Equipoise in action: A qualitative synthesis of clinicians’ practices across six randomised controlled trials

Background Randomised controlled trials (RCTs) are essential for evidence-based medicine, and increasingly rely on front-line clinicians to recruit eligible patients. Clinicians’ difficulties with negotiating equipoise is assumed to undermine recruitment, although these issues have not yet been empirically investigated in the context of observable events. We aimed to investigate how clinicians conveyed equipoise during RCT recruitment consultations across six RCTs, with a view to i) identifying practices that supported or hindered equipoise communication, and ii) exploring how clinicians’ reported intentions compared with their actual practices. Methods and Findings : Six pragmatic, UK-based RCTs were purposefully selected to include several clinical specialties (e.g. oncology, surgery) and types of treatment comparison. The RCTs were all based in secondary-care hospitals (n=16) around the UK. Clinicians recruiting to the RCTs were interviewed (n=23) to understand their individual sense of equipoise about the RCT treatments, and their intentions for communicating equipoise to patients. Consultations in which these clinicians presented the RCT to trial-eligible patients were audio-recorded (n=105). The consultations were analysed using thematic and content analysis approaches to identify practices that supported or challenged equipoise communication. A sample of consultations was independently coded by three researchers to optimise reliability in reported findings. Clinicians and patients provided full written consent to be interviewed and have appointments audio-recorded. Interviews revealed that clinicians’ sense of equipoise varied: although all were uncertain about which trial treatment was optimal, they expressed different levels of uncertainty, ranging from complete ambivalence to clear beliefs that one treatment was superior. Irrespective of their personal views, all clinicians intended to set their personal biases aside to convey trial treatments neutrally to patients (in accordance with existing evidence). However, equipoise was omitted or compromised in 48/105 (46%) of the recorded consultations. Three commonly-recurring practices compromised equipoise communication across the RCTs, irrespective of clinical context. First, equipoise was overridden by clinicians offering treatment recommendations when patients appeared unsure how to proceed, or asked for the clinician’s expert advice. Second, clinicians contradicted equipoise by presenting imbalanced descriptions of trial treatments that conflicted with scientific information stated in the RCT protocols. Third, equipoise was undermined by clinicians disclosing their personal opinions or predictions about trial outcomes, based on their intuition and experience. These broad practices were particularly demonstrated by clinicians who had indicated in interviews that they held less balanced views about trial treatments. The findings were limited by clinicians volunteering to take part in the study who might have had a particular interest in improving their communication skills. However, the frequency of occurrence of these issues across the RCTs lends some assurance that these practices are likely to be transferable and relevant to clinicians more widely. Conclusions Communicating equipoise is a challenging process that is easily disrupted. Clinicians’ personal views about trial treatments unwittingly encroached on their ability to convey equipoise to patients. Clinicians should be encouraged to reflect on personal biases and be mindful of the common ways in which these can arise in their discussions with patients. Identification of common pitfalls that recurred irrespective of RCT context indicate opportunities for specific training in communication skills that is broadly applicable to a wide clinical audience.