Article Text

Download PDFPDF
EP12 Using electronic patient record (EPR) routine data to inform decisions about which ambulance stations should participate in randomised controlled trials
  1. Joshua Miller,
  2. Jenny Lumley-Holmes
  1. West Midlands Ambulance Service University NHS Foundation Trust, UK


Background Timely recruitment of patients to randomised controlled trials (RCTs) is vital to delivering high-quality, cost-effective research, but RCTs often fail to meet recruitment targets, or meet them more slowly than planned. Ambulance services using electronic patient records (EPR) may be able to use this data to better inform decisions about which sites should participate in future RCTs.

Methods We used off-the-shelf business intelligence software to examine electronic patient records for all emergency and urgent cases within one English ambulance service for the 287 days 1st January 2020 to 13th October 2020. We adapted inclusion criteria from an upcoming RCT and applied it to this dataset: patient age; trauma mechanism; IV morphine administered. These potentially eligible cases were then displayed by ambulance station, an approximate daily rate per station calculated, and this rate applied to the RCT recruitment period of 304 days. This gave a potentially eligible patient population per station. Finally, we applied a patient conversion rate of 20% to estimate how many participants would be recruited from each station, basing this on recent experience of recruitment to another RCT. The RCT target is 111 patients per station.

Results 783186 records were reviewed, with 7667 of them potentially eligible for the study. When applied to the proposed study period, there is a potential population of 8121 across the 14 stations. Estimated recruitment per station varied (median 100, range 60-209) and with a 20% conversion rate, 9/14 stations would under-recruit.

Conclusions Using retrospective data from electronic patient records could be a simple, low-cost way to optimise patient recruitment. Other study criteria could be applied easily, as well as differing conversion rates. We identified ambulance stations where patient recruitment would not be likely to reach study targets, helping inform our decisions around stations as potential study settings.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.