CONCLUSIONS
It is widely recognised that humans in the workplace create safety far more than they erode it. Human capacity for recognising problems and adjusting behaviours and actions in the moment will almost always prevent an accident rather than cause one. This premise was strongly supported in our observations of work processes in OxCRF. We observed the type of attributes and behaviours that support a strong safety culture in both the leadership teams and staff working in the centre including: encouraging and valuing diversity of opinion; a constructive dialogue about risk and an acceptance that just because processes are running smoothly in the moment, they may not do so reliably in future. It is this pro-active approach to safety that gave rise to this study in the first place.
CRFs are operated by NHS Trusts, pharmaceutical companies, contract research organisations or academic institutions, routinely staffed by a core team of healthcare professionals supplemented by trial-specific staff and charged with the delivery of multiple externally-generated protocols, often concurrently. This environment, especially during periods of high activity where IMP or interventions with divergent risk profiles are being evaluated, presents unique challenges where risk is concerned, and it is therefore vitally important to have robust safety frameworks in place that can apply across studies. Whilst the tool traditionally perceived to guarantee this is adherence to guidelines and regulations (with accompanying documentation), there is real danger these distract from core, often common sense, measures that involve consulting with the correct stakeholders with the relevant training, experience and local knowledge to instigate proportionate and focused measures to mitigate risk to participants
This is the first time, to our knowledge, that human factors methods have been explicitly used to analyse work systems in a CRF and protocol elements of an experimental medicine study to provide recommendations that improve the safety of clinical research. Our findings support the further investigation and validation of their value in this context with a view to routine implementation, not just in retrospect to the investigation of safety incidents, but proactively to help avert them
Acknowledgements: Author contributions – J.F conceived the study. H.H and L.M designed and performed the research and analyzed the data. H.H, J.F and L.M drafted the manuscript which was edited by C.C, H.McS and D.R and reviewed by all authors. JM, AM and SJ were Trial Clinicians for the COV-CHIM01 study and took part in the research. RLR was the Research Matron overseeing all nursing aspects of the COV-CHIM01 study and EH was the Research Nurse.
The Climax donation to the University of Oxford funded the reported work. The Wellcome Trust funded the COV-CHIM01 study for which the University of Oxford acted as sponsor.
The authors would like to acknowledge the contribution of the full COV-CHIM study team, all staff members at the OxCRF and the technical and administrative team in OxSTaR. We would also like to thank all the individuals who volunteered to participate in the COV-CHIM01 study, especially those who agreed to be observed as part of this nested pilot project.
Conflict of interest statement : RLR and SJ have previously contributed to intellectual property licensed by Oxford University Innovation to AstraZeneca. All authors declared no competing interests for this work
Funding Information: The Climax donation to the University of Oxford funded the reported work. The Wellcome Trust funded the COV-CHIM01 study (Wellcome Trust reference:  222305/Z/21/Z) for which the University of Oxford acted as sponsor.
Data availability statement: The data that support the findings of this study are available from the corresponding author upon reasonable request.