Article Text
PDF
Abstract
Background Neonatal research, particularly neonatal emergency research is a challenging area, notably in relation to obtaining valid prospective informed consent. The aim of this study is to determine parental perceptions of the consent process involved in performing research in newborn care, to explore methods used to obtain consent and their acceptability to parents.
Methods A parental questionnaire was developed that examined attitudes towards research and hypothetical research studies, in which the acceptability of various methods of consent was examined (informed, waived, deferred). These research scenarios were of varying time sensitivity and perceived risk level. The study setting was an Irish maternity hospital.
Results There were 600 responses to the questionnaire. In 93% of cases, parents felt that their involvement in the consent process was essential. In emergency situations, 52% felt full prospective informed consent was necessary; however, almost 28% of parents would feel pressure to consent. Most (75%) parents would prefer to be approached to discuss neonatal research studies antenatally, irrespective of study type and 40% of parents felt that neonates involved in research studies received overall better care. Acceptability of deferred consent was greater than waived, and was highest for the more emergency-based scenarios presented.
Discussion Parents feel that they should play a central role in research involving their children. There were differences in the acceptability of various consent methods with strongest agreement for informed consent and lowest agreement for waived consent. Parents were more willing to accede to deferred consent in the cardiopulmonary resuscitation scenario study. These findings provide useful insights to consent strategies in future newborn research studies.
- neonatology
- health services research
- ethics
Statistics from Altmetric.com
What is already known on this topic?
Valid prospective informed consent is a foundation on which ethically robust research is conducted; however, there may be challenges in obtaining such consent in neonatal research, particularly pertaining to emergency research.
Existing evidence highlights concerns that current methods of consent are unsatisfactory for parents and raises questions about the validity of informed consent in such research studies.
What this study adds?
Parents feel that they should play a central role in research involving neonates, and the majority would like to be approached to discuss research studies in the antenatal period.
When comparing informed, deferred and waived consent level of acceptability for parents was highest for informed consent, followed by deferred with the lowest agreement for waived consent across all study types.
Background
Neonatal research, particularly neonatal emergency research is challenging, most notably in relation to obtaining valid prospective informed consent. There is widely held agreement that parents have the opportunity to consent or decline to participate in research studies via an informed consent process. However, there is evidence that the existing consent procedures are unsatisfactory.1 This is particularly true when applied to time critical delivery room and resuscitation research in which there may be limited time to obtain prospective and truly informed consent. Parents subsequently questioned about enrolment at the time of delivery raised concerns about timing of approach, the amount of information they received and the lack of time to consider it sufficiently.2 Mothers approached for consent in the immediate postpartum period expressed concern that their physical, mental and emotional state may have precluded them fully understanding the information being provided.3 These findings question whether consent given under such circumstances can be considered valid. A key element of prospective informed consent includes provision of information in a timely fashion, allowing parents time to process and understand this information prior to making an informed decision, which may prove challenging in neonatal research.
Previous studies have argued that full informed consent is impossible in emergency situations as an adequate explanation of the research being undertaken is difficult and voluntariness is compromised due to parental anxiety.1 To address such concerns, a number of modifications to the consent process have been suggested.4 Alternative proposed consent processes include antenatal consent, adopting a deferred process, a waived consent process and in some cases using a continuous consent process, all of which have their own potential limitations.1 4 5
Deferred consent refers to the process of conducting research without prior prospective informed consent but later approaching parents for consent for a child to remain in the study or for the information collected as part of the study to be used. While it raises ethical questions about respect for patient autonomy, deferred consent has the benefit of facilitating research in situations where prospective informed consent would not be possible.6 The Declaration of Helsinki makes provisions for research involving patients from whom it is not possible to obtain informed consent in advance.7 8 Furthermore, European Union legislation has evolved to account for this challenge. Under Article 35 of EU Regulation Number 536 Clinical trials on medicinal products for human use, 2014, deferred consent may be acceptable under certain circumstances. The requirements are that (i) research decisions cannot be reasonably delayed until a parent is present to give prospective informed consent, (ii) the trial is expected to provide a direct clinically relevant benefit for the patient or the diagnosis of the patient’s condition, (iii) the parents be approached for consent as soon as practical after study involvement and (iv) deferred consent has had prior approval as a consent procedure for that study by an ethics committee.9 This approach has been evaluated in peri-partum research in the CONsent methods in childreN’s emergEncy medicine and urgent Care Trials (CONNECT) study which found that parents and practitioners, while initially expressing concerns about deferred consent in the CATheter infections in CHildren trial (CATCH) trial, later described support for deferred consent in emergency trials to facilitate progress in this area, if the intervention was already in use in routine clinical practice.10 A survey of parents whose babies had participated in a randomised controlled trial evaluating two different methods of placental transfusion to premature infants which involved a deferred consent process reported an overall positive (69%) response to their baby’s involvement.11 ‘Waiver of consent’ describes a different process, whereby informed consent is not a requirement of the study. This again presents concerns for patient autonomy and ethically conducted research. Neonatal trials where low-risk interventions or standard of care are being evaluated have employed a waiver of consent; however, parental understanding and acceptance of this has not been evaluated in depth.11
Another example is the suggestion that in the case of emergency neonatal trials antenatal information should be given to all parents and consent obtained prior to delivery so that they may ‘opt out’ at this time or continue to have their babies enrolled in the case of an emergency.4 In qualitative retrospective interviews, parents report that they would have liked information about the trial earlier in their pregnancy. However, limitations to this approach included discomfort in consenting to something that had not yet happened and an inability to make a proper decision until the situation applied directly.2 However, a Danish study found that parents were comfortable or very comfortable with giving consent at the time of resuscitation if prior information about the study had been given in the antenatal period. With regard to deferred or waived consent, the majority of parents in this series were uncomfortable with this option. A smaller number would, hypothetically, be comfortable with waived or deferred consent under the assumption that physicians would act in the best interest of their child.5 This approach has potential disadvantages however. The results of the SUPPORT trial demonstrate that those who were approached for antenatal consent tended to have more favourable sociodemographic variables and were more likely to have received antenatal steroids when compared with those not approached for consent.12 This raises concerns about the external validity and generalisability of the research. A further concern is that providing information about ongoing research studies in the antenatal period may be unnecessarily burdensome to parents or increase parental anxiety.13
The aim of this study is to determine parental perceptions of the consent process involved in neonatal research. We wished to explore methods used to obtain consent for neonatal research and their acceptability to parents. In better understanding parental views on this subject, we hope to inform practice, the design of further research studies and provide additional information to stakeholders in this area.
Methods
A parental questionnaire was designed to elicit parental views on research involving neonates and in particular on issues involving consent procedures (see online supplementary material 1). The questionnaire included 32 items, which examined general attitudes towards research involving neonates, the importance of consent in neonatal research and also provided parents with four hypothetical research studies to assess acceptability of different methods of consent. These research scenarios were of varying time sensitivity and perceived risk level. Parents were asked to rate the acceptability of participating in the study and various methods of consent provided (informed, waived, deferred). Input into the design of the hypothetical research studies was sought from two neonatologists, an obstetrician and a medical ethicist. A summary of the research studies is included in box 1. The final section included demographic information. The questionnaire used structured questions with yes/no and Likert scale answers to facilitate quantitative analysis. The questionnaire was piloted on parents in the neonatal unit and the original questionnaire was subsequently approved and used.
Research studies presented to parents in questionnaire sections 2–5
Study 1: which is the best way to feed newborns born prematurely?
Newborn babies who are born prematurely are fed by putting small amounts of milk into their tummies by a tube. We normally start with small amounts of milk and build up over the days after birth. We do not know if it is better to increase the milk intake over 5 or 7 days. In this study, babies will be randomised to one of the following groups.
Reach full feeds by 6 days old.
Reach full feeds by 8 days old.
Study 2: what is the best way to give newborns extra blood from the placenta just after the baby is born?
Following delivery, a baby can be given a little extra blood from the placenta by either of these two ways:
Allowing the baby to remain attached to the umbilical cord for an extra minute after birth.
Squeezing the blood in the umbilical cord into the baby after birth.
Both methods work and both are performed routinely around the world. However, it is not clear which is the best for babies. To find out which method is best, newborns would have to be randomised before they are born to decide which babies get which method of giving extra blood from the placenta.
Study 3: resuscitation research
A small number of babies require extra resuscitation after birth. This might mean providing cardiac massage to the baby. There are two ways of giving cardiac massage:
The current way involves giving three chest compressions to one breath in a synchronised manner.
The other way is to give continuous compressions with prolonged breaths provided.
No studies have been done to see which method of cardiac massage works best for newborns. This study would randomise newborns that need resuscitation to one or the other method. Since we cannot predict which babies might need cardiac massage, it is difficult to ask parents for consent for this possibility in advance, as it is an absolute emergency.
Study 4: which is the best method to give a medicine into the lungs of babies with breathing difficulties after birth?
Babies who are born prematurely often have difficulty breathing due in part to reduced levels of a substance called surfactant in their lungs. To help treat this, doctors place a breathing tube into the lungs and surfactant is given directly into the lungs along this tube. A new method has been developed that allows the medicine to be given without having to put the tube into the baby’s lungs. This is done by vaporising the medicine into an aerosol form using a nebuliser and letting the baby inhale the medicine. This study would randomise preterm babies with breathing difficulties into two groups:
Babies would receive the nebulised surfactant medicine.
Babies would receive a nebulised placebo (nebuliser without surfactant).
If the babies in either group had breathing difficulties, they would all receive the surfactant medicine via the breathing tube.
Supplemental material
This study was carried out in Cork University Maternity Hospital, a 126-bed maternity hospital with approximately 8000 births and 90 000 outpatient attendances per year. Study participants were invited to complete the questionnaire in five different settings: antenatal outpatient appointments, antenatal inpatient wards, postnatal inpatient wards and neonatal outpatient clinic appointments. Non-probability convenience sampling was used. Parents were recruited from two antenatal outpatient clinics per week and one neonatal outpatient clinic per week over the study period. Parents and prospective parents were approached by study investigators in the outpatient clinic waiting rooms and invited to participate. Questionnaires were distributed to parents who were willing to participate and later returned to a secure box in the outpatient department. Antenatal and postnatal inpatients were invited to participate by hospital staff on admission to the ward and completed questionnaires returned to a secure box on the ward. The time period of the study was 6 months from January to June 2017. A returned completed questionnaire was considered implied consent to participate.
Only anonymised data were collected, inputted into Microsoft Excel and analysed using SPSS.
Continuous data were described using mean and SD and categorical data using number (percentage). For Likert-scaled questions, number (percentage), mean (SD) and median (IQR) were reported. Responses to Likert-scaled questions for the hypothetical research studies were compared using the Friedman test. If statistically significant differences were found between the four studies, pairwise comparisons were performed using Dunn-Bonferroni tests. For binary response questions, responses were compared using the Cochran Q test. If statistically significant differences were found between the four studies, pairwise comparisons were performed using McNemar’s test, with Bonferroni correction. The McNemar-Bowker test was used to compare responses with nominal scaled questions.
A parent’s overall likelihood to consent to a study was measured by averaging their responses to the questions on1) participation in the study, 2) deferred consent to the study and 3)waived consent to the study. The possible range of scores was 1–5 with a higher score reflecting a higher likelihood to consent to the study. The reliability of this scale was assessed using Cronbach’s alpha. Repeated measures analysis of variance (ANOVA) was used to compare likelihood with consent across the hypothetical studies. If a statistically significant difference was found between the four studies, pairwise comparisons were performed using paired t-tests, with Bonferroni correction. All statistical analyses were performed using IBM SPSS Statistics V.24. All tests were two-sided and a p value <0.05 was considered to be statistically significant.
Results
Demographic information
In total, 673 questionnaires were distributed and there were 600 questionnaires returned completed in full by parents or prospective parents and included in this study. Of the total respondents, 72% were mothers. The parent’s ages ranged from 16 to 50 years with a mean (SD) of 32.4 (5.3) years. The majority of respondents were married (64%) and had completed third-level education (75%). Most parents completed the questionnaire in antenatal clinic (69%) or antenatal ward (17%) and the mean gestation at the time of filling the questionnaire was 28.1 weeks (range 8–41 weeks, SD 9.2, n=429). Demographic details are outlined in table 1.
Parent information, n=600
Attitudes towards consent in neonatal research
Of the 600 parents who completed the questionnaire, 112 (19%) had previously been approached to participate in a research study. Of those, 101 (90%) consented to participate in the research study on that occasion. Parental attitudes towards the consent process are summarised in table 2. Of the total respondents, 555 (93%) felt that they should be asked for consent in research involving their babies. When asked specifically about informed consent for research in emergency situations, 311 (52%) felt that full informed consent was necessary while 132 (22%) were unsure.
Parental opinion on research involving newborns, n=600
Further questions examined whether parents would feel pressured if they were approached for consent for their child to participate in a research study. In a non-emergency situation, 446 (74%) said they would not feel pressured, 95 (16%) were unsure. In emergency situations, 319 (53%) would not feel pressured and 115 (19%) were unsure. Responses were significantly different depending on whether it was an emergency or non-emergency situation (p<0.001 from McNemar-Bowker test) with parents more likely to feel pressurised to give consent in an emergency situation.
When questioned as to when they would prefer to be approached to consent for research studies involving neonates, 450 (75%) would prefer to be approached before birth and 89 (15%) would prefer to be approached after immediate birth. Of the 600 parents questioned, 40% (n=238) felt neonates involved in research received better overall care, 2% (n=9) felt care would be worse and 59% (n=353) were unsure.
Parental attitudes towards hypothetical research studies
Agreement to participate in hypothetical study
The responses to the statement "I would agree to my baby participating in this study" are summarised in table 3. Agreement was strongest for study 3 (resuscitation) and lowest for study 4 (new drug study). A statistically significant difference was identified (p<0.001 from Friedman test). Post hoc tests revealed that the differences were between study 3 (resuscitation) and each of the other studies (study 3 vs study 1) (feeding trial) (adjusted p=0.001); study 2 (cord clamping) (adjusted p=0.012); study 4 (new drug study) (adjusted p<0.001).
Parental opinion on informed, waived and deferred consent n=600
Timing of informed consent
For all four studies, the majority of parents would prefer to be approached to discuss participating in the study before their baby was born. Preference to be approached prior to the birth of the baby was highest for study 2 (80%) and lowest for study 3 (69%). There was a statistically significant difference between the four studies (p<0.001 from Cochran’s Q test). Post hoc tests revealed that the differences were between study 2 (cord clamping) and each of the other studies (study 2 vs study 1) (adjusted p=0.014); study 3 (adjusted p<0.001); study 4 (adjusted p=0.048) (table 4).
Preferred timing to be approached to discuss participating in the study, n=600
Parental attitudes towards deferred consent in neonatal research
Responses to the statement “In circumstances where there was not enough time to get consent, I would be happy for my baby to be enrolled in the trial and approached for consent after my baby was enrolled" are summarised in table 3. Agreement to deferred consent was strongest for study 3 and weakest for study 1. A statistically significant difference was found between the studies (p<0.001 from Friedman test). Post hoc tests revealed that the differences were between study 3 and each of the other studies (study 3 vs study 1) (adjusted p<0.001); study 2 (adjusted p<0.001); study 4 (adjusted p<0.001), and between study 1 and study 4 (adjusted p=0.001).
Parental attitude towards waived consent in neonatal research
Responses to the statement "I would be happy for my baby to be a participant in this study without it ever being discussed with me" are summarised in table 3. Agreement to waived consent was strongest for study 3 and weakest for study 1. A statistically significant difference was found between the studies (p<0.001 from Friedman test). Post hoc tests revealed that there were statistically significant differences between all four studies (adjusted p<0.001 for all except study 3 vs study 4 (adjusted p=0.001).
Overall parental likelihood to consent to research study
The scales were found to have moderate-to-high reliability. Cronbach’s alpha was 0.64 for study 1 (moderate reliability), 0.71 for study 2 (high reliability), 0.83 for study 3 (high reliability) and 0.82 for study 4 (high reliability). Agreement to participate in the study was strongest for study 3 (mean (SD): 3.19 (1.01)), followed by study 4 (mean (SD): 2.89 (0.97)) and study 2 (mean (SD): 2.81 (0.88)) and weakest for study 1 (mean (SD): 2.60 (0.76)). A statistically significant difference was found between the studies (p<0.001 from repeated measures ANOVA). Post hoc tests revealed that there were statistically significant differences between all studies (adjusted p<0.001 for all) except for study 2 and study 4 (adjusted p=0.254).
Discussion
Informed consent forms the foundation on which all ethically conducted research is based; however, this is a challenging area when neonates are involved. The results of this study demonstrate that parents feel that they should play a central role in research involving neonates, and the majority would like to be approached in the antenatal period. In general, parental attitudes were positive and a general belief that neonates involved in clinical research receive the same or better care. In the emergency setting however parents reported higher rates of ‘pressure’ and uncertainty surrounding the informed consent process.
We found that most parents would prefer to be approached to discuss research prior to delivery regardless of the research study. This was particularly true for the cord milking/delayed cord clamping study. Previous studies have reported that parents may experience anxiety enrolling in research studies when birth is imminent citing inadequate information provision, insufficient time to consider decision sufficiently and a feeling that their decision was rushed. These findings were replicated in a study specifically addressing consent to delayed cord clamping as was reported in our parent group.2 Parental ability to understand and retain information pertaining to a research study is undoubtedly undermined in the immediate perinatal period.14 15 Parents questioned about their consent to participate in research in labour describe dissatisfaction with the information provided, insufficient time to consider the information and some not recalling the consent process at all.16 These issues are of obvious concern where full informed consent is concerned raising questions about the validity of such consent. Parental preference towards antenatal consent as described in this study reflects these concerns.
Providing information about ongoing research to parents during the antenatal period could help to alleviate anxiety and improve informed consent practices; however, this must be balanced with the risk of causing undue concern among expectant parents.4 While we did not specifically address this process, there may be a role for provision of written information on various studies, displayed on hospital websites and displayed on information screens when people are attending outpatients. An additional method is a continuous informed consent framework in which initial informed consent is the first step of an ongoing discussion and information sharing with parents throughout recruitment and during the study. This approach has resulted in a positive parent experience in other studies.17 Another study addressing consent to NICU research found that for 38% of parents recruitment increased stress levels in an already stressful situation.18 Parents report altruistic motivations for consenting to neonatal research studies including making a contribution to research and wanting to help future families.2 19 Perceived benefits for their baby were also a motivating factor and that was also highlighted by parental responses in this and other studies.20 Parents wish to be centrally involved in research involving their children. Moving beyond parental involvement on an individual level in the consent process towards an institutional level with parental representation on institutional review boards and liaising with parents in research protocol design allows parental opinion to be considered from study inception.
Overall parents were most likely to consent to study 3 (‘resuscitation study’) and least likely to consent to study 1 (‘feeding study’). We hypothesise that parents may be more likely to consent to higher risk research where infants are more acutely unwell but where the potential benefits of improving treatment are much greater. In study 1, the perceived risks and benefits of participation may be of less tangible importance to parents resulting in lower rates of parental consent. The concept and language in study 4, the new surfactant delivery study, were perhaps more difficult to understand and unfamiliar to parents resulting in less parental likelihood to participate.
When comparing informed, deferred and waived consent, the strongest agreement was for informed consent, followed by deferred with the lowest agreement for waived consent in all four studies. There was however a difference in acceptability of various consent methods to parents based on the underlying study and its perceived risk. Deferred and waived consent were less acceptable in perceived non-urgent situations such as the feeding study. Parents were more willing to accede to deferred or waived consent in the time-sensitive resuscitation study. In the resuscitation research scenario, however, the questionnaire provided a rationale for the timing of consent. Further justification for timing of consent in this study may have influenced parental responses and the acceptability of alternative consent methods in this study.
Parents in this study demonstrate understanding of the challenges surrounding informed consent in emergency research. Deferred consent was preferable to waived in all situations emphasising once again parental desire to be a part of the research process and their understanding of the issues. Improving the management of newborns requiring resuscitation is an important area in which ethically conducted research is a challenge. Further work into the feasibility and acceptability of novel methods of consent in time-sensitive studies among parents, patients, ethicists and clinicians is mandatory to facilitate improved clinical outcomes for these patients.
This study has several potential limitations. Given that it is a questionnaire-based study using convenience sampling, there is the potential for response bias among those choosing to participate. In our study, 75.3% of participants had or were completing a third-level qualification. This is higher than the 56.2% of those aged 15–39 years who have such a qualification as reported by the Central Statistics Office, Ireland in 2016 suggesting self-selection bias. Mothers were over-represented among the respondents due to the nature of healthcare attendance among expectant parents with mother’s presence mandatory in the outpatient and antenatal ward setting. This pattern of gender distribution has also been seen in other studies addressing consent in neonatal research.17 21 This is a potential source of bias in this and other studies and future studies in this area should consider this in study design, perhaps using quota sampling to ensure a more equally distributed study population. Furthermore, 86% of the responses were from the antenatal outpatient or ward setting. A more even distribution of antenatal and postnatal parents would have been preferable as parental experience throughout delivery and the postnatal period may influence decision-making. Furthermore, the questionnaire did not address consent in particular circumstances surrounding delivery, for example, parental attitudes towards consent in the setting of pre-term labour. This is an important consideration and one that would benefit from further investigation in future studies. The design of hypothetical situations and the limited information provided for each study could potentially have limited full parental understanding and therefore accuracy of response. The questionnaire was designed to be as brief as possible to reduce question fatigue and further potential bias; however, this possibility cannot be excluded. The synchronous presentation of four research studies for consideration may have resulted in different parental opinion than if the studies were presented in isolation. The hypothetical nature of this study is also a potential limitation as parents may have differing attitudes if approached to participate in these studies in reality. Despite these limitations, we believe our findings highlight important aspects surrounding parental attitudes towards consent in neonatal research.
Conclusion
Parents believe informed consent is an important prerequisite to neonatal research. Current consent methods are undermined by difficulties with timing and the vulnerability of parents in the perinatal period. Neonatal research continues to progress at a rapid pace. Incorporating parental views into the consent process is key to advancing newborn care.
References
Footnotes
Funding This work is supported by a Science Foundation Ireland Research Award (INFANT-12/RC/2272).
Competing interests None declared.
Ethics approval Clinical Research Ethics Commitee of the Cork Teaching Hospitals.
Provenance and peer review Not commissioned; externally peer reviewed.