Background: The electronic health record (EHR) and health care provider workflow process may contribute to patient misidentification or wrong-patient errors. When self-caught by the provider, these errors are classified as near-miss errors. When these errors reach the patient, they can result in serious harm. The Office of the National Coordinator for Health Information Technology Patient Identification SAFER Guide recommends displaying patient photographs in EHRs to help reduce wrong-patient errors. Preliminary data from our randomized controlled trial shows a significant reduction in wrong-patient orders among providers assigned to see a patient photograph (intervention) when placing electronic orders versus providers assigned to no photographs (control). To better understand why these errors continue to occur despite the presence of photographs in the EHR, we conducted real-time surveys of providers following these near-miss events.
Methods: This study is a secondary, qualitative analysis of a larger, randomized controlled trial at a major academic medical center. For the main study, patients who presented to the center underwent a routine protocol to have their photographs taken during registration. Providers were randomized in a 1:1 scheme to either see photographs (intervention) or not see photographs (control) when placing electronic orders. We employed the Wrong-Patient Retract-and-Reorder (RAR) measure, which is endorsed by the National Quality Forum as a measure of wrong-patient orders, the primary study outcome. The measure is an electronic query that detects RAR events, defined as one or more orders placed for an initial patient, then retracted shortly thereafter, and replaced by the same provider for a different patient. We developed a system that would run the query every 30 minutes and notify our research team of all the events that had occurred in the prior 30 minutes. In real time, we conducted brief interviews in a convenience sample of providers who made wrong-patient errors, despite the display of a patient photograph in the EHR. After obtaining verbal consent, providers were asked to describe the event, provide feedback about the intervention, and identify the root cause.
Results: In the majority of interviews conducted, providers had either not noticed the photograph prior to placing the order or had seen the photograph but had never met the patient prior to placing an order. Most providers also noted during the error they were rounding, working on multiple patients, and were interrupted. Limitations noted by providers are illustrated by the following quote: “I think sometimes in the photos it’s hard to see the patient, maybe it’s really zoomed out or the patient is wearing a hat, or there’s something distracting in the background. The photos could be larger as well and set on a white background…. A lot of it is about how you set up your chart, so if you can filter so it’s only on your patients vs. all of the patients in the ED that is helpful.”
Conclusions: Preliminary results from our interviews indicate that providers either did not see the photograph or did not know what the patient looked like when placing the initial order on the wrong patient. This suggests that improving the way photographs are employed in the EHR may help to reduce these wrong-patient orders further.