Article Text
Abstract
As guidelines for neonatal resuscitation evolve from a growing evidence base, clinicians must ensure that practice is closely aligned with the available evidence, based on methodologically sound and ethically conducted research. This paper reviews ethical, legal and risk-management issues arising during the design of a quality-assurance project to make video recordings of neonatal resuscitations after high-risk deliveries. The issues, which affect patients, researchers, staff and the hospital at large, include the following: 1) Informed consent for research involving emergency procedures is often not possible, for lack of time to provide sufficient information. The mental capacity of the subject or parent may be compromised by the impending emergency, and freedom of choice is threatened by the time pressure to consent. 2) Video recording of the inevitable medical errors raises issues of whether participating staff may be identifiable and accountable, affecting their willingness to participate in such research. The approach to staff participation and identification is reviewed. 3) The use of video data for education threatens the privacy of research subjects. The ethics of maintaining privacy is balanced with the ethics of using the data to improve practice of resuscitation. 4) The research subjects (patients, or the staff whose performance is being monitored) must be defined. 5) There are legal and ethical aspects of management and ownership of data. 6) The role of the Human Research Ethics Committee in protecting the research subject and possibly the medicolegal interests of the hospital is discussed. This paper reviews the literature and discusses the issues.
Statistics from Altmetric.com
In 2007, the Department of Paediatrics, Mercy Hospital for Women, embarked on a project of video recording of neonatal resuscitation. In planning this project a number of ethical, legal and risk-management issues arose, affecting patients, researchers, staff and the hospital at large. These issues concerned the validity of the consent process, the privacy of staff and research subjects, the management of video data and the role of the Human Research and Ethics Committee (HREC). We review the literature surrounding these issues, explore how they relate to our project, and discuss our approach to them.
OUR PROJECT
This project entailed the video recording of resuscitation after high-risk deliveries, primarily for the purpose of quality assurance. These were predominantly the births of premature infants or other deliveries in which the likelihood of the need for resuscitation could be predicted in sufficient time to allow consent processes and the organisation of the necessary team. A number of deliveries were recorded at short notice, when unexpected emergencies occurred at a time when the team and equipment were available.
A digital video camera was fixed to the radiant warmer, the field of view including the bed, therefore capturing only the hands of the operators and the baby. Sound recordings were also made and voices could be recognised.
Video data were assessed with a validated scoring sheet and the performance of each resuscitation was analysed. A secondary objective was to use the data for intradepartmental education and quality assurance meetings.
PATIENT CONSENT
We review here the principles of valid consent, describe various approaches and finally argue that, despite the perceived low clinical risk to the subject in our study, informed consent is warranted.
The anticipated birth of a baby requiring resuscitation is an event that is preceded by much anxiety among clinicians and parents and is not a time conducive to valid informed consent. The issues to be balanced in the conduct of emergency neonatal research include the best interests and freedom of choice of the subject and family, the suitability of parents as proxies, the ethical validity of sometimes partially informed consent and the need to advance neonatal knowledge and practice.
The Nuremberg code, subsequently advanced by the Helsinki declaration (http://www.wma.net/e/policy/b3.htm), was devised 60 years ago to guide the ethical conduct of human research. Yet physicians still struggle to satisfy these guidelines, and more recent guidelines, as medical care and research become more complex.1–4 The three principles that underpin valid consent are freedom of choice, provision of sufficient information and having the mental capacity,5 all of which are compromised when recruiting subjects for emergency neonatal research.
Recruitment for trials in neonatal resuscitation is necessarily attempted when the validity of consent is compromised, such as an impending high-risk delivery, which is often unexpected and poses substantial risk for both mother and baby. Subjects’ ability to take in sufficient information and to make informed decisions when time is short is flawed, as was demonstrated by retrospective interview with both parents and clinicians about their participation in neonatal research.4 Parents of 200 babies from across Europe who had been asked to participate in neonatal research were interviewed to assess the validity of their consent. Over 50% of parents who were asked to participate in emergency research identified problems in the domains of competence, information and understanding, and nearly a quarter were concerned about voluntariness. Clinicians also recognised flaws in the process, with only three of 107 having no problem with obtaining valid consent. Problems with adequacy of information, competence in decision-making, understanding and voluntariness of participation in emergency research are commonly reported. Overwhelming emotional distress may burden the process, and a parent’s decision to consent may be influenced by perceived obligations to the clinician, the neonatal unit and society and by their own desperation.5 The power imbalance between clinical staff and families is at its greatest at times of critical clinical need, especially if no prior relationship has been able to be established.
Several approaches have been proposed to overcome such obstacles in obtaining informed consent during neonatal emergencies,6 7 as outlined in Appendix 1. The request for consent from potential subjects or their parents in antenatal clinics satisfies the requirement for mental competency, as the family are not under duress, and provides time for a greater exploration and understanding of the information. It provides more time for the decision-making process but has the negative implications of causing unnecessary anxiety about delivery to a large number of potential research subjects, for whom the chance of having a high-risk delivery is small. Another concern is that antenatal consent about a hypothetical situation does not have the immediate relevance that is perhaps necessary for truly informed consent.
Alternative methods attempt to remove the burden of information-processing and decision-making for parents. For instance, the waiving of parental consent has been considered in randomised controlled trials that examine two equally acceptable treatments.7 This approach relies on both treatment arms providing interventions that are commonly used. It has the significant disadvantage of being paternalistic, in that it assumes that the researchers and/or the HREC have managed to assess and understand all factors that the family may consider to be ethically or practically important. This is a significant assumption, especially in a multicultural society, in which no assumptions can be made about families’ views of the meaning or value of research or of the importance of factors such as privacy or other issues, such as the making of images.
Such considerations are not immediately relevant to a study that will not change the care of a particular baby and that is inherently “low risk”. Different risk profiles may, however, have very different meanings to some families, who may value factors not immediately obvious to the researchers or the HREC. It is important to note that despite seeing the value of neonatal research, parents overwhelmingly desire input into the decision-making process.4 8 In the study by Morley and colleagues,8 parents were asked about their opinions of being involved in multiple research studies. Forty per cent of respondents were happy to be involved in several studies, but, importantly, 98% of parents wanted to decide about this participation. Parents are also a valuable protection for their children against potentially harmful or unethical research.
HRECs have been considered an alternative as providers of consent.7 HRECs are entities that are independent, expert and informed and could shoulder the burden of responsibility. The HREC evaluates the scientific merit and protects the subject from harm but cannot ultimately ensure the autonomy of the subject.
The Australian national statement on ethical conduct in human research (http://www.nhmrc.gov.au/PUBLICATIONS/synopses/e35syn.htm) was developed by the National Human Medical Research Council (NHMRC), the Australian Research Council and the Australian Vice Chancellor’s Committee. Section 2.3.6 outlines items that must be satisfied before consent can be waived. It specifies that so long as the study is low risk, benefits outweigh the risk of harm and obtaining consent is impractical, then consent may be waived. In addition, the privacy of the participants must be protected and the data kept confidential.
Truog and colleagues suggest conditions allowing informed consent to be waived.9 They include the following:
the treatments can be used outside the trial without consent
the treatments do not confer more than minimal additional risk when compared with the alternatives
genuine equipoise exists
an institutional review board determines that no preference exists between two treatments
patients are informed of these standards by the treating institutions.
Policy B26 of the Helsinki declaration (www.wma.net/e/policy/b3.htm) also provides for research to be undertaken without consent, including proxy consent, but only if the physical or mental condition that prevents obtaining informed consent is a necessary characteristic of the research population—arguably, neonatal emergencies, although the importance of proxy consent from parents is a core value in paediatrics, if valid consent can be obtained.
In 1996 the US Food and Drug Administration ruled on circumstances where exclusion from informed consent could be justified so as to provide protection to those with diminished autonomy and to permit emergency research (www.hhs.gov/ohrp/documents/100296a.pdf). Appendix 2 lists those criteria. Its application has been tested in a hypothetical setting of parents with a child in intensive care,10 when 63% agreed to participation in research under exclusion from informed consent with a prospective opt-out clause. Community consultation is important in this process, as differing cultural and religious approaches to research must be fully accommodated, together with the issues of alienation and power imbalance in dealing with minority communities.
Others have suggested providing the information to potential subjects in a stepwise fashion or at multiple times throughout a study11 with the intention of strengthening autonomy and informed consent. This would be difficult to achieve in emergency research, involving a “one-off” intervention.
Parental preferences for consent to neonatal research have been examined8 12 13 In one study,12 107 Canadian expectant parents or parents of infants requiring intensive or nursery care were asked to determine their level of comfort with 15 hypothetical research scenarios and to specify factors important in consent to research. Being well informed in advance was a central theme. Having the opportunity to decide, trusting the researcher or clinician and knowing that there may be benefit for the child were also important. Prospective consent, either immediate or during the antenatal period, was clearly most preferred; waived or retrospective consent scored less well. Opting out scored the least, which indicates the importance of choice in the consent process. Societal benefit, personal benefit, improved care of future babies5 and perceived risk of harm were also factors shown to be important in the decision to participate in research.13
Previous experience exists in video recording of resuscitation in neonatal units and in paediatric and adult emergency departments over the past 20 years in both Australia and the USA.14–21 The primary objectives of our study closely align with these. All the studies were guided by local HRECs or hospital administration with respect to obtaining consent. Consent was commonly deemed unnecessary, as the nature of the study was an audit of standard practice and clinical quality assurance.17–20 Hospital legal opinion was sought with regard to handling of data20 and privacy of staff.18 Data deletion was common17 19 21 and privacy legislation was deemed irrelevant on the grounds that the video recording was an audit tool.
We believe that participation and particularly storage of data require informed consent at some level. This is particularly important because data storage and its subsequent use as an educational tool threatens the privacy of both families and staff. Hair-splitting about whether audit is research and does not require consent cannot change the very real privacy risks to family and staff inherent in such studies, together with the fact that families will inevitably discover that a critical event in their lives has been filmed. Some will want access to the data for their own purposes and all are entitled to a degree of control over how such personal information may be handled and the uses to which it may be put. Without clear resolution of these issues, there is the possibility of causing real trauma to some families, not least through a lack of clarity concerning their entitlement to subsequent access to such data.
Our approach was to obtain prospective consent from parents of all eligible subjects; this would almost invariably occur during duress of some kind. Where circumstances permitted, we sought consent as early as feasible so as to remove the immediacy. In the event of acute emergencies or time restriction, we sought retrospective consent and destroyed the data if consent was denied. While this approach threatens freedom of choice, it maximised the opportunity to collect data in inherently unpredictable situations. The principle of freedom of choice was only partly maintained, in cases in which retrospective consent was sought. This was justified by the assumption that the intervention had little foreseeable adverse impact on the standard of care provided.
The researchers are active clinicians in our unit. All have contact with parents as carers and this dual relationship potentially creates problems, as families may be reluctant to refuse or revoke consent to studies being undertaken by medical staff caring for their baby. This is common to much clinical research being undertaken in neonatology. Our approach to this, and other studies at Mercy Hospital for Women, is that consent for research participation is not obtained by staff who are rostered to provide clinical care at the time of enrolment. This partially separates the two functions, although it is more difficult in the case of retrospective consent procedures, in which situation relationships may have been formed with multiple treating clinicians.
Guidelines such as those of Truog and colleagues have mostly been devised around the issue of risk, to manage situations in which risk is low but it is difficult or impractical to obtain informed consent. Our research is of low risk to the baby involved, being purely observational, and would therefore be encompassed in such guidelines, but the key issue in this case is not that of direct clinical risk to the baby as a subject of research. The situation is best seen as one of low clinical risk to the primary subject (the baby) but with considerable privacy risks for other parties (parents and staff). It could also be seen as one in which the primary subject of research is the clinical staff, for whom the issues are professional risk, medicolegal consequences and privacy, with the issues of the family’s and the baby’s privacy being important additional issues.
It could also be argued that clinician performance may be altered as a result of being observed—the Hawthorne effect. A Hawthorne effect could impact performance in either direction, as scrutiny could enhance vigilance or be distracting and thereby add another dimension to the concept of risk in this circumstance.
STAFF CONSENT
Using video data as a means of assessing the quality of resuscitation was perceived by some as threatening and risky in the event of litigation. During our study design phase, we conducted information sessions with medical and nursing staff to provide a forum in which concerns could be raised and to discuss the research aims and methodology. Reactions from staff were mixed, with most willing to participate, some refusing and some undecided. Concerns centred primarily on the identification of clinicians, accountability and the generation of “evidence” in the event of medical/nursing error. Also raised were the consent process, data management and availability and the identification of the subjects.
Our position on these issues evolved into the following -
When resuscitation occurs, staff attendance is recorded and medical notes are documented. The identification of staff present at resuscitation is therefore mandatory irrespective of the act of video recording.
Disclosure of medical error is standard practice whether it is captured on a video recording or not.
Video data is likely to document some medical errors. It is, however, more likely to provide evidence of good practice, rather than errors that may adversely affect outcome.
It is incumbent upon us to subject our performance to quality assurance.
If a staff member attending a high-risk delivery had refused to participate, we planned not to make a video record of these deliveries. Our approach was designed to ensure that staff did not feel pressured to participate and that they maintained involvement with neonatal resuscitation, with the hope that once the project commenced reservations would subside and participation would become accepted. Interestingly, many staff felt discomfort about being identifiable on video, despite their presence having been routinely documented in the medical records for many years.
DATA OWNERSHIP, MANAGEMENT AND DISCLOSURE OF ERROR
Our project, being based on individually identifiable data, raises the issue of the ownership and management of such data. We chose to retain the data in media file format. We minimised the identification of subjects and staff so the images could be used in educational activities within the hospital with maximal protection of privacy.
As previously noted, video recording has been considered exempt from freedom-of-information legislation in Australia and the USA18–20 and has thus often been destroyed after use. In Australia, freedom-of-information legislation exists at state and federal level that permits patients access to their records, with a few exceptions.22 Whether the data collected should be considered part of the medical record is controversial. The items that constitute medical records include progress notes, operative reports, diagnostic reports, letters and referrals, and may include photographs and information stored electronically.22 In the state of Victoria, the Health Records Act 2001 outlines what constitutes health information and how it should be handled. It could be interpreted to mean that video recordings, despite being used for research purposes, constitute part of the medical record and are therefore subject to the Act. This would mandate retention of the data for 25 years and making it accessible to the subjects.
The storage of such data potentially retains evidence of error in resuscitation that may not have been detected or thought relevant by the resuscitation team. The discovery of error and, equally important, areas requiring improvement is central to this process and leads to the issue of its disclosure. We argue that there is an ethical obligation to provide the information that a family may need in pursuing action in the event of any mishap, as the primary focus must always be the welfare of the patient and family.
The incidence of medicolegal action in our department is low, which is in part attributable to a culture of disclosure of adverse events as a result of error or unforeseen events. The ethical principles surrounding disclosure of error include fiduciary obligations, autonomy, truth-telling, respect for persons and professional standards.23 The benefits of disclosure extends to the patient and family, the practitioner and the healthcare system, yet the ethical obligation to disclose information may often be over-ridden by the perceived interests of the hospital and staff. For the patient, autonomy and the doctor–patient relationship are preserved by disclosure. Practitioners also benefit from being able to improve patient safety and potentially decrease litigation,24 and the healthcare system may benefit from increased confidence in the medical profession23 and by decreased costs of litigation.
It should also be noted that a leading reason for litigation (an expensive, draining and damaging experience for all concerned) is the need to discover “what happened” or perhaps also the need to really ensure that nothing has been concealed. Much of this could be short-circuited by open availability of information.
Kraman and Hamm reported from a Veterans Affairs Hospital in the USA24 that a shift in culture and policy towards active disclosure of medical errors resulted in less litigation. Comparable levels across other Veterans Affairs institutions with the traditional approach to disclosure were observed. However, it is recognised that these data were not validated and that the legal framework governing these institutions and physicians25 does not represent the majority of US medical institutions.26 Despite this, the growing movement towards active disclosure26 and subsequent improvements in patient safety practices are ethically valid, altruistic and in the best interests of the patient and potentially of the healthcare systems themselves.
We elected to retain the data for the purpose of education, potential further research and the best interests of the patient.
THE ROLE OF THE HREC
In Australia, the NHMRC defines the function of HRECs in reviewing research (http://www.nhmrc.gov.au/ethics/human/hrecs/index.htm). Chapter 5.2 of the National Statement on Ethical Conduct in Human Research (2007) (http://www.nhmrc.gov.au/PUBLICATIONS/synopses/e72syn.htm) is dedicated to the role of the HREC. The main responsibilities encompass ensuring good governance of the committee, ensuring the adherence of the researchers to their responsibilities and preserving the subjects’ best interests.
In assessing our study, the HREC highlighted risk-management issues regarding data management. Issues surrounding medicolegal risk for the hospital were also raised and discussed with researchers, the view of some being that these were factors to be considered in the decision of the HREC concerning acceptance or rejection of our proposal. The committee ultimately decided that it was not the role of the HREC to manage any risk posed to the hospital by the retention of the data and that their primary role was protection of the participants’ interests in the conduct of research.
The composition of institutional HRECs in Australia commonly includes representatives of hospital administration, senior clinicians and legal representatives. The legal representatives are not usually those who would normally advise or represent the hospital in any action, their function on the HREC being to advise on legal matters relevant to consent, privacy and statutory obligations in the conduct of research. This mixture of interest groups can lead to some role confusion, there being multiple pressures acting on and through such representatives. Clinicians present to advise on the practicability and scientific and ethical validity of studies may identify with the need to reduce medicolegal risk for fellow clinicians, while administrators and lawyers may feel an obligation to protect the hospital from any risk, even at the risk of compromising valuable research.
We support the view ultimately taken by our HREC—that its over-riding function is the protection of subjects and that considerations of medicolegal safety have to be dealt with elsewhere, if indeed they should be allowed into such decisions at all. Given the nature and composition of hospital HRECs and the need to ensure that ethically appropriate research is not hindered by such considerations, there is perhaps a need for more formal, open exploration of this issue than has hitherto occurred.
CONCLUSIONS
To date, 24 parents have been approached to participate in this project, of whom 23 have consented. One family refused and three families, when approached antenatally, continued the pregnancy to a low-risk gestation. Twenty video recordings have been obtained. Sixteen families were consented retrospectively, three prospectively but in the intrapartum period and one prospectively in the antenatal period. Altruism was the main reason for consenting to participation and most families appreciated that the research would not benefit their baby. No staff members refused to participate.
One family has sought access to the recordings. This family’s child had a severe diaphragmatic hernia, was successfully resuscitated and subsequently died in the first days of life. The reason for the request was that the video recording of the resuscitation proved to be the only images available of the baby while still alive. The images were made available on compassionate grounds, after considerable discussion at administrative and HREC level.
Research involving video recordings to assess the performance of resuscitation is a worthwhile endeavour with great value in the training of medical and nursing staff but raises the ethical and legal issues of informed consent, privacy, data management and the role of the HREC.
Although it can be challenging to obtain valid informed consent, this can be achieved, and family autonomy, patient safety and respect for the privacy of both patients and staff can best be served if significant efforts are made to include all parties in consent to such procedures.
Appendix 1
Modes of consent in emergency research include prospective, retrospective, waived consent and an opt-out approach.
Prospective consent occurs prior to the performed study. Prospective consent can also be obtained from potential subjects from a wider population in the event they meet eligibility, for instance, in antenatal clinics.
Retrospective consent involves the obtaining of consent after subject recruitment has occurred.
Waived consent applies when emergency circumstances such as a life-threatening illness preclude the subject or proxy from providing consent.
“Opting out” assumes automatic inclusion, with subjects being able to withdraw consent prior to or after the trial has commenced6 This method aims to remove the burden of decision-making from the research subjects
Appendix 2
Summary of the exception from informed consent requirements for emergency research:
1) The patients are in a life-threatening situation, available treatments are unproven or unsatisfactory, and the collection of valid scientific evidence, which may include evidence obtained through randomized placebo-controlled investigations, is necessary to determine the safety and efficacy of particular interventions.
2) Obtaining informed consent is not feasible because
a) the patients will not be able to give their informed consent as a result of their medical condition;
b) the intervention involved in the research must be administered before consent from the patients’ legally authorized representatives is feasible; and
c) there is no reasonable way to identify prospectively the individuals likely to become eligible for participation in the research.
3) Participation in the research holds the prospect of direct benefit because
a) patients are in a life-threatening situation that necessitates intervention;
b) appropriate animal and other preclinical studies have been conducted, and the information derived from those studies and related evidence support the potential for the intervention to provide a direct benefit to the individual patients; and
c) risks associated with the research are reasonable in relation to what is known about the medical condition of the potential class of patients, the risks and benefits of standard therapy, if any, and what is known about the risks and benefits of the proposed intervention or activity.
4) The research could not practicably be carried out without the waiver.
5) The research protocol defines the potential therapeutic window, and the investigator has committed to attempting to contact a legally authorized representative for each patient within that window of time.
6) The Institutional Review Board has reviewed and approved informed consent documents to be used in the cases when informed consent from the patient or his or her representative is feasible.
7) Additional protections will be provided including at least the following:
a) consultation with representatives of the community in which the research will be conducted and from which the patients will be drawn;
b) public disclosure to the communities in which the research will take place and from which the patients will be drawn;
c) public disclosure at the completion of the investigation of the study results, including demographic characteristics of the research population;
d) establishment of an independent data monitoring committee;
e) commitment by the investigator, if feasible, to attempt to contact, within the therapeutic window, a family member of the patient and ask whether he or she objects to the patient’s participation in the research
REFERENCES
Footnotes
Competing interests: None declared.