Macnair, Archibald James Cameron; (2022) Utilising electronic health records (EHR) to improve the efficiency of oncology clinical trials. Doctoral thesis (M.D(Res)), UCL (University College London).

Preview

Text Utilising electronic health records (EHR) to improve the efficiency of oncology clinical trials.pdf - Submitted Version Download (5MB) | Preview

Abstract

INTRODUCTION: Randomised controlled trials (RCTs) are typically considered the gold standard for investigating oncology treatments. However, RCTs can be resource intensive for healthcare organisations. Electronic health records (EHRs) comprising routinely collected data have the potential to improve the efficiency of clinical trials by supporting recruitment, follow-up and lowering costs. However, there is limited information about the reliability (accuracy and completeness) of EHR and how they can be used to support clinical trials in the UK. METHODS: The utility of EHR data was examined by: 1) comparing trial data relating to cardiovascular serious adverse events with data from two EHR databases using sensitivity and predictive value. 2) assessing the practicality and validity of data from a national cancer registry as a passive method of long-term follow-up in an established RCT and 3) whether data on aspirin use from a pre-existing RCT dataset, could provide further supporting evidence for the initiation of a new trial. RESULTS: Access to national EHRs was challenging with time from application to receiving data ranging from 8-15 months. Some cardiovascular inpatient events i.e. acute coronary syndrome matched well to trial data (sensitivity 0.89). However, both the trial data and EHR found separate events meaning poor comparability within these datasets. Long-term passive follow-up with EHR could be used for mortality data (sensitivity 100%) and potentially cancer recurrence if multiple cancer registry datasets were used. A large trial evaluating aspirin supported by EHR is feasible, but this initial study did not provide statistically supportive evidence for the use of aspirin in primary chemoprevention. CONCLUSION: EHR have the potential, in the future, to support clinical trials particularly for long-term follow-up. However there remain major hurdles to resolve prior to their widespread use especially in access to appropriate datasets and understanding the limitations of the data.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item