MR376 - Ischaemic Stroke and Transient Ischaemic attacks (TIA)

TIA-Survive is a follow-up study to the TIA cohort study. The aim of TIA-Survive is to estimate the long-term survival of patients following a transient ischaemic attack (TIA), explore the main causes of death and identify prognostic factors associated with long-term survival. Civil Registry mortality data are required for individual participants to calculate the survival time in days from TIA occurrence to death and to identify the main causes of death and facilitate prognostic modelling.

The University of Oxford is processing data under the General Data Protection Regulation (GDPR) Articles 6(1)(e) ’task in the public interest’ and 9(2) (j) ‘research’.

Article 6(1)(e): The sharing of information will allow estimation of survival time following a TIA. Decisions based on the results of the study are likely to benefit society by providing GPs and medical professionals the information to inform patients following a TIA of their likely prognosis; and by identifying risk factors, medical staff have potential to treat those at high risk, with potential to prolong the lifespan and quality of life of patients.

Article 9(2)(j): Processing is necessary for statistical purposes to ensure accurate estimation of survival can be calculated to enable research into the long-term prognosis of persons following a TIA. The research will be carried out by an educational establishment in the area of primary care health research, to ensure high standards of quality and safety of health care. By sharing mortality data with the University of Oxford, the proposed analyses, researching the relationship between TIA and long term survival and likely cause of death, will identify causes that may be treatable thus improving survival and impacting on the quality and safety of health care.

Transient ischaemic attack, or TIA, occurs when there is interruption to the blood supply to the brain. Patients can have slurred speech, weakness on one side and problems with vision but symptoms resolve fully within 24 hours. The long-term outcome for patients following a TIA is not fully understood.

A total of 290 participants were recruited (between 1981 and 1986) to the TIA cohort study from Oxford Community Stroke project and a hospital based study. The TIA cohort study reported on short-term survival of persons following a TIA.

The TIA cohort team is based at the Nuffield Department of Primary Care Health Sciences at the University of Oxford and will investigate the long-term follow-up of persons following a TIA (TIA-Survive study) – this is a follow-up study, linking up with results found earlier in the TIA cohort short term follow-up analysis.

TIA-Survive will link the NHS Digital’s civil registration death data to the TIA cohort study data, using a pseudonymised study ID, to calculate survival time from diagnosis to death. Researchers will report the survival rates, mortality rates and cause of death, for the whole cohort in aggregated form, following ONS recommendations on small sample sizes avoiding potential identification of individuals. This work will determine how long people live for following a diagnosis of TIA, and whether they die from stroke-related problems or something else, to help doctors and patients understand more about the condition. The original short-term follow-up found a higher risk of death from myocardial infarction and coronary heart disease than expected in the general population and recommended preventative treatment for patients following a TIA. Research into the longer term outcomes by estimating mortality rates from different causes is now warranted to further benefit treatment of patients. This long-term survival analysis has not been undertaken yet.

To address the common law duty of confidentially, the University of Oxford have destroyed all fields that NHS Digital have classed as identifiable. This includes identifying details in both the data that NHS Digital (including any predecessor organisations) has provided the University of Oxford and fields that the University of Oxford have obtained from other sources. Only pseudonymised data is now held at the University of Oxford. Researchers do not require further data but wish to request permission to continue the processing of these data.

NHS Digital, as a trusted party, will retain the cohort identifiers as provided under historic Agreements. Under previous versions of this Agreement, Mortality data (including identifying data and date and cause of death) were disseminated to the University of Oxford linked via a pseudonymised study ID.

NHS Digital has supplied The University of Oxford with MRIS Cause of death, Members and Postings and Event notification reports from 1991 to 2019. The earliest reports included backdated data to the start of the recruitment period in 1981. Researchers use the Cause of Death report to estimate the death rates by causes of death in aggregated form, ensuring the tabulation of results will not include small counts that may risk disclosure of participants’ identity. The Members and Postings reports have been used to track death data received for whole cohort. Event notification reports provides the dates and causes of death to allow researchers to calculate survival time and mortality rates by cause for the TIA cohort participants.

Researchers have previously requested data from 1991 to 2018 to enable them to calculate long term survival and mortality rates and also to allow them to identify risk factors associated with the prognosis of persons following a TIA. Researchers require access to all of these years in order to increase the statistical power of the survival analysis performed and allow them to estimate mortality rates at 5 -10 yearly intervals post diagnosis.

The University of Oxford is the data controller and also processes the data for this study. No other organisations process the data for this purpose.

NHS Digital supplied Event and mortality data with identifying data to the University of Oxford via their secure portal. All identifying information supplied by NHS Digital and in the original participant information has since been destroyed mitigating the risk of re-identification.

The pseudonynised data are stored on an encrypted network at the University of Oxford and are merged with the TIA cohort data using a pseudonymised ID as previously described to allow calculation of survival times.

There will be no attempt to re-identify individuals.

Data processing is only carried out by substantive employees of the University of Oxford who have been appropriately trained in data protection and confidentiality. The data are protected by access control and encryption. Data are accessed both locally and remotely using University of Oxford encrypted PCs and laptops. The combined dataset is not copied or transferred from the University of Oxford network. Data will be held securely in accordance with the University’s policies and procedures. Further information is available on the University’s Information Security website: https://www.infosec.ox.ac.uk

The University of Oxford will use the linked civil registration mortality data to report the long-term survival of participants in the cohort. This analysis has commenced but not been completed to date. It should take a further 3 to 6 months to complete. The initial findings from the study will be presented in an abstract for submission to the Society for Academic Primary Care Annual Scientific Meeting in July 2020. The University of Oxford team will also write up the findings into a paper and submit to a peer-reviewed journal for publication such as the European Heart Journal or Stroke. The team hope to have successfully published the work by September 2020. The target audience will be cardiologists, stroke physicians and family doctors caring for patients with TIA and policymakers responsible for planning TIA and stroke services. The team will also work with their patient and public involvement (PPI) representatives to disseminate the findings to the public through PPI newsletters, charity websites and social media when the paper is published.

All output data have been aggregated with small numbers supressed in line with the HES analysis guide.

Following the original study, a paper was published in the Journal of Neurology, Neurosurgery and Psychiatry in 2003 titled 'Long term risks of stroke, myocardial infarction, and vascular death in "low risk" patients with a non-recent transient ischaemic attack'. This is the only output so far from this cohort which is why it is important to report the mortality figures.

Linking mortality data will allow the scientific community to understand more about the impact of TIA on longevity, and the reasons why people died. This type of epidemiological research is important for public understanding of TIA and how things may have improved over time. Prognostic research is important for patients to understand survival times following a vascular event, to clinicians looking after them to enable evidence-based discussions about treatment and outlook and to NHS commissioners to ensure there are appropriate healthcare services for people with TIA.

The updated linking mortality data will allow the University of Oxford to report the survival of the entire cohort. The participants were recruited in the 1980s and well-phenotyped. Outcome data from this period is lacking. Accurate survival rates following a TIA diagnosis, and cause of death, will improve the University of Oxford’s understanding of the pathophysiology of this condition.

The only publication to date from the study was published in 2003. The paper was published in the Journal of Neurology, Neurosurgery and Psychiatry and titled 'Long term risks of stroke, myocardial infarction, and vascular death in "low risk" patients with a non-recent transient ischaemic attack'. The paper has been downloaded a total of 3,926 times and has been cited 177 times. This work has added to understanding of the long-term cardiovascular risks in people with a history of TIA. Since 2003, the national stroke strategy, published by the Department of Health in 2007, has aimed to improve outcomes for people with stroke by raising awareness of symptoms and includes earlier diagnosis and management of people with TIA. No further analysis has been completed to date.