Prospective data linkage to facilitate COVID-19 trials – A call to action

Main Article Content

P Alison Paprica, Professor
https://orcid.org/0000-0001-6362-7087
Matthew Sydes, Professor
https://orcid.org/0000-0002-9323-1371
Kimberlyn McGrail, Professor
https://orcid.org/0000-0002-9349-1915
Andrew Morris, Professor
https://orcid.org/0000-0002-1766-0473
Michael J Schull, Professor
https://orcid.org/0000-0002-1318-6364
Rhoswyn Walker, Dr.
https://orcid.org/0000-0001-5605-2879

Abstract

Many countries around the world are creating COVID-19 trial datasets and databases of COVID-19-related data such as test results that cover entire populations. It is our strong recommendation, and sincere hope, that these data assets will be brought together through record linkage so that their scientific value and impact for society can be amplified. We call on trialists, data stewards and research funders to work together so that prospective linkage of trial data to medical databases becomes the norm, starting with COVID-19 trials.

The international research community is mobilising an unprecedented response to COVID-19, including more than 2500 registered clinical trials [1, 2]. There are also at least 100 observational studies based on administrative and routinely collected healthcare data [3, 4]. The small upfront effort to allow linkage of administrative data, registries, electronic medical records, and personal health records (collectively referred to as medical databases) to the trial data should create significant returns in scientific insight. However, to our knowledge, only a small number of COVID-19 trials are leveraging medical databases by linking to them.

The solution is simple and is a golden opportunity which should not be missed. As demonstrated by the West of Scotland Coronary Prevention Study (WOSCOPS) 20-year Follow-up Study [5], a major advantage of linkage is that administrative data can be used for follow up so that researchers can identify events that occur, including those that happen long after the study’s original time period for active assessment of participants has ended. More specifically, the WOSCOPS 20-year Follow-up Study found that statin treatment for five years in the original 1989 trial was associated with a legacy benefit, with improved survival and a substantial reduction in cardiovascular disease outcomes over a 20-year period in a high-risk cohort of men [6]. Further, the cost of follow-up assessment using medical databases was very low compared to the bespoke data collection of the original trial, costing on the order of tens of thousands rather than millions of dollars [7].

More recently, the UK RECOVERY trial of therapies for COVID-19 was the first to identify a survival benefit of dexamethasone in high-risk hospitalised patients with COVID-19. The study included linkage of trial-specific data and data from routine healthcare and registry sources. Using expedited access to existing medical databases, RECOVERY investigators were able rapidly to determine information on vital status (e.g., date and cause of death); discharge from hospital; intensive care use; and renal replacement therapy [8]. The RECOVERY trial’s speed and comprehensiveness deserve praise [9], both of which were enabled by linkage to medical databases.

Linkage to medical databases in trials can reduce cost (and thereby allow larger trials within an existing budget), increase speed (particularly if near real-time data flows can be achieved), support more complete long-term follow-up, and allow multiple outcomes to be monitored. In the case of COVID-19, it could also help researchers go beyond prospective analysis and forecasts based on COVID-19 cases, bringing in additional information about patient characteristics and past health system use. Combining clinical and health services research expertise will increase our understanding of the relevance of comorbidities, demographic factors and prior health system utilisation on the effects of COVID-19. As per the WOSCOPS 20-year Follow up Study example, linkage could also enable the assessment of the long-term impacts of interventions for COVID19, including experimental drugs and vaccines, years after clinical trials end, and at greatly reduced cost relative to bespoke data collection. Rare outcomes could be studied at scale and provide robust evidence regarding the safety of new vaccines, and thereby reassure regulators and the public that potential risks are carefully monitored [10].

There are, of course, caveats. Use of medical databases in trials requires full understanding of data quality limitations, excellent knowledge of database holdings, and case validation work [7]. It is also important that the public and trial participants support use of their data in trials [7, 11, 12]. Effective partnerships between trialists and medical database stewards will be essential, and Health Data Research UK (HDR UK) and Health Data Research Network Canada (HDRN Canada) have already begun to support this. For example, HDRN Canada organisations have developed standardised text that trialists can use when seeking participant informed consent for data linkage [13, 14], and the HDR UK Health Data Hub for clinical trials, NHS DigiTrials, delivered in partnership with NHS Digital, is providing services to improve the assessment of clinical trial feasibility, which has successfully contributed to the RECOVERY trial [15, 16].

Many countries around the world are creating COVID-19 trial datasets and databases of COVID-19-related data such as test results that cover entire populations. It is our strong recommendation, and sincere hope, that these data assets will be brought together through record linkage so that their scientific value and impact for society can be amplified. We call on trialists, data stewards and research funders to work together so that prospective linkage of trial data to medical databases becomes the norm, starting with COVID-19 trials.

Statement of Competing Interests

The authors declare that they have no conflcit of interest.

References

  1. ClinicalTrials.gov. Search Results: COVID [Internet]. 2020 [2020 July 15]. Available from: https://clinicaltrials.gov/ct2/results/details?cond=COVID

  2. World Health Organization. International Clinical Trials Registry Platform (ICTRP) [Internet]. 2020 [2020 July 15]. Available from: https://www.who.int/ictrp/en/

  3. Health Data Research Network Canada. COVID-19 Response [Internet]. 2020 [2020 July 15]. Available from: https://www.hdrn.ca/en/covid

  4. Health Data Research UK. HDR UK COVID-19 Strategy. London: HDR UK. 2020 Apr 4 [2020 July 15]. Available from: https://www.hdruk.ac.uk/covid-19/

  5. Barry SJE, Dinnett E, Kean S, Gaw A, Ford I. Are Routinely Collected NHS Administrative Records Suitable for Endpoint Identification in Clinical Trials? Evidence from the West of Scotland Coronary Prevention Study. PLOS ONE. 2013; 8(9): e75379 10.1371/journal.pone.0075379

    https://doi.org/10.1371/journal.pone.0075379
  6. Ford I, Murray H, McCowan C, Packard CJ. Long-term safety and efficacy of lowering low-density lipoprotein cholesterol with statin therapy: 20-year follow-up of West of Scotland Coronary Prevention Study. Circulation. 2016;133(11):1073-80 10.1161/CIRCULATIONAHA.115.019014

    https://doi.org/10.1161/CIRCULATIONAHA.115.019014
  7. Canadian Institutes of Health Research. Best Brains Exchange – Using Administrative, Electronic Medical Record and Patient Health Record Data in Clinical Trials. Ottawa, ON: Canadian Institutes of Health Research; 2016. Available from: https://www.ices.on.ca/~/media/Files/Data-and-Analytic-Services/2019/Medical-databases-and-trials-BBE-report-17May16.ashx?la=en-CA

  8. Horby P, Lim WS, Emberson J, Mafham M, Bell J, Linsell L, et al. Effect of Dexamethasone in Hospitalized Patients with COVID-19: Preliminary Report. medRxiv. 2020 and , preprint.10.1101/2020.06.22.20137273

    https://doi.org/10.1101/2020.06.22.20137273
  9. Wilkinson E. RECOVERY trial: the UK covid-19 study resetting expectations for clinical trials. BMJ. 2020; 369:m1626 10.1136/bmj.m1626

    https://doi.org/10.1136/bmj.m1626
  10. Liu EY, Smith LM, Ellis AK, Whitaker H, Law B, Kwong JC, et al. Quadrivalent human papillomavirus vaccination in girls and the risk of autoimmune disorders: the Ontario Grade 8 HPV Vaccine Cohort Study. CMAJ. 2018;190(21):E648-E655 10.1503/cmaj.170871

    https://doi.org/10.1503/cmaj.170871
  11. Paprica PA, Sutherland E, Smith A, Brudno M, Cartagena RG, Crichlow M, et al. Essential requirements for establishing and operating data trusts: practical guidance co-developed by representatives from fifteen Canadian organizations and initiatives, IJPDS, in press

  12. Aitken, M., Cunningham-Burley, S., Darlington, A., Elstub, S., Escobar, O., Jones, K.H. et al. Why the Public Need a Say in How Patient Data are Used for Covid-19 Responses. IJPDS. 2020;5:2:01 10.23889/ijpds.v5i2.1357

    https://doi.org/10.23889/ijpds.v5i2.1357
  13. Population Data BC. Data Linkage [Internet]. 2020 [2020 July 15]. Available from: https://www.popdata.bc.ca/datalinkage

  14. Institute for Clinical Evaluative Sciences. Linking Trial Data with ICES Data [Internet]. 2020 [2020 July 15]. Available from: https://www.ices.on.ca/DAS/Trial-Data

  15. Health Data Research UK. NHS DigiTrials - The Health Data Research Hub for Clinical Trials [Internet]. 2020 [2020 July 16]. Available from: https://www.hdruk.ac.uk/help-with-your-data/our-hubs-across-the-uk/nhs-digitrial/

  16. NHS. NHS DigiTrials [Internet]. 2020 [2020 July 16]. Available from: https://digital.nhs.uk/services/nhs-digitrials

Article Details

How to Cite
Paprica, P. A., Sydes, M., McGrail, K., Morris, A., Schull, M. J. and Walker, R. (2020) “Prospective data linkage to facilitate COVID-19 trials – A call to action”, International Journal of Population Data Science, 5(2). doi: 10.23889/ijpds.v5i2.1383.

Most read articles by the same author(s)

1 2 3 > >>