New research on EHR safety performance finds link between usability and improved EHR safety

An article published this week in Jama Network Open, has highlighted the link between EHR frontline usability and EHR safety performance. The article, by Classen, et al., (2023), findings suggest that “current broad efforts to improve EHR usability may be associated with improvements in EHR safety performance”.

The new cross-sectional study was based on a sample of “the 127 hospitals that took the Leapfrog Hospital Survey, including the EHR/CPOE Evaluation Tool, in at least 1 year from 2017 to 2018 and also participated in the KLAS ARCH Collaborative (ie, collected survey data) in the same years as the Leapfrog Hospital Survey EHR/CPOE Evaluation Test”. Of the 127 hospitals in the initial sample, 15 were excluded due to “incomplete ARCH Survey Data”.

Authors “found that frontline EHR user experiences of usability were associated with the hospital safety performance of their EHR”. For Leapfrog safety measures primarily focusing on “a single critical part of the overall EHR use process—prescriber medication ordering”, there was a “significant association between the safety of the operational EHR and the experience that frontline clinicians have in using it”. This, they discussed, was “probably related in part to the frustration that they experience with medication ordering in poorly designed EHRs”, such as “too many alerts that would not only frustrate physicians but also cause them to ignore the alerts”.

The largest association between user experience was found to be for the ARCH survey areas of external integration, functionality, efficiency, time, and ease of learning, which the authors conclude is “not surprising because these areas reflect the users’ actual experience with the systems in terms of satisfaction, efficiency, and analytics ratings”.

Upon evaluating individual categories of the Leapfrog EHR/CPOE test, findings showed “certain categories of safety were associated with the overall user satisfaction score, especially drug-drug interactions, drug-allergy, therapeutic duplication, drug-age, drug dosing, and drug-route; other categories were not associated, such as drug-diagnosis, drug-laboratory, or drug-monitoring”.

Whilst authors “did not attempt to find direct causation between user satisfaction and safety performance of EHRs”, they did speculate that “this link may represent an association between how well a hospital implements and optimises their EHR for both user satisfaction and safety performance”.

Specific challenges noted in the study for EHR usability include that they are often approached as “separate silos”, and that “EHR vendor software is highly configured by local organisations before it is operationalised by users”.

The study’s authors concluded that they had found “a positive association between the safety performance of EHRs using an objective nationally endorsed standard and frontline users’ perceptions of EHR usability and experience”. This means “both health systems and vendors need to consider usability not only as critical for the frontline users, but also as a critical safety issue in the design, development, implementation, and maintenance of these complex EHR systems, and they should work together with frontline users and organisations to improve usability without compromising the integrity of safety performance”.

Last month, we looked at research published in Frontiers journal, which put forward a case study proposing an “omni-channel, outcomes-focused approach to scale digital health interventions in resource-limited populations”.

Another study from Sage Digital Health journal aimed to deliver a digital health intervention to improve health awareness for infection prevention, hygiene, and sanitation to assess its impact, and identify the risk of multimorbidity in women of reproductive years from low socio-economic background.

Classen, D.C., Longhurst, C.A., Davis, T., Milstein, J.A., and Bates, D.W. (2023) Inpatient EHR User Experience and Hospital EHR Safety Performance. Jana Network Open 6(9):e2333152. DOI:10.1001/jamanetworkopen.2023.33152