The ethics of ‘Trials within Cohorts’ (TwiCs): 2nd international symposium

On 7-8 November 2016, 60 people with an interest in the ‘Trials within Cohorts’ (TwiCs) approach for randomised controlled trial design met in London. The purpose of this 2 TwiCs international symposium was to share perspectives and experiences on ethical aspects of the TwiCs design, discuss how TwiCs relate to the current ethical framework, provide a forum in which to discuss and debate ethical issues and identify future directions for conceptual and empirical research. The symposium was supported by the Wellcome Trust and the NIHR CLAHRC Yorkshire and Humber and organised by members of the TwiCs network led by Clare Relton and attended by people from the UK, the Netherlands, Norway, Canada and USA. The two-day symposium enabled an international group to meet and share experiences of the TwiCs design (also known as the ‘cohort multiple RCT design’), and to discuss plans for future research. Over the two days, invited plenary talks were interspersed by discussions, posters and mini presentations from bioethicists, triallists and health research regulators. Key findings of the symposium were: (1) It is possible to make a compelling case to ethics committees that TwiCs designs are appropriate and ethical; (2) The importance of wider considerations around the ethics of inefficient trial designs; and (3) some questions about the ethical requirements for content and timing of informed consent for a study using the TwiCs design need to be decided on a case-by-case basis. Main report On 7-8 November 2016, 60 people with an interest in the ‘Trials within Cohorts’ (TwiCs) design met in London for the 2 TwiCs international symposium. The symposium was supported by the Wellcome Trust and NIHR CLAHRC Yorkshire and Humber and organised by members of the TwiCs network led by Clare Relton. As well as UK participants, people came from the Netherlands, Norway, Canada and USA. Over the two days, the invited plenary talks were interspersed by discussions, posters and mini presentations from bioethicists, triallists and health research regulators. © The Author(s). 2017 Open Access This artic International License (http://creativecommons reproduction in any medium, provided you g the Creative Commons license, and indicate if (http://creativecommons.org/publicdomain/ze On the first day (7 November, 2016), Jon Nicholl (University of Sheffield, UK) opened the meeting, welcoming all to the symposium. He described how the first international symposium in 2014 brought together triallists using the design for the first time, and led to this, the 2 symposium which aimed to provide a forum in which to discuss and debate ethical issues including how the TwiCs approach relates to the current ethical framework. What are TwiCs? Clare Relton (University of Sheffield, UK) set the scene by outlining the Trials within Cohorts (TwiCs) approach as described in the original article (Fig. 1) in the BMJ in 2010 [1], and the 7 key features of the design: (I) Recruitment of a large observational cohort of patients/ people with the condition of interest (II) Regular measurement of outcomes for the whole cohort (III) Capacity for multiple randomised controlled trials over time. Then for each randomised controlled trial: (IV) Identification of all eligible people in the cohort (V) Random selection of some individuals from all eligible people in the cohort, who are then offered the trial intervention (VI) Comparison of the outcomes in randomly selected people with the outcomes in eligible people not randomly selected; that is, those receiving usual care (VII) “Patient centred” informed consent; that is, the process of obtaining information and consent aims to replicate that in routine health care as far as is possible. Ethics in current use Clare described how more than 20 studies using the TwiCs design now had ethics board approval from boards in Australia, Canada, Finland, France, Germany, Mexico, Netherlands, Spain, UK, and the USA. These studies were recruiting cohort populations (e.g. early life, children and adolescents, young indigenous, adults, older people) in a variety of settings (e.g., hospital, primary care), in order to facilitate trials in diverse health areas (e.g., attention deficit hyperactivity disorder, breast cancer, colo-rectal cancer, bone metastases, depression, hepatitis C, HIV, hip fracture, falls prevention, long term conditions, severe mental illness, scleroderma). Embedded within these cohorts, there were at least 20 randomised trials testing a wide range of interventions, and various approaches to informed consent were being used in these studies. Jon Nicholl (University of Sheffield, UK) then gave an example of an emergency medicine research study where it was not possible to obtain informed consent prior to randomisation and have a viable trial, which illustrated that informed consent for participation in a trial is not always required. Jon argued that although all trial participants in TwiCs should receive information about data collection, storage and sharing and all other non-therapeutic research processes, only those in the intervention group need to receive information about the intervention. TwiCs designs randomly select from the cohort and offer the intervention being tested, those unselected are not actually allocated to ‘treatment as usual’, so there is no ethical obligation to tell those unselected about those who were selected, or about le is distributed under the terms of the Creative Commons Attribution 4.0 .org/licenses/by/4.0/), which permits unrestricted use, distribution, and ive appropriate credit to the original author(s) and the source, provide a link to changes were made. The Creative Commons Public Domain Dedication waiver ro/1.0/) applies to the data made available in this article, unless otherwise stated. Trials 2017, 18(Suppl 2):244 Page 2 of 10 the treatment they are not being offered. Jon offered the analogy of lottery winners who are ‘selected’, where there is no sense in which ticket holders who don’t win are ‘allocated’ to a losers group. Jon concluded by offering a ‘Sheffield’ position statement for discussion “In cohort trials, members of the cohort who are not selected to be offered a new treatment do not need to be told about the trial intervention (s)”. Merrick Zwarenstein (Western University, Ontario, Canada) set the context for the design by clarifying how pragmatic trials provide evidence to inform decision making and explanatory trials test whether or not an intervention causes an outcome. Merrick suggested that the PRagmatic Explanatory Continuum Indicator Summary -2 (PRECIS-2) framework could help designers of TwiCs trials match their design to their intentions. James Flory (Memorial Sloan Kettering Cancer Centre, New York, USA) outlined his review [2] of proposals for randomised controlled trials (RCTs) where randomisation occurred without prior information being given that interventions would be allocated at random (Randomisation without Consent). He described 6 different approaches found in the literature including emergency medicine research, Zelen designs and TwiCs designs. The morning session concluded with two researchers reporting their practical experiences with the TwiCs design and the ethical questions that were generated and/or resolved through the use of the design. Rudolf Uher (Dalhousie University, Canada) described the FORBOW cohort of youth at high risk of severe mental illness and the first RCT (Skills for Wellness) embedded within this cohort. He described the advantages of using the design in a situation where most children at risk were not seeking help. No concerns had been raised about the TwiCs design during institutional review board process for FORBOW. Then Anne May (University Medical Centre, Utrecht, Netherlands) described the exercise-based FIT trial which is embedded within the hospital-based breast cancer ‘UMBRELLA’ cohort which uses the staged consent version of the TwiCs design [3]. She highlighted the possible pros (fast recruitment, no contamination) and cons (nonacceptance in the intervention group) using the TwiCs design. Ethical perspectives The afternoon session began with bioethicist Scott Kim (National Institute for Health, USA) who provided an overview of the ethical questions that pragmatic RCTs raise and an ethical analysis of two variations of TwiCs designs, those where information about (and consent for) future RCTs (i.e. the possibility of being randomised to the offer of a therapeutic intervention) was provided at enrolment to the cohort , and those where this information was only provided after randomisation to those in the intervention group. He concluded with system level ethical questions for broad population based TwiCs cohorts and learning healthcare systems. This was followed by Shaun Treweek (University of Aberdeen, Scotland) who focussed on the wider ethical question of the ethics of inefficiency, describing the lack of evidence to inform trial process decisions (e.g. ‘opt out’ vs ‘opt in’ for recruitment), and highlighting the potential waste of resources and participant goodwill. He argued that inefficiency is an ethical problem and how methodologists must generate evidence to support their decisions about trial processes. Tjeerd van Staa (University of Manchester, UK) described how TwiCs designs are suited for pragmatic trials in the era of big and ubiquitous data collection, but highlighted that refusal of treatment in the intervention arm could result in bias and loss of power. Andrew Vickers (Memorial Sloan Kettering Cancer Centre, USA) emphasised the benefits of integrating patient reported outcomes into routine clinical practice for optimizing clinical care, reusing these data for observational and experimental research such as TwiCs, and improving response rates. Towards the end of the day, Kirsty Wydenbach from the MHRA (Medicines & Healthcare products Regulatory Authority) in the UK, emphasised that they were familiar w