Man vs. manikin revisited – the ethical boundaries of simulating difficult airways in patients

Airway-related research in anaesthesia is, arguably, dominated by manikin-based studies, despite clearly articulated concerns over the validity of extrapolating their findings to real patients 1, 2. Notwithstanding advances in simulation technology, manikins can still not accurately reproduce patient airway anatomy, lacking, in particular, the ability to replicate the feel and compliance of human tissues, cervical spine mobility, the variability of anatomical proportions and pathologic changes 3-5. As has been said, “Manikin studies are all well and good, providing your patients are manikins” 6. Despite this critical lack of fidelity, airway simulators are omnipresent in the evaluation of airway tools and techniques, and have infiltrated airway management skills training and clinician performance assessment, even extending into guideline development 7. The benefits of employing manikins in such studies are self-evident and already well described, namely: the reproducibility of experimental conditions; the scope for repeated or unlimited interventions; the opportunity to practice rare clinical events; increased safety and the absence of potential patient harm; and avoidance of ethical approval and a laborious recruitment process. Accordingly, it is comparatively rare for investigators in this field to undertake patient-based research, especially when simulating difficult tracheal intubation, as the risk of an adverse patient event and the inherent ethical issues are frequently deemed insurmountable. To reconcile the two approaches, Cook argued in an editorial that ‘bench’ and manikin studies should be used first as screening tests for airway devices and be followed by studies with human participants, before adoption into clinical practice 8. One such study is published in this month's issue of Anaesthesia. Kleine-Brueggeney et al. 9 randomly allocated patients to two different models of rigid laryngoscope (Bonfils and SensaScope®) in a simulated difficult airway situation. Whilst its value as a clinical study is undeniable, this must be balanced with careful consideration of the ethical acceptability of simulating a difficult airway in patients for the sole purpose of research. Intuitively, the practice of simulating a difficult airway in a healthy patient with a predicted ‘normal’ airway in a controlled situation, seems more ethically appealing (and undoubtedly safer) than experimenting on patients with known difficult airways or the physical characteristics predictive of one. Simulating a difficult airway in patients for the purpose of clinical study is not a new concept, with several approaches described. Two commonly adopted methods include: undertaking direct laryngoscopy with a Macintosh laryngoscope blade, then reducing traction, allowing the epiglottis to fall posteriorly, thereby creating an obscured view of the vocal cords with an adjustable degree of difficulty 10, 11; and application of a rigid cervical collar, as favoured by Kleine-Brueggeney et al., designed to create a physical restriction of neck flexion, head extension and mouth opening 12, 13 as might occur in a patient with cervical spine instability. Since there are advantages and disadvantages of both of these techniques (in terms of airway accessibility, practicality, reliability, reproducibility and generalisability), the choice of which technique to adopt is largely governed by the nature of the difficult airway intended and the type of airway device or technique under investigation. Regardless of the chosen method, the physician-researcher is obliged to critically evaluate the risk of potential harm as a direct consequence of this. Most studies employ some simple safeguards to reduce risk. Subject recruitment should be restricted to ASA 1-2 patients, and those with a history of airway-related problems or features suggestive of a potentially difficult tracheal intubation (limited mouth opening, reduced thyromental distance, limited neck movement, modified Mallampati score of 3 or 4, upper airway disease) should be excluded. During the study, multiple attempts at tracheal intubation should be avoided, and after one (or a maximum of two) failed attempts, the investigator must restore the airway to its normal condition and revert to the standard technique for tracheal intubation to prevent further harm. New devices should be evaluated in patients by experts only, and medical students, novices and junior staff should not be involved in airway studies (unless the primary aim is to investigate the learning of skills). Despite employing these simple measures, the risk of an adverse event cannot be eliminated and, therefore, the physician-researcher must consider what constitutes an acceptable risk to their patients. This tends to be greater in studies involving therapeutic interventions as it can be offset by the potential benefit of the intervention itself, whereas in a study where a non-therapeutic intervention is being employed (such that the patient will not personally benefit), the burden of risk must be kept to a minimum. In the study by Kleine-Brueggeney et al., patients incurred a high incidence of mucosal bleeding. This is a relatively common and generally mild complication of using a rigid scope for tracheal intubation. Nevertheless, given that a previous study investigating the Bullard rigid scope in patients with identically simulated rigid collar induced-difficult airways 13 demonstrated a similarly high incidence of tissue trauma, is it reasonable to subject these patients to such a predictable adverse event? Furthermore, repeated attempts with an unfamiliar device could convert an ‘easy’ (but artificially difficult) intubation into one that is impossible, due to oedema and bleeding, such that there is a small, but significant risk of major harm and even death in such studies. Were the patients fully cognisant of these possible complications? It is the responsibility of the physician-researcher and, ultimately the institutional review board, to decide whether the limits of acceptable practice are being breached. For example, a study investigating the use of suxamethonium for insertion of a supraglottic airway device 14 could be deemed to have crossed this boundary 15. Committees are often comprised of members with limited understanding of these issues and may not have a representative knowledgeable in the area, e.g. would a committee know that it is inappropriate to use suxamethonium for airway insertion in a normal patient undergoing elective surgery? Therefore, the physician-researcher must comply fully with the institutional review board requirements and ensure that clear, unambiguous study protocols are constructed such that the potential for adverse events is communicated explicitly. Specifically, the physician-researcher has an obligation to explain to the institutional review board what constitutes standard anaesthetic care and how their proposed protocol differs, along with the risks and benefits associated. Independent specialist review (e.g. by an experienced anaesthetist from another institution) would also provide additional assurances. The risk of harm associated with a study method must be balanced against the usefulness of the data derived from the study, as well as the true value of the study in the context of the research field in question. For example, a pioneering study, a novel device, or a novel technique that offers the possibility of innovative treatment or a new insight or a different perspective from pre-existing studies, may merit a higher degree of patient risk, within reason of course. After all, clinicians must uphold their duty to first do no harm 16. In fact, the responsibilities of the physician-researcher extend far beyond non-maleficence. In quasi-legal terms, the physician-researcher can be regarded as a fiduciary to their patient participants and, as such, is obliged to put the interests of their patients first, before their own academic goals. Despite the recent shift towards a model of shared decision-making, as long as the power and knowledge imbalance between physician and patient still exists, this rather paternalistic construct will remain relevant. Consequently, the investigators’ responsibility must be primarily as their physician and patient advocate and only secondarily as a researcher 17. In simple terms - an investigator may have a very interesting hypothesis to test but, if the patient is unlikely to gain from the intervention while possibly incurring harm, it may be breaching this responsibility. The World Medical Association Helsinki Declaration (2013) is explicit in this regard, echoing the points made above and concluding that: “The health of my patient will be my first consideration; it is the duty of the physician to promote and safeguard the health, well-being and rights of patients, including those who are involved in medical research; The primary purpose of medical research is to generate new knowledge, [but] this goal can never take precedence over the rights and interests of individual research subjects” 18. Differences of opinion can, of course, arise with individual researcher's interpretations of well-being or interests, but the overall message is clear – the patient must come first. Indeed, clinical ethics should be viewed as being inseparably linked with common morality and social order, such that general principles of avoiding causing pain or suffering, preventing harm occurring to others, and treating all persons with equal moral consideration should be as important in clinical studies as they are in day-to-day living. In pharmaceutical trials, patients may receive potentially beneficial or expensive treatments (that they would otherwise not be able to access) or they may have failed standard treatment modalities and the trial offers a novel, albeit experimental or unproven therapy. However, the motives of patients recruited for studies where they are unlikely to reap any personal benefit (and may incur harm) are somewhat different. They are either driven by the philanthropic precept of community spirit or solidarity, where they are willing to participate to improve the care of future patients (for the common good) rather than for personal gain or, rather less altruistically, by financial incentives. The recruitment and consent process must be rigorous and physician-researchers must be wary of co-ercion, over-zealous recruitment or misrepresentation (over-selling the benefits and under-selling the risks). Study participants must understand the methods and potential risks involved in deliberately creating a difficult airway, and must be able to distinguish between clinical practice and clinical research. Practice refers to any intervention designed to enhance the well-being of the patient (with reasonable expectation of success) whereas research (applicable here) refers to the process of hypothesis testing, through which conclusions can be drawn to produce a body of generalisable knowledge 19. With equipment studies, patients should also know whether the researcher stands to make any personal gain, apart from academic, such as a financial relationship with the device manufacturer. Taking into account the points discussed, we recommend that airway researchers should follow the proposed consensus on airway research ethics (CARE - Table 1). This is intended to promote good research practice in the area of airway management studies, providing physician-researchers with additional guidance on the ethical issues (and boundaries of acceptability) of recruiting and subjecting patients to airway interventions or techniques. It should assist physician-researchers in minimising potential risk to patient participants from the airway device/technique under investigation by ensuring that patient well-being is paramount. Finally, it should assist researchers, institutional review boards and journals in evaluating patient risk and balancing this with the value of any proposed study. It is widely acknowledged that the correlation between man and manikin is relatively poor and, therefore, the applicability of findings from airway simulation studies are very often limited. While patient-based studies are certainly preferable (and necessary, if a device is to meet the ADEPT criteria 20), they carry an inherent degree of risk to the participant, and patients must be made fully aware of this. It is questionable whether institutional review board members are always completely cognisant of the potential for harm, therefore it is incumbent upon journals and professional organisations (such as the Difficult Airway Society) to determine the safety standards for such airway studies and to ensure, where a potential conflict of interest exists, that the clinical usefulness of the study has been transparently balanced with the potential risk of harm to the patient. To this end, the proposed CARE guideline is designed to protect patients and enable researchers, institutional review boards and journals to make such critical evaluations. No external funding or competing interests declared. In light of the above points, and the importance of these issues with respect to patient safety, the journal Anaesthesia will insist that authors/investigators follow the above consensus on airway research ethics (CARE) when undertaking any airway studies submitted for publication in Anaesthesia. This will be added to our Author Guidelines (available at http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1365-2044/homepage/ForAuthors.html).