Statistics from Altmetric.com
Editor—The international randomised controlled trial by Dr Baumer1 concluded that there was no benefit of patient triggered ventilation (PTV), but an added risk of increased pneumothorax in those less than 28 weeks gestation. In the same issue, Beresford et al 2 concluded in a similar trial (with slightly more mature newborns 29 weeksv 27), that PTV was feasible with no significant differences noted in medium term outcomes.
Chronic lung disease (CLD) is multifactorial in origin and in Baumer's trial significant factors, like ventilation pressures, intrauterine growth retardation, use of postnatal steroids, and nasal continuous positive airway pressure (nasal CPAP) use, have been not compared between the groups.
Intrauterine growth retardation has been shown to be an important risk factor for CLD.3 In a trial where primary outcome is CLD, omission of the data regarding use of postnatal steroid use is quite surprising. Meta-analysis of the use of dexamethasone in very low birthweight infants has shown marked reduction in CLD at 36 weeks post conceptional age.4 The use of nasal CPAP that has been shown to influence the incidence of CLD,5 is not mentioned in their trial. Beresford et al did not use nasal CPAP in any of their infants, but used synchronised intermittent mandatory ventilation (SIMV) in the weaning phase.
To the clinician, who has to make the choice of using these commonly available tools to reduce death and morbidity in this vulnerable group of infants, is the choice clear or more unclear? I leave the readers to decide.
Dr Baumer responds on behalf of the trigger trial collaborators:
Visveshwara emphasises that the results of the trigger trial should not be interpreted as demonstrating lack of benefit for patient triggered ventilation using other sensors or ventilators. I would concur with this statement, which was emphasised in the paper.
However, Visveshwara should not be surprised to find different outcome rates in the patients whose results he presents, as they are a different group of infants from those reported in either study. The evidence for benefit from the impedance technique is unconvincing, based as it is on a controlled study of 40 infants and a further 110 uncontrolled cases from one centre. A multicentre randomised controlled trial of sufficient power is needed to demonstrate benefit from the impedance and patient terminated ventilation techniques he describes. To date, such a study has not been performed.
Yadav suggests that important risk factors have not been compared in the study. He describes two different types of risk factor, namely inherent factors in the infant (intrauterine growth retardation) and treatments applied to the infants (ventilation pressures, use of postnatal steroids and of continuous positive airway pressure (CPAP)).
In a large randomised controlled study, individual patients will have varying degrees of risk for the outcomes being measured. The purpose of the study design is to allocate patients in such a way that the overall risk for each arm is the same. The larger the study, the less likely that there will be an unequal balance of risk, assuming that the randomisation process is performed correctly. We reported very similar birthweights and gestations in the two groups. The proportion of growth retarded infants was therefore allocated equally, and will not have biased the results.
A study comparing two modes of ventilation cannot be conducted with the attendant clinicians blind to treatment allocation. The study protocol required all other treatments to be applied equally to infants in both arms of the study. There were written treatment protocols for each mode of ventilation. However, it is still possible that other treatments could have been applied unequally, with the possibility of bias resulting.
Interpretation of ventilator pressures is difficult, as in the trigger ventilation technique weaning was undertaken at lower peak inspiratory pressures. In Plymouth, trigger ventilated infants entered in the trial had slightly lower peak inspiratory pressures in the first 72 hours, consistent with the different weaning policy. However, as the duration of ventilation did not differ between the groups, it is reasonable to conclude that there was no systematic bias in the application of ventilation.
Information on the postnatal use of steroids was collected in the trial. There was no difference in the proportion of infants receiving postnatal steroids (25.5% v 26.0%), nor in the postnatal age at which they were first administered (median 15v 17 days). There is therefore no evidence of bias resulting from their use.
The use of CPAP for weaning from ventilation has not been demonstrated to reduce chronic lung disease in randomised controlled trials. The paper Yadav cites discusses the possible benefits of a policy of early use of nasal CPAP. This is not relevant to the trigger trial, as infants recruited were by definition already being ventilated. I would also like to qualify Yadav's statement that we found an increased risk of pneumothorax in infants less than 28 weeks gestation. The difference was not statistically significant, suggesting that the observed difference may have occurred as a result of chance.
The trial therefore shows no evidence of bias, and the finding that patient triggered ventilation has no additional benefit over intermittent mandatory ventilation using the ventilators and techniques studied remains valid. The trial cannot assist clinicians in their choice of other modalities of support such as early use of CPAP or postnatal steroid use, and is applicable both to growth retarded as well as appropriately grown preterm infants.
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.