that the benefits of dexamethasone derive, at least
in part, from the improvements in pulmonary compliance demonstrated by Couser 1992, an effect that may have diminished over a longer period.
An additional limitation of this review results from the changes in neonatal intensive care since these studies were performed. The use of exogenous surfactant, increased use of antenatal steroids and the trend to extubate early to NCPAP have reduced the duration of endotracheal intubation. In addition, an increasing number of survivors at 22 to 24 weeks forms a new population in whom this treatment may have a different safety/efficacy profile.
Two dosage regimens were used by trials in this review. Meaningful comparison is only possible with respect to the outcome of stridor. An apparent advantage of the higher, multiple dose strategy requires further investigation.
Epinephrine has been used topically to reduce airway oedema and to facilitate extubation and is the subject of another systematic review (Davies,1998).
Dexamethasone reduces the need for endotracheal reintubation of neonates after a period of IPPV. In view of the lack of effect in low risk infants and the documented and potential side effects, it appears reasonable to restrict its use to infants at increased risk for airway edema and obstruction, such as those who have received repeated or prolonged intubations.
Issues of dosage and applicability to the extremely low birthweight population should be evaluated. Longer term outcomes such as chronic lung disease, duration of assisted ventilation and length of hospital stay should also be examined.
None
Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
Courtney 1992 | 1) Blinding of randomization - yes: random number table, cards drawn by pharmacy personnel 2) Blinding of intervention - yes 3) Complete followup - yes for the important outcome of endotracheal reintubation, incomplete for some secondary outcomes 4) Blinding of outcome measurement - yes | Included:
infants > 1000g, intubated for 3 to 30 days. Primary conditions comprised
RDS (38), pneumonia (8), airleak (2), asphyxia (1), meconium aspiration (1)
and persistent fetal circulation (1). Excluded: anomalies of the airway or lungs, evidence of active infection, hypertension, hyperglycemia or bleeding diathesis or had received steroids (including maternal) within 5 days of extubation. | Dexamethasone 0.5 mg/kg in 3 doses, 8 hours apart, the last dose given 1 hour before extubation or normal saline placebo. | Reintubation within 24 hours of extubation. Postextubation atelectasis within 24 hours on CXR. Pulmonary mechanics using an esophageal ballon and a heated pneumotachometer. Side effects of dexamethasone: BP and bedside glucose estimations every 4 hours, observation over 1 week for evidence of sepsis (documented on blood, urine, CSF cultures) | Larger infants - mean weight at extubation = 2.43 kg, intubated for mean of 9 to 11 days. No sample size calculations provided. 9 of 51 (18%) were excluded after randomization although data available on all for the outcome of endotracheal reintubation. | A |
Couser 1992 | 1) Blinding of randomization - Yes: method unspecified, performed by hospital pharmacy 2) Blinding of intervention - Yes 3) Complete followup - Yes 4) Blinding of outcome measurement - Yes | Included: infants at high risk for airway edema: traumatic or multiple endotracheal intubations. Excluded: congenital central airway or lung parenchymal anomalies, open use of dexamethasone for CLD, sedative or muscle relaxant medication within 12 hours of extubation. Reached predetermined minimal levels of respiratory support with no atelectasis on CXR and normal blood gases. | Dexamethasone 0.25 mg/kg per dose at 8 hourly intervals for 3 doses beginning 4 hours before extubation or normal saline placebo. | Reintubation: for severe stridor
plus bradycardia, oxygen saturation <85% in > 80% inspired oxygen and
unresponsive to racemic epinephrine or required frequent bag and mask ventilation. Stridor: defined as a high-pitched inspiratory sound associated with signs of upper airway obstruction. Atelectasis apparent on CXR performed within 24 hours postextubation. Blood gas analysis at 2 to 4 hours and 18 to 24 hours postextubation. Pulmonary mechanics using heated pneumotachometer. Side effects of dexamethasone: BP estimations every 3 to 4 hours, screening for glycosuria followed by bedside and formal blood glucose estimations as required. | High risk infants with mean weight at entry ~ 1200g. NCPAP and methylxanthine use at discretion of attending neonatologist. Sample size calculation performed on the basis of rates of extubation failure. | A |
Ferrara 1989 | 1) Blinding of randomization - yes: method unspecified 2) Blinding of intervention - yes 3) Complete followup - yes 4) Blinding of outcome measurement - yes | Included:
Single intubation, ventilated for at least 48 hours (mean = 6 days), larger
infants (mean birthweight ~2200g. Extubated from low ventilator settings
- <35% oxygen, rate < 6 breaths/minute, PEEP < 4cm. Excluded:respiratory acidosis (pH < 7.30, pCO2 > 50), > 1 intubation, already receiving steroids, congenital anomalies. | Dexamethasone as a single 0.25 mg/kg dose, 30 minutes before extubation or saline placebo. | Assessed
at 30 minutes, 6 hours and 24 hours postextubation: audible stridor, blood
gas estimations, oxygen requirements, Downes' score, the need for reintubation,
apnea and bradycardic episodes. Assessed at 4 and 24 hours: CXR for lobar atelectasis | Infants < 1500g managed with NCPAP postextubation One infant excluded after randomization because of central hypoventilation syndrome. Sample size calculations performed on the basis of rates of "respiratory complications" | A |
Courtney SE, Weber KR, Gumo SS, Spohn WA, Bender CV, Malin SW. Randomized trial of dexamethasone (D) for prevention of post extubation laryngeal edema following extubation. Pediatr Res 1989;25:34A.
* Courtney SE, Weber KR, Siervogel RM, Spohn WA, Guo S, Malin SW, Bender CV. Effects of dexamethasone on pulmonary function following extubation. J Perinatol 1992;12:246-251.
Courtney SE, Weber KR, Spohn WA, Bender CV, Malin SW, Guo S, Siervogel RM. Dexamethasone (D) for prevention of post extubation laryngeal edema following extubation. Pediatr Res 1988;23:502A.
Couser 1992 {published data only}
Couser RJ, Ferrara TB, Falde B, Johnson K, Schilling CG, Hoekstra RE. Effectiveness of dexamethasone in preventing extubation failure in preterm infants at increased risk for airway edema. J Pediatr 1992;121:591-596.
Ferrara 1989 {published data only}
Ferrara TB, Georgieff MK, Ebert J, Figher JB. Routine use of dexamethasone for prevention of postextubation respiratory distress. J Perinatol 1989;9:287-290.
* indicates the primary reference for the study
Arensman RM. Surgical management of the airway. In: Goldsmith JP, Karotkin EH, editor(s). Assisted ventilation of the neonate. Philadelphia: WB Saunders, 1988:342-356.
Carlo WA, Martin RJ, Fanaroff AA. Assisted ventilation and the complications of respiratory distress. In: Neonatal-Perinatal Medicine: Diseases of the fetus and infant. 1992:820-834.
Davies MW, Davis PG. Nebulized racemic epinephrine in infants being extubated (Cochrane Review). In: The Cochrane Library, Issue 3, 1998. Oxford: Update Software.
Durand M, Sardesi S, McEvoy C. Effects of early dexamethasone therapy on pulmonary mechanics and chronic lung disease in very low birth weight infants: A randomized, controlled trial. Pediatrics 1995;95:584-590.
Tellez DW, Galvis AG, Storgion SA, Amer HN, Hoseyni M, Deakers TM. Dexamethasone in the prevention of postextubation stridor in children. J Pediatr 1991;118:289-294.
Davis PG, Henderson-Smart DJ. Intravenous dexamethone in neonates being extubated (Cochrane Review). In: The Cochrane Library, 1999. Oxford: Update Software.
Davis PG, Henderson-Smart DJ. Intravenous dexamethasone for extubation of newborn infants (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Oxford: Update Software.
Comparison or outcome | Studies | Participants | Statistical method | Effect size |
---|---|---|---|---|
01 Dexamethasone vs Placebo (all infants) | ||||
01 Endotracheal reintubation | 2 | 160 | RR (fixed), 95% CI | 0.18 [0.04, 0.97] |
02 Postextubation stridor | 2 | 109 | RR (fixed), 95% CI | 0.39 [0.16, 0.93] |
03 Postextubation atelectasis | 1 | 110 | RR (fixed), 95% CI | 0.54 [0.19, 1.53] |
04 Glycosuria | 1 | 50 | RR (fixed), 95% CI | 12.86 [0.77, 213.62] |
05 Subglottic stenosis | 1 | 50 | RR (fixed), 95% CI | 0.28 [0.03, 2.55] |
02 Dexamethasone vs Placebo (high risk patients) | ||||
01 Endotracheal reintubation | 1 | 50 | RR (fixed), 95% CI | 0.10 [0.01, 1.68] |
02 Stridor | 1 | 50 | RR (fixed), 95% CI | 0.17 [0.04, 0.70] |
03 Subglottic stenosis | 1 | 50 | RR (fixed), 95% CI | 0.28 [0.03, 2.55] |
03 Dexamethasone vs Placebo (multiple dose) | ||||
01 Postextubation stridor | 1 | 50 | RR (fixed), 95% CI | 0.17 [0.04, 0.70] |
04 Dexamethasone vs Placebo (single dose) | ||||
01 Postextubation stridor | 1 | 59 | RR (fixed), 95% CI | 0.97 [0.27, 3.51] |
The review is published as a Cochrane review in The
Cochrane Library, Issue 1, 2005 (see http://www.thecochranelibrary.com for
information). Cochrane reviews are regularly updated as new evidence emerges
and in response to comments and criticisms, and The Cochrane Library should
be consulted for the most recent version of the Review. |