There is considerable overlap between the data that clinical triallists need to assess the impact of interventions on patients and what is already routinely collected and stored in patients’ records from healthcare interactions. These routinely collected data are reflected in local, regional or national datasets, collectively labelled here as ‘healthcare systems data’ (HSD).
Timely and reliable access to HSD has the potential to transform the conduct of many clinical trials, making them considerably more efficient and cost-effective in a number of ways. These include identification and recruitment of trial participants, through to follow-up for outcomes and participant retention.
Some notable studies have successfully accessed and used HSD to identify potential recruits, support randomised trials, deliver interventions, and supplement or replace trial-specific follow-up.
Our work focusses on developing evidence for how useful and accessible such data is, by collating researchers’ experiences with its use, identifying key issues and barriers to using HSD, and working to solve them.
We work closely with colleagues in NHS DigiTrials and the Data Access Request Service teams of NHS England to improve data access to trialists, which includes assessing the integrity and utility of frequently accessed datasets.
This work is carried out in partnership with NHS England, together with Health Data Research (HDR) UK and the NIHR-MRC Trials Methodology Research Partnership Health Informatics Working Group.
We are involved in the Trials Stakeholder Prioritisation Forum in conjunction with HDRUK Alliance to bring together key stakeholders in a four-nation approach. It includes data providers and custodians; users of these data in the trials community; and users or interpreters of trial findings, including medicines regulators and NICE, as well as trial funders.
The forum aims to identify and address barriers to HSD use in trials, disseminate output from the HDRUK programme of work and set priorities for it.
We run data utility comparison studies, which investigate how well healthcare system datasets match up to data collected in clinical trials. HSD is usually collected for management of patients, and reimbursement to healthcare providers for the services provided, so not for research purposes.
The main aim of data utility comparison studies (DUCkS) is to understand if specific HSD is suitable to replace some or all of trial specific data collection, by comparing levels of agreement or concordance between the trial collected data (from case report forms) and HSD. Some outcome measures, like deaths, are already reliably collected. Others, like safety, need detailed work to meet regulatory reporting requirements as many healthcare systems datasets are not updated in real-time.
If you are using or planning to use healthcare systems data in your clinical trial, please contact Macey Murray to collaborate on a data utility comparison study. You will receive a free crocheted duck in return!
Trialists need to understand how data providers manage and produce the HSD they use in their trials. This is known as data provenance, the detailed record of the origins of the data, the processes, and the methods by which it is produced.
Understanding provenance enables trialists to ascertain the integrity of such data: the extent to which the data are complete, consistent, accurate, and reliable throughout the data lifecycle. Through a secondment at NHS DigiTrials, over two projects we documented and assessed four key datasets: Admitted Patient Care, Outpatients, and Critical Care data of Hospital Episodes Statistics, and the Civil Registration of Deaths (from the Office for National Statistics).
The first project documented two datasets in a published report, and the second used a data intelligence tool to ingest metadata of four datasets into NHS Digital’s Central Metastore. The outputs are publicly available in the DEDICaTe page.
We run training and knowledge transfer and exchange initiatives to help others make the most of HSD.
Listen to our Trials Talk podcast episode on how clinical trials can make the most of healthcare systems data, featuring Macey Murray and Matt Sydes:
Watch Macey Murray's British Heart Foundation Data Science Centre lunchtime webinar on assessing the provenance and integrity of healthcare systems data: