Why participants in The United Kingdom Rotator Cuff Tear (UKUFF) trial did not remain in their allocated treatment arm: a qualitative study.

OBJECTIVE: The UKUFF trial was a three-way parallel group randomised trial comparing surgical and non-surgical treatments for people with rotator cuff tears of their shoulder. High crossover between arms in the UKUFF led to the original trial design being reconfigured; 'Rest then Exercise' was halted. This study explored why participants recruited did not remain within allocated treatment arms and explored crossover and surgical decision making.

DESIGN: A qualitative phenomenological approach.

PARTICIPANTS: Purposive sampling (n=18) included participants randomised to 'Rest then Exercise' arm considered least likely to proceed to surgery but who had surgery, plus participants from all arms not having surgery.

METHODS: In-depth, semi-structured interviews were recorded and transcribed. Field-notes, memos, member-checking and a reflexive diary were used.

DATA ANALYSES: In accordance with Interpretative Phenomenological Analysis. Peer review, code-recode audits and constant comparison occurred throughout.

RESULTS: 1. Impact of symptoms and diagnosis: these influenced crossover; long durations of severe pain and failed conservative treatment increased eagerness for surgery. 2. Perceptions and expectations of treatment: surgery provided hope for participants, especially when "Rest then Exercise" was perceived as having previously failed. Surgeons were perceived to believe "tears need repairing". 3. Professionals know best: autonomy and communication: patients deciding not to have surgery had to actively leave the surgical waiting list. Increasing age, carer role, self-employment, co-morbidity and improving symptoms were reasons described for declining surgery.

CONCLUSIONS: Most participants had failed conservative treatment before trial entry and described strong preferences regarding treatment. Trials should demonstrate patient and clinician equipoise but participants' rarely described equipoise. If conservative treatments are usually provided sequentially in clinical practice, it may be inappropriate to include them as comparators in surgical trials. This is a qualitative study and not eligible for trial registration since it was carried out independently of the UKUFF trial (UKUFF ISRCTN97804283 Date assigned 29/06/2007).