From the 2021 HVPAA National Conference
William Yang (University of Maryland Medical Center), Rosanne Rouf (University of Michigan), Venkat Gundareddy, Regina Kauffman, Amy Knight
Background
Electronic medical records’ (EMRs) user interfaces often present information in an inefficient manner leading to difficulty accessing relevant clinical information, frustration with usability, and adverse clinical events. Clinical information, such as labs and medications, are often displayed on different screens, rather than collated and correlated in intervention-centric or diagnosis-centric views which better facilitate complex clinical decision-making. Safe and effective inpatient diuresis of patients undergoing treatment for acute decompensated heart failure (ADHF) requires the synthesis of many different types of clinical information available in the EMR.
Objective
To create a clinician-centered EMR dashboard to facilitate the treatment of patients hospitalized with ADHF which is more usable and efficient than the standard methods of data review available in the EMR.
Methods
We surveyed and observed clinicians pre- and post-implementation of the EMR dashboard. Clinicians who treat ADHF patients and routinely place orders in the EMR were eligible for inclusion in the study. Participants subjectively rated their perception of the EMR using descriptive Likert scales in the context of ADHF treatment. We also observed participants reviewing information in the EMR prior to ordering diuretics for a patient, recording the elapsed time, mouse clicks required, and EMR screens accessed.
Results
Prior to dashboard implementation, 30 providers were observed treating 34 patients, while following implementation, 27 providers were observed treating 30 patients. On a 5-point descriptive Likert scale, providers reported a significant increase in both efficiency of information retrieval (3.88 vs 2.73, p=0.0034) and EMR satisfaction (4.24 vs 2.87, p=1.4×10-6) post-implementation of the EMR dashboard. Providers were observed to make 43% fewer clicks (8.9 vs 15.6, p=1.5×10-7) and spent 23% less time reviewing information per patient (75 vs 98 seconds, p=0.0026). The average number of discrete EMR activities accessed decreased from 4.6 to 2.1, with the greatest reductions in accessing the ‘Vitals’, ‘Weight’, and ‘Med History’ activities, while ‘Results Review’ and ‘I/O’ saw small decreases and remained relatively popular post-implementation.
Conclusions
A clinician-centered EMR dashboard for treating ADHF patients increased provider satisfaction and efficiency with the EMR.
Clinical Implications
There is unrealized potential in existing EMRs to better support clinical workflows. Intervention-centric and diagnosis-centric dashboards designed by clinicians using existing EMR functionality can improve EMR usability and have the potential to improve patient care.