There has been increasing academic and policy interest in the ethics of research over the past few decades. However, attention has largely been focussed on consent [1], and issues that have been neglected relate to ethics of evidence and clinical trials, and the role of other moral requirements such as equipoise and risk. Over 30 years ago Benjamin Freedman introduced the concept of equipoise as “uncertainty or equal belief that there is equal balance between the two treatments in the trial and that one is equally effective than the other” [2].
Whilst this concept has since served as a foundation of research ethics a debate has persisted over whether collective uncertainty amongst recruiting clinicians is sufficient or whether each individual clinician themselves needed to exist in a personal state of uncertainty. Freedman referred to ‘individual equipoise’ as when an individual clinician is uncertain and ‘collective or clinical equipoise’ as when a community of clinicians are uncertain. This is termed ‘clinical or collective equipoise’ and often regarded as the most robust ethical principle on which conducting medical research and randomisation can be morally justified [3]. Appropriately executed RCTs are generally accepted to be the most dependable approach for comparing health technologies, Lilford et al [4], while equipoise as the state of epistemic uncertainty is traditionally regarded as both the necessary and sufficient ethical condition stipulated by regulators such as Food and Drug Administration (FDA) Agency and National Institute of Clinical Excellence (NICE) to justify randomisation in clinical trials [4].
The scientific community continues to endorse clinical or collective equipoise as the most rigorous and ethical basis for undertaking randomised clinical trials based on the rationale that collective uncertainty accommodates the apparent fragility of individual equipoise [5]. However, acknowledging that there must be a scientific rationale behind an intervention to bother to trial it at all, there has recently been a resurgence of interest in equipoise due to new trial designs. The promise of innovative trial designs, such as Response Adaptive Randomisation (RAR) [6], is that they offer a more ethical alternative conventional RCTs, in that they can use accumulating data and respond to threats to individual equipoise [7]. Deng et al [8] propose that regulatory bodies should look at more innovative trial designs to cope with the above problem. For example, a response-adaptive cross-over trial design utilises the outcome data to decide whether to move any patient into the most effective treatment arm, so that the patient’s exposure to the inferior arm is reduced and, as a consequence, minimises the risk to the patient whilst the clinical trial is underway. Some argue that this approach could be justified, even though there would be lack of individual equipoise, if there would be a risk that the patient would be denied treatment by remaining in the inferior arm and where no other treatment exists [9]. A study by Legocki et al [10] concluded that although clinical experts accept advantages of Adaptive Clinical Trials (ACT) such as having great potential to efficiently identify patients who will be helped most by specific treatments [11], however, ACTs have a potential of disrupting equipoise. They also acknowledge other ethical challenges given differences in the weightings respondents placed on randomisation ratios according to clinical specialties.
As the controversy and debate continues, Hey et al [6] state that “many clinical trials include procedures with some level of “net risk” to participants, meaning that the procedures are done purely for research purposes and hence do not promote participants’ “best clinical interests”. They propose ‘Net Risk’ as an alternative to the concept of clinical equipoise.
Whichever conception is adopted, De Meulemeester et al [12] in their cross-sectional analysis of published literature on RCTs conclude that equipoise, although widely accepted as a scientific criteria and moral justification for conducting trials, was mentioned inconsistently and was often misunderstood when reported. De Meulemeester et al [12] argue that the utility of the concept of clinical equipoise as an ethical standard for justification of RCTs should therefore be challenged.
There is little analytic or empirical literature on the potential role of equipoise in justifying randomisation within two arms where the intervention is psychological intervention plus Treatment As Usual (TAU) versus TAU (which can include drug therapy). Indeed, this is the first published study of its type in the cultural and religious context of Pakistan.
Self-harm is a major risk factor for eventual suicide in Pakistan, and the prevention of self-harm is therefore a key focus for suicide prevention efforts [13]. Young people are especially at risk of suicide and self-harm [14].. For example, the prevalence of self-harm over three months in young people in India was 3.9–25.4%[i][13]. As such young people with a history of self-harm are a priority group for interventions. There is growing evidence that psychological therapies, including those based upon Cognitive Behavioural Therapy (CBT) principles, can help prevent further self-harm in those at risk, including in young people [15]. However, while there have been robust RCTs establishing efficacy, the available evidence has been limited to western and higher income countries. Only one study with adolescents SH (12–18) from a Low and Middle Income Country (LMIC) is relevant to how we think about implementation methods of talking therapies in this age group [16].
Based on the above analysis and importance of the issue, the current paper seeks to investigate whether clinicians recruiting patients to an RCT of YCMAP (The Y-CMAP Programme has been culturally adapted with permission from a self-help guide called “Life After Self-Harm” and “Cutting down: A CBT workbook for treating young people who self-harm” [17] to fit with the client’s problems and primarily utilises problem-solving, cognitive-behavioral assessments of self-harm, and dialectical therapy strategies to bring about change) and those treating patients not involved in the trial had a personal preference for the intervention or whether they were in a state of ‘individual’ equipoise. For those who did express a preference for a particular arm, we were interested in which one they favoured, and whether they regarded participation in an RCT as scientifically and ethically important and necessary despite these personal preferences.
Objectives
The objectives of the survey were to 1) assess the preferences of clinicians involved in recruiting to a RCT, and those not involved in the trial, regarding whether they preferred YCMAP plus TAU or TAU only for young patients at risk of self-harm and suicide and 2) to determine the extent to which there is collective uncertainty in clinicians (i.e. equipoise) to justify conducting a RCT despite personal preferences.
[i] Wide range may denote uncertainty of scale