22. European Stroke Conference 562 © 2013 S. Karger AG, Basel Scientific Programme Patients with scan timing issues Patients with multiple scan episodes or a single scan ep-isode (exclud-ing patients with standard two scan epi-sodes & event scans) Patients with Scan Trans-fer Form re-lated issues Patients with oth-er scan info issues Patients with imaging issues Patients with Scan Transfer issues Num-ber of pati-ents 4 (R scan > 6h) 37 (R scan < 6h) 249 (P scan > 48h) 216 (>3 scan episodes) 40 (1 scan epi-sode) 24 (Date/Time issues) 33 (Problem-atic STFs) 4 (Scan date/time issues) 7 (Wrong scan label-ling) 5 (DOB discrepan-cies) 18 (Incomplete scans) 6 (Low quali-ty images) 23 (CT scan re-lated) 18 (MR scan related) 28 (JPEGs/Bit-map) 19 (Other) 18 (CD) 15 (FTP/Web) Per-cent-age of total num-ber of pa-tients (%) 0.13 (R scan > 6h) 1.21 (R scan < 6h) 8.20 (P scan > 48h) 7.11 (>3 scan episodes) 1.31 (1 scan epi-sode) 0.8 (Date/ Time issues) 1.1 (Problem-atic STFs) 0.13 (Scan date/time issues) 0.23 (Wrong scan label-ling) 0.16 (DOB di-screpan-cies) 0.59 (Incom-plete scans) 0.19 (Low quality images) 0.75 (CT scan related) 0.59 ( MR scan related) 0.92 (JPEGs/ Bitmap) 0.62 (Other) 0.59 (CD) 0.49 (FTP/ Web) 24 Large clinical trials (RCTs) Medical image management for multicentre trials. Experience from the Third International Stroke Trial (IST-3) with 6576 scans E. Sakka1, D. Perry2, D. Buchanan3, K. Innes4, J. Drever5, P. Sandercock6, R.L. Lindley7, J.M. Wardlaw8 Edinburgh University, Edinburgh, UNITED KINGDOM1, Edinburgh University, Edinburgh, UNITED KINGDOM2, Edinburgh University, Edinburgh, UNITED KINGDOM3, Edinburgh Uni-versity, Edinburgh, 4, Edinburgh University, Edinburgh, UNITED KINGDOM5, Edinburgh Univer-sity, Edinburgh, UNITED KINGDOM6, Edinburgh University, Edinburgh, UNITED KINGDOM7, Edinburgh University, Edinburgh, UNITED KINGDOM8 Background: Medical imaging is a powerful tool for clinical trials. We aimed to highlight the image management challenges for multicentre trials based on our experience. Methods: We evaluated the process of image management in the Third International Stroke Trial (IST3) of i.v alteplase versus control in ischaemic stroke. IST3 recruited 3035 patients from 135 centres in 12 countries from 2000 to 2012. Patients had pre- (R) and post-randomisation (P) brain imaging at 24-48hr; and a further scan in case of adverse neurological events. A plain CT or MR brain scan was required as a simple scan acquisition protocol was preferred. The Scan Housekeeping System registered received scans and tracked progress of queries on individual scans. Where possi-ble, the Imaging Manager attempted to resolve imaging issues without contacting the investigator. We evaluated the scan management processes, by analysis of the data integrity reports in the scan dataset. Results: In total there were 6576 scans (2.16 scanning episodes/patient on average), most (approx-imately 70% sent on CD) for which there were 764 queries (0.25 queries/patient on average). Scan timing produced most queries (commonest: P scan at >48hrs; 8.2%). 7.1% of patients had >2 scan episodes that were not clearly linked to an adverse neurological event, making it difficult to label them correctly. The Scan Transfer Forms were incomplete in 1.1% of queries; scan mislabelling oc-curred in <1% of queries. Receiving non-DICOM images was the main issue for imaging (approxi-mately 1%) whilst CD errors was the one for Scan Transfer methods (0.6%). On a random sample of 38 queries sent, the average time taken to resolve a query was 11.8 days. Conclusions: Imaging data management and quality control in multicentre trials is complex and time consuming. A well designed and adequately resourced image management system is essential to en-sure high quality scan data and reliable clinical trial results.
Karger_ESC London_2013
To see the actual publication please follow the link above