A qualitative study of how treatment recommendations are followed in clinical trials


In the past, qualitative trial research has mainly focused on whether patients and staff understand how trials work, or tried to understand their motivations for taking part. Rarely has research looked into why treatment recommendations are followed, or not, during the trial. In Trials, we present a study that looks to understand why success in achieving a particular glycaemic target was limited in the Treating to Target in Type 2 Diabetes Trial (4-T), in the hope that the findings could inform future trials.

We conducted in-depth interviews with 45 patients and 21 health professionals, including physician and research nurses, recruited from trial centres in the UK. Most patients were committed to taking insulin as recommended by 4-T staff, but to avoid hypoglycaemia, they occasionally altered or skipped insulin doses, normally in consultation with professionals. It became obvious that patients were usually unaware of the trial's glycaemic target; indeed, positive staff feedback sometimes lead patients to believe they had been 'successful' even when their glycaemic level exceeded the target.

While some staff felt that the system for automating insulin dosages used in the trial had increased their confidence to prescribe larger doses than usual, they all described situations where they had not followed its recommendations. In fact, they regarded the application of a 'one size fits all' glycaemic target during the trial as contradicting routine clinical practice where they would tailor treatments to individuals. They also expressed concerns that 'tight' glycaemic control might impose an unacceptably high risk of hypoglycaemia, thus compromising trust and safety, especially amongst older patients. To address these concerns, staff tended to adapt the trial protocol to align it with their clinical practices and experiences.

So, to understand trial findings, encourage attainment of endpoints, and promote protocol fidelity, it may be necessary to look beyond individual patient characteristics and experiences. We suggest that the context of trial delivery, the impact of staff involvement, and the difficulties staff may encounter in balancing 'clinical' and 'research' roles may need to be addressed.