Randomised control trial
Courtney SE, Durand DJ, Asselin JM, et al.High-frequency oscillatory ventilation versus conventional mechanical ventilation for very-low-birth-weight infants.
New Engl J Med
2002;347(9):643-52.
- Submitted by:Sachin Shah - Fellow
- Institution:Hospital for Sick Children and University of Toronto
- Date submitted:12th June 2003
1
Objectives and hypotheses
1.1
Are the objectives of the study clearly stated?
Yes. The objective was to determine whether early institution of high frequency oscillatory ventilation would improve the pulmonary outcome of very low birth weight infants without increasing the incidence of intracranial hemorrhage or periventricular leucomalacia
2
Design
2.1
Is the study design suitable for the objectives
Yes. Multicenter, non crossover, randomised controlled trial
2.2
Who / what was studied?
Infants between 601-1200 grams, appropriate for gestational age, received one dose of surfactant, requiring CV with FiO2 of at least 0.25 and mean airway pressure of 6 cms, < 4 hours of age, and expected to need mechanical ventilation for more than 24 hours.
2.3
Was this the right sample to answer the objectives?
Yes. Sample size calculations were performed.
2.4
Is the study large enough to achieve its objectives? Have sample size estimates been performed?
Sample size estimates have been performed assuming that 50% of the study population would be alive and not requring supplemental oxygen at 36 weeeks CGA and that assignment to HFOV would improve this outcome to 65%. Two tailed alpha error was 0.05 and power was 90%.
2.5
Were all subjects accounted for?
Yes. 2226 infants between 601-1200 grams were admitted to the study sites. Of these, 895 met the entry criteria; 87 were not enrolled due to non availability of equipment;158 refused consent and no attempt was made to obtain consent for 150 infants. 500 infants were enrolled; 245 in HFOV group and 254 in SIMV group. 2 were diagnosed to have congenital heart disease later and were excluded.
2.6
Were all appropriate outcomes considered?
Yes. Outcomes were CLD, death, age of first extubation, airleaks, PIE, IVH, PVL, PDA, NEC, ROP, NEc, time to reach full feeds.
2.7
Has ethical approval been obtained if appropriate?
Yes. The study was approved by the institutional review at each participating center.
2.8
Were the patients randomised between treatments?
Yes.
2.9
How was randomisation carried out?
Central randomisation at a clinical coordinating centre in California. Eligible infants were stratified according to birth weight, study site and exposure to antenatal corticosteroids.
2.10
Are the outcomes clinically relevant?
Yes. All important outcomes are considered.
3
Measurement and observation
3.1
Is it clear what was measured, how it was measured and what the outcomes were?
Yes.
Need for oxygen at 36 weeks CGA was assessed by giving a trial off oxygen for 24 hours. If at any time during the 24 hour evaluation period the infant required oxygen to maintain SaO2 > 87%, the infant was considered to be dependent on oxygen.
Successful extubation was defined as remaining extubated for at least 2 weeks
Need for oxygen at 36 weeks CGA was assessed by giving a trial off oxygen for 24 hours. If at any time during the 24 hour evaluation period the infant required oxygen to maintain SaO2 > 87%, the infant was considered to be dependent on oxygen.
Successful extubation was defined as remaining extubated for at least 2 weeks
3.2
Are the measurements valid?
Yes. All the measurements were objective.
3.3
Are the measurements reliable?
Yes. Standard objective measurement tools have been used.
3.4
Are the measurements reproducible?
Yes
3.5
Were the patients and the investigators blinded?
Investigators were not blinded.
4
Presentation of results
4.1
Are the basic data adequately described?
Yes. Clinical characteristics which are described include birth weight (different strata), gestational age, antenatal corticostroids, maternal chorioamnionitis, maternal GBS status, PIH, toxemia, race, antenatal care, use of illicit drugs, number of caesarean sections, sex, apgar scores, age and respiratory parameters at randomisation
4.2
Were groups comparable at baseline?
Yes. There was no difference in the characteristics described above.
4.3
Are the results presented clearly, objectively and in sufficient detail to enable readers to make their own judgement?
Yes. Results are presented for all the relevant outcomes. However, results are not presented as relative risk, risk difference abd confidence intervals.
4.4
Are the results internally consistent, i.e. do the numbers add up properly?
Yes.
4.5
Were side effects reported?
All important side-effects were reported such as Intraventricular hemorrhage, periventricular leucomalacia, pneumothorax, sepsis, patent ductus arteriosus, necrotising enterocolitis, intestinal perforation, retinopathy of prematurity and hearing loss.
5
Analysis
5.1
Are the data suitable for analysis?
Yes
5.2
Are the methods appropriate to the data?
Yes. Categorical outcomes were compared using Fisher's exact test. Normally distributed continuous outcomes were compared with the use of unpaired student's t test, and non parametric continuous outcomes with the use of Mann Whitney U test.
5.3
Are any statistics correctly performed and interpreted?
Yes. However, treatment estimates such as relative risk, risk difference and confidence intervals are not reported.
6
Discussion
6.1
Are the results discussed in relation to existing knowledge on the subject and study objectives?
Yes
6.2
Is the discussion biased?
Yes. Some of the discussion addresses issues not presented in the results section (e.g NNT). Also, the findings of this study have not been compared with the more recent HFOV trials.
7
Interpretation
7.1
Are the authors' conclusions justified by the data?
The authors conclude that there was a small but significant benefit of HFOV in terms of pulmonary outcome for VLBW infants without an increase in other complications of premature birth.
These conclusions are not justified by the data because the difference in CLD or mortality was of borderline statistical significance.
These conclusions are not justified by the data because the difference in CLD or mortality was of borderline statistical significance.
7.2
What level of evidence has this paper presented? (using CEBM levels)
Level 1b
7.3
Does this paper help me answer my problem?
Yes. Also, the paper does reassure that use of HFOV using high volume strategy is not associated with increased risk of IVH or PVL. However, long term neurovelopmental follow up is also needed.
8
Implementation
8.1
Can any necessary change be implemented in practice?
N/A. This will depend on the existing practice.
8.2
What aids to implementation exist?
No increase in complications of prematurity with the use of HFOV with a trend towards decreased mortality and CLD.
8.3
What barriers to implementation exist?
The difference in the incidence of CLD or mortality is of borderline statistical significance. The clinical significance remains uncertain.
Protocol based approach to HFOV.
Conservative extubation criteria for CV (FiO2 < or = 0.25 and MAP < or = 5).
Long term neurodevelopmental outcome not reported.
Protocol based approach to HFOV.
Conservative extubation criteria for CV (FiO2 < or = 0.25 and MAP < or = 5).
Long term neurodevelopmental outcome not reported.