PROspectiVe imaging research DEsign and coNducT (PROVIDENT): Considerations for clinical trials and studies using imaging (Part II)

OBJECTIVES: Imaging is used in a wide range of contexts in clinical research projects, but adds complexity to the design, conduct and analysis. This paper is the second of two in which we use a consensus approach to combine multidisciplinary perspectives on the challenges in conducting prospective clinical trials and other research studies involving imaging. Here we consider challenges in image interpretation and quantification, quality assurance and quality control (QA/QC); scanner imaging acquisition, data flow and storage, health economics (HE) decision modelling, costings for running a trial; and commercialisation. KEY FINDINGS: Availability of scanners and staff can impact deliverability. Pre-specification of key procedures, roles and responsibilities via appropriate documentation is important; ensuring compatibility across different sites and machines is challenging and requires advance input from multiple stakeholders. Testing critical procedures, including the flow of images and derived data between participating sites and/or external legal entities, can avoid delays. Effective QA/QC is conducted at regular intervals; relevant staff should be involved at the planning stage. Identifying appropriately qualified readers and arranging for image hosting takes time; this should be done prior to image acquisition. Testing image interpretation burden informs feasibility and costings. Cost estimates for research involving imaging and HE modelling of imaging interventions can be complex due to the interplay between local and national policies, and the extent to which the research imaging is integrated with standard care. CONCLUSION: These considerations derived from a multidisciplinary team will be useful for funding applications, protocol design, trial implementation, conduct and commercialisation and uptake of new imaging techniques. IMPLICATIONS FOR PRACTICE: Many prospective imaging studies could be improved by the upfront awareness of potential challenges and understanding of real-world examples these considerations provide.