UCLH clinicians and researchers are to test a novel way of running clinical trials using the hospital’s electronic health records system.
The new approach is designed to reduce research costs, answer clinical questions difficult to answer through traditional trials and enable many more studies to take place. The approach uses a computer prompt to ask a clinician making an everyday care decision, especially where there is no clear cut wrong or right answer, whether the patient should be randomised into a study or not.
Randomised controlled trials (RCTs) are the ideal way of evaluating treatments and interventions. But RCTs are expensive to carry out, and consent processes and randomisation of patients can be labour-intensive.
Clinicians collectively make hundreds of thousands of everyday treatment decisions based on observation and their own experience. There is often little evidence of what the right decision is and costly, complex RCTs may not be feasible to provide the evidence.
For example, magnesium is routinely given in intensive care to prevent abnormal heart rhythms, but clinicians vary in how they administer magnesium and optimal practice is unknown. A traditional RCT looking at dosing of magnesium could be done, but this would be expensive and time consuming due to the variation seen between patients.
For the PROSPECTOR-Critical Care study, led by Dr Matthew Wilson, researchers will assess the feasibility of using computer prompts in Epic to ask clinicians deciding on magnesium supplementation whether their patient could be randomised into a study looking at the effectiveness of liberal or restrictive approaches to magnesium supplementation.
The prompt will remove the need for manual patient randomisation and consenting – patients will have consented to take part in advance – and the clinician will be able to override the prompt if preferred.
When the clinician declines randomisation, the patient contributes data to an observational study which will still add to understanding of magnesium supplementation.
The research team, which includes BRC-supported Prof Folkert Asselbergs, Dr Steve Harris and Ruben Miguel, will evaluate two ways of prompting randomisation, and explore two models of patient consent. They will later seek clinician and patient feedback on each. Patients and the public have already inputted into the design of the study.
The study will shed light on how electronic health records systems may be used to deliver rigorous learning from routine clinical decision making. It will also collect pilot data to allow planning of a large-scale trial of magnesium supplementation strategies in the future.
Dr Wilson said: “Randomised controlled trials are the gold standard in terms of evaluating treatment approaches, but they can prove challenging to carry out in pragmatic situations.
“In future we hope it will be possible for many more trials to be embedded within electronic health record systems in the way we are looking at in this study – so that we can generate robust answers to many currently unanswered questions.
