Article Text

Download PDFPDF

Parental consent and neonatal delivery room trials: walking an ethical tightrope
  1. Louise S Owen1,2,3,
  2. Peter G Davis1,2,3
  1. 1 Dept of Obstetrics and Gynaecology, University of Melbourne, Parkville, Victoria, Australia
  2. 2 Clinical Sciences, Murdoch Children's Research Institute, Melbourne, VIC, Australia
  3. 3 Newborn Research, The Royal Women's Hospital, Melbourne, Victoria, Australia
  1. Correspondence to Associate Professor Louise S Owen, Newborn Research, The Royal Women's Hospital, Parkville, VIC 3052, Australia; louise.owen{at}

Statistics from

Vohra et al present the results of a secondary analysis of the Vermont Oxford Network Heat-Loss Prevention (HeLP) Trial. The HeLP trial examined use of polyethylene wrap in the delivery room (DR), to reduce mortality in extremely preterm infants.1 More than 800 infants from 38 centres across the USA and Canada were enrolled. The study stopped at 50% recruitment, on grounds of futility. All centres used prospective antenatal consent, four sites also used a consent waiver. Seventeen per cent of recruits (n=133) were enrolled via waived consent. Vohra’s analysis evaluates whether the consent mode used was associated with differences in recruits.

Neonatal resuscitation research unfortunately lags behind other areas, largely due to the complexities of conducting DR studies meaning that many DR interventions have never been rigorously tested. The Declaration of Helsinki guides all research consent; but gaining ethically appropriate antenatal parental consent for DR research is challenging. Antenatal consent is not always possible, and researchers must avoid duress resulting from time pressure or parental distress.

Vohra rightly notes that restricting recruits for DR studies to only those in whom antenatal consent is possible, results in systematic exclusion of an important subset of the population, an effect highlighted by a secondary analysis of the SUPPORT Study (Surfactant Positive Airway Pressure and Pulse Oximetry Randomized Trial).2 SUPPORT required antenatal consent for DR interventions, and recruited infants with fewer risk factors than those eligible but not recruited. Enrolled infants were born in better condition and had better outcomes, with the differences attributable to differences in their birth characteristics.2 Dependence on antenatal consent resulted in biased enrolment; enrolled infants came from a lower-risk subset of the eligible population who were more likely to do well. Infants at highest risk of adverse outcome were excluded. Incorporation of consent processes that don’t require antenatal consent could avoid selective sampling, reduce bias and increase external validity.

The HeLP Trial permitted both waived consent and what was termed ‘delayed consent’. Any form of waiver, or after-the-fact consent is more accurately termed ‘research without prior consent’ (RWPC). Experts advocate for this terminology,3 as strictly, consent cannot be given for something that has already occurred. RWPC potentially balances the need for unbiased research, with respect for autonomy and protection of patients, especially important in the vulnerable neonatal population. High-quality research is necessary to evaluate new therapies and standard but untested treatments. When antenatal consent is appropriate it should be used, however a parent’s capacity to give ethically valid consent may be impaired in the immediate perinatal period. Parents frequently struggle to assimilate standard consent processes, and the anxiety around preterm birth and a potentially sick child further erodes their capacity.3 The UK’s Royal College of Obstetricians and Gynaecologists guideline for obtaining consent for perinatal research when consent is time critical, advocates for prospective ‘brief oral consent’ followed later by written consent.4 Whilst this provides an alternative pathway, others have warned that parental recall of such assent may be very limited.5 6

RWPC has been used in more than a dozen neonatal trials since 2000, but its implementation remains controversial. Public perceptions that clinical trials constitute experiments persist, and research ethics committees (RECs) and many clinicians are more familiar with using unproven interventions, than enrolling patients into clinical trials.3 Legislation regulating RWPC varies; Vohra details the US and Canadian regulations; others have compared these with Australian and European regulations.7 Most stipulate that the research is of ‘minimal risk’, although how risk is measured is subject to debate. When two accepted clinical practices are compared the assumption is that the research risk is minimal. However, if the outcome sought is a difference in mortality, and one intervention ultimately results in higher mortality than the other, is the risk still minimal?

All study sites in HeLP could apply for RWPC, and four gained permission. Whether other sites sought approval and were denied is unclear. The HeLP Trial described RWPC options as either ‘consent waiver’ or ‘delayed consent’. The latter required parental permission to use collected data. Vohra doesn’t specify whether the four RECs sanctioning RWPC endorsed full waiver, or stipulated ‘delayed consent’. Of 137 infants recruited using RWPC, 133 had data included, and four families ‘withdrew’. Although, this proportion is small, managing data from randomised infants when consent is declined (or not sought), is an important aspect of RWPC. Of the 133 included infants, 30 died before 36 weeks corrected age. It is not clear what process was required when an infant died before consent was sought, or whether this occurred. However, if data from those who die without consent being sought are disregarded, systematic bias is introduced, and the true effect of the intervention may be missed. Studies addressing this issue recommend including these data, but regulations are inconsistent. When designing trials using RWPC, consideration should be given to seeking RECs’ approval to use anonymised data from deceased patients if no consent from bereaved parents is obtained.3

Vohra’s analysis highlights differences in the mothers and infants enrolled via the two consent pathways. A higher proportion of mothers recruited antenatally received antenatal steroids, and their infants had higher Apgar scores. No differences were seen in neonatal survival or important morbidities. Similarly in 2018, Songstad et al published a secondary analysis of the High Flow Nasal Cannulae as Primary Support in the Treatment of Early Respiratory Distress Trial that compared initial non-invasive respiratory support modes in preterm infants in eight centres across Australia and Norway, mostly using prospective consent.8 RWPC was included at one Australian site, commencing partway through recruitment. Songstad compared recruitment at that site whilst RWPC was available, with earlier recruitment when it was not. He reported that when RWPC was available, recruitment of eligible infants was higher (82% vs 50%), and included mothers received less antenatal steroids and more antibiotics. No important difference in primary outcome was seen.

Vohra’s and Songstad’s data support the hypothesis that consent mode systematically alters the enrolled population. Although neither study demonstrated differences in trial outcome by consent mode, this potential exists. Vohra rightly notes that if the requirement for antenatal consent selects a sample more likely to do well, then outcomes obtained from that study are only applicable to the subpopulation enrolled. Whereas, if recruitment comes from the entire eligible population the true effect is more likely to be found, and the results applicable to the whole population.

Different countries, and RECs within countries, interpret and apply guidelines for alternative consent modes differently;9 such disparity risks multisite trials recruiting from different subpopulations, leading to less uniform, less valid study outcomes.7 This concern has motivated regulatory bodies to advocate for single-site ethical approval for international multisite studies. den Boer et al describe a framework for this, whilst acknowledging that achieving agreement will be challenging.7

Vohra suggests that RWPC ‘shares the burden’ of consent with the hospital and the REC, allowing parental involvement ‘at a more appropriate time’. Vohra argues that researchers should actively pursue RWPC as it is ethically acceptable in the perinatal period. Others might argue that RWPC erodes parental trust and that some shared decision-making is feasible, and desirable, in even the most complex situations. Several qualitative studies have explored parental opinions of RWPC in paediatric emergencies, and found it broadly acceptable. Limited data exist for newborns: Rich et al reported 69% of recruited families recalled a positive experience,6 and more recently Sloss et al reported that 89% of parents of newborns with recent experience of RWPC found the process acceptable.10

However the controversies are viewed, RWPC facilitates recruitment of a broader, higher-risk population, the impact of which is still to be determined. Neonatal DR studies using alternative consent modes should report the effects of consent mode on trial outcomes.

Parents and infants deserve fair access to neonatal research and to optimal neonatal care; parents should be involved in research decisions but also protected from onerous decision-making during periods of high stress. Community consultation, already mandated in some countries,7 and specifically including parents of preterm infants should commence during study design, informing decisions regarding acceptable consent processes.

As researchers, we strive to run efficient, robust DR studies balancing the ethical imperatives to provide high-quality, evidence-based care whilst respecting the child’s and family’s autonomy. Guidance and support from regulatory and funding bodies, parent and community groups, and RECs are required to achieve this aim.

Ethics statements

Patient consent for publication



  • Correction notice This article has been corrected since it first published. The provenance and peer review statement has been included.

  • Contributors LSO conceived, wrote and edited the first and final drafts of this manuscript. PGD contributed to and edited the manuscript.

  • Funding This study was funded by the National Health and Medical Research Council, grant number: (105911,1113902,1159444).

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

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.

Linked Articles