Archiving a Complex Dataset:: IST-3 - A case study
The Third International Stroke Trial (IST-3) was a long running, large-scale international multi-centre randomised controlled trial of a thrombolytic drug, coordinated from Edinburgh, that recruited 3035 patients with stroke from 156 hospitals in 12 countries. Data on the patient’s demographics and clinical condition were collected at baseline, 7 days, 6 and 18 months. Digital copies of brain imaging studies (over 6000 separate brain imaging studies) were collated in a secure archive. The trial meta-data included paper and electronic records describing the trial methods, protocols, data collection forms and records relating to all the ethical regulatory approvals required to conduct the research. To meet the requirements of the funding agency and the regulatory authorities, the trial team had to create an effective, efficient and accessible archive, which could meet the requirements of several different types of users, and remain useable for at least 15 years. Those who might wish to access the archive would include (but not be limited to) drug regulatory agencies, research governance organisations, lawyers and scientists wishing to analyse the data under appropriate data sharing licences. The trial datasets therefore had to be understood as a whole so that any issues were identified and resolved. We anticipated that, after 15 years the originators of the archive may not be available to explain it, therefore documentation needed to be explicit, clear and concise. Given the nature of the personal and health information stored, strict confidentiality and security was paramount, and anonymity preserved.