Planned study within a trial using a stepped wedge trial design: evaluating a decision aid on patients’ decisional conflict in PRIMETIME


Session type:


Indrani Bhattacharya1,Charlotte Coles2,Natalie Atkins3,Micki Hill3,Alexa Gillman3,Lisa Fox3,Anna Kirby4,Lesley Turner5,Hilary Stobart5,Jo Haviland3,Penny Hopwood3,Judith Bliss3
1Institute of Cancer Research,2Cambridge Oncology Centre,3Institute of Cancer Research - Clinical Trials and Statistics Unit,4Royal Marsden Hospital,5Independent Cancer Patients' Voice



PRIMETIME is a prospective biomarker directed cohort study aiming to identify a group of breast cancer patients who can safely avoid adjuvant breast radiotherapy following breast conserving surgery (1). This group is deemed to be at such a low risk of local relapse that the potential benefits of radiotherapy are unlikely to outweigh the risks. The uncertainty patients face regarding healthcare decisions is known as ‘decisional conflict’ (2). Decision aids are interventions which help patients to weigh up the risks and benefits of treatments (3). Evidence suggests decision aids reduce decisional conflict (3). The study within a trial (SWAT) concept enables trialists to conduct research embedded within a larger trial in an economic and efficient manner (4). This SWAT is designed to investigate the effect of a decision aid on patients’ decisional conflict within PRIMETIME.




The PRIMETIME SWAT will utilise a cluster stepped wedge trial design. The decision aid will be in video format. Decisional conflict will be assessed using a validated decisional conflict scale (2) in clusters prior to and following implementation of the decision aid. All clusters will receive the standard patient information sheets and be randomised to receiving the decision aid video at increasing intervals from when their centre began recruiting to PRIMETIME using minimisation. The primary outcome is reduction in decisional conflict. The secondary outcome is to assess acceptance of entry into PRIMETIME.


The cluster stepped wedge trial design ensures that by the end of the study all centres will have use of the decision aid as opposed to a parallel design which may be considered less favourable as some clusters would never introduce the decision aid. If we are able to determine decision aids reduce decisional conflict this would provide evidence to support increasing resources into the development of decision aids.