Sam Cross is a statistician working with the South London Applied Research Collaboration at King’s College London in the Institute of Psychiatry, Psychology & Neuroscience (IOPPN). After completing her bachelor’s degree in mathematics in the US, she attained an MSc in medical statistics from LSHTM. She worked as a clinical trial statistician after her master’s degree before transitioning to her current role at KCL. She currently provides statistical input and analysis for a variety of projects, including investigating the effects of the pandemic on mental health service use, a program to reduce stigma around mental illness, and of course the HARPdoc trial.
Sam discovered the field of implementation science after starting her role in the IOPPN in 2021 and has been enthusiastic to learn more ever since. She is interested in continuing to do work to help bridge the gap between clinical research and implementation.
Sam is a co-first author of this work along with Dr. Tayana Soukup. Tayana contributed to this trial in her role as a research fellow in the Centre for Implementation Science in the IOPPN. She is also an Assistant Professor at The State University of New York at Buffalo, Senior Research Fellow at Imperial College London, and a Senior Fellow in Multidisciplinary Teams at the University of Suffolk, UK.
An established blood glucose awareness training (BGAT) was compared to a novel Hypoglycaemia Awareness Restoration Programme despite optimised care (HARPdoc) in a randomised hybrid trial.1-2 While HARPdoc was not shown to be superior in reducing severe hypoglycaemia (SH) over 12 months, it was effective in reducing cognitive barriers to avoiding SH.3 We report a comparative analysis of the implementation of HARPdoc to BGAT, and explore whether self-reported implementation outcomes are associated with clinical outcomes.
This was an effectiveness-implementation hybrid type 2 trial (NCT02940873) occurring 2016-2021 in the UK and USA. Both BGAT and HARPdoc arms were rated for acceptability, appropriateness, and feasibility by the programme participants (n=45), their relatives (n=6), and healthcare providers (HCPs; n=27), totalling 48 assessments for HARPdoc and 41 for BGAT. Previously developed scales (AIM, IAM and FIM, respectively; 4 items each with a 5-point Likert scale)4 were used for these implementation assessments.
All study participants rated HARPdoc higher on acceptability, appropriateness, and feasibility. Total implementation scores were significantly higher for HARPdoc (M=3.67, SD=0.80) than BGAT (M=4.22, SD=0.79) for the programme participants (difference=0.86, 95%CI:0.37–1.34, p=0.001) and for all participants (difference=0.55, 95%CI:0.22-0.89, p=0.01). A statistically insignificant 35% decrease in SH events at 12 months with each point increase in implementation rating was estimated. Secondary outcomes were inversely associated with higher implementation ratings.
We found evidence that programme participants, their relatives, and HCPs find HARPdoc more implementable than BGAT. This warrants further investigation of the implementability of the two programmes within a larger sample.