Statistics from Altmetric.com
Editor—We read with great interest the recent study by Baumer1 which compared patient triggered ventilation (PTV) with conventional intermittent mandatory ventilation (IMV) in preterm infants with respiratory distress syndrome. Baumer's trial is the largest to date.2 He concludes that there is no observed benefit to PTV and a trend towards a higher incidence of pneumothorax in infants less than 28 weeks of gestation. Methodological problems in this study, however, prompt us to question the validity of these conclusions.
While an open trial is perhaps the fastest means of achieving a large sample size, it has inherent problems of unequal experience and expertise among participating centres and unequal enrolment of patients. This study had a rather high rate of non-enrolment of eligible patients and a significant number of patients were not even offered the assigned mode of ventilation. Baumer provides little information regarding the experience of each centre with PTV and one might infer that some of the centres had little or none. Perhaps this might also be an explanation as to why more PTV patients were crossed over to IMV. The overall mortality rate in this study seems high (192 of 924 randomised patients, 20.8%), but no comment is made.
The issue of pneumothorax may have been related more to the strategy and/or triggering device used than the ventilatory mode.3Inspiratory pressures may not have been weaned fast enough during PTV. If inadequate tidal volumes are provided during PTV, the triggered breath rate may increase to compensate and hence maintain minute ventilation. This can lead to gas trapping and alveolar overdistension unless breaths are flow cycled and inspiration ends at a percentage of peak flow rather than persisting for a fixed time limit. The pressure triggering device of the SLE may be inappropriate for immature infants. In such a population, it may lead to patient fatigue and subsequent trigger failure; it performs less well than certain types of airflow trigger, as indicated by a higher degree of asynchrony.4It is of interest that the single centre study of Beresford and colleagues,5 which utilised the SLE for both PTV and conventional ventilation, but included infants of birth weights 1000–2000 g, found no difference in the pneumothorax rate. Although 386 patients were recruited into that study, only a 50% reduction in chronic lung disease could be confidently detected; no such effect was demonstrated.
Although we applaud Baumer's effort in organising a trial of this magnitude, we must caution that the conclusions should be confined to the specifics of this trial and may not be representative of PTV in this population.
Dr Baumer comments on behalf of the trial collaborators:
It was heartening to see the interest shown in the trigger ventilation trial, and I share the correspondents' caution that this trial can only show the effectiveness or otherwise of this mode of ventilation using the ventilators and techniques adopted for the trial.
The correspondents suggest that some of the participating centres were inexperienced or inexpert in the technique. As was discussed in the paper, the trial coordinator visited each participating centre and briefed staff in the technique; written protocols on both modes of ventilation were available in all centres; most participating centres already had experience of the technique (typically up to two to three years before entering the trial); centres were enrolling patients for up to four years. In addition, those few centres without prior experience were advised not to start randomising patients until they were confident with the technique. This was a pragmatic trial. If the technique of patient triggered ventilation (PTV) cannot be applied successfully even with the education and support offered in this trial, it is reasonable to conclude that it is not likely to be any more effective in widespread use.
On reviewing the dataset, 139 infants were enrolled within three months of their centre's first infant being randomised. There was no significant difference in rates of death or chronic lung disease, duration of ventilation, cerebral ultrasound abnormality, or departure from the assigned mode of ventilation between these infants and those enrolled after the first three months. There was, however, a significantly lower rate of pneumothorax in the first three months of each centre starting the trial (5.0% versus 13.0%, p < 0.02), but with no observed difference in pneumothorax rate between the two modes of ventilation. This effect does not take into account other possible confounding factors (such as centre effects). This difference in pneumothorax rate was no longer apparent if the infants enrolled within the first year at their centre were considered separately.
The finding of an apparent time effect should be interpreted with great caution. It might suggest that the education provided by the trial coordinator had a short term benefit in reducing pneumothorax rates. If so, this is an important finding that could point to the need for improved ongoing education programmes in neonatal intensive care units, and which also needs to be considered in the design of randomised controlled trials involving comparison of different techniques of ventilation.
The rate of non-enrolment (33%) is not surprising when one considers the difficulties of obtaining informed consent in such circumstances. In addition, the patient characteristics of non-enrolled patients were reported, showing similar gestation and birth weight. It is therefore reasonable to conclude that the results of the trial should be generalisable to the population served by the participating centres. The proportion of infants initially ventilated with the incorrect mode of ventilation was very small (1%). These infants were appropriately analysed on an “intention to treat” basis, and even when infants receiving only their assigned mode of ventilation were considered, the results were the same. It should be emphasised that the trial protocol allowed clinicians to change the mode of ventilation at their own discretion. This was interpreted more liberally by one or two centres. It is therefore of note that 79% of infants only received their assigned mode of ventilation throughout. Thirteen per cent of infants in the trigger arm of the Beresford study were changed to an alternative mode of ventilation, even when the study design precluded crossover of treatment strategy. As reported, the most immature infants and those subsequently dying were more likely to be switched to the other mode of ventilation. This does not suggest that centres were lacking in experience.
It is not possible to draw any inferences from the observed mortality rate. This study was conducted in the surfactant era, and the median duration of ventilation (six days) attests to the severity of the respiratory problems, including the more mature infants, despite surfactant therapy in 93% and antenatal steroid administration in 63%. The lower rate of adverse outcomes in the study of Beresford and colleagues1-1 was not surprising given the fact that infants below 1000 grams were excluded.
The hypothesis advanced by the correspondents about the consequences of too slow a rate of weaning of inspiratory pressures in PTV mode is intriguing. However, it is unsupported by any references, suggesting that it is a purely hypothetical explanation for the development of pneumothorax.
The suggestion that the pressure triggering device may be less able to achieve synchrony than the Dräger babylog 8000 was acknowledged in the discussion (reference 31). We found no significant difference in pneumothorax rates. There was also no observed benefit in the more mature infants in the study, as also reported by Beresfordet al.1-1
No modification to the trigger ventilation technique used in this study has been shown to be required based on the available research evidence from the intervening seven years. Until further evidence is published that shows a better method of PTV than the one used in this trial, it is reasonable to conclude that a similar study conducted now using the SLE 2000 would reach a similar conclusion. The limited information from this study on the Dräger babylog 8000 does not suggest that different results would have been found if the trial had used only this ventilator.
The correspondents' concerns about methodology therefore are very unlikely to account for the disappointing lack of observed benefit from PTV.
Morphine was used frequently or routinely in both the PTV trials, and theophyllines were not used except during weaning. This differs from the practice at the time of our initial report in 1993,1-2and could have had a significant impact on the success of PTV, especially in the more immature infants. Further research is also needed to improve triggering devices, to find better methods for detecting asynchrony, and to investigate the use of different approaches to PTV such as the ones suggested by the correspondents.
I would like to use this opportunity to pay tribute to the two trial coordinators (Sue Ellis and Tom Mill), to the trial statistician (David Wright), and to the data monitoring committee (David Field and Diana Elbourne), whose details were inadvertently omitted from the final paper, and without whom, together with the trial collaborators, the study would not have been possible.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.