EHR integration of radiology alerts speeds up—but doesn’t significantly change rate of—follow-ups

Integrating nonurgent but clinically significant radiology alerts into the EHR offers a workable way to manage follow-up patient care—and the integration doesn’t impede closed-loop communication between radiologists and referring physicians.

That’s the main takeaway from a study at Brigham and Women’s Hospital in Boston, the results of which are running in the Journal of the American Medical Informatics Association.

The study cohort consisted of 171 primary care physicians (PCPs) who received 5,931 Level 3 (nonurgent, clinically significant) radiology alerts via both EHR and an Alert Notification of Critical Results (ANCR) system.

The doctors acknowledged the bulk of the alerts via ANCR, with 16 percent using the EHR for the communication. Meanwhile, they acknowledged alerts much more quickly when using ANCR rather than the EHR (median 36 minutes for ANCR and median seven hours for EHR).

However, follow-up rates were acceptable or better regardless of acknowledgement vehicle: Over a 24-month postintervention period, PCPs acted on 97 percent of actionable alerts acknowledged in ANCR and on 79 percent acknowledged in the EHR.

Led by Stacy O’Connor, MD, the authors cite several factors that may explain the relative parity on follow-ups.    

For one thing, “although PCPs have established workflows that can take advantage of ANCR-EHR integration (e.g., reviewing test results in the EHR-based results management application while seeing patients in clinic) and may prefer notification in the EHR over email notification, this is not a consistent preference,” they write.

Secondly, ANCR relies on “active” external notifications—a heads-up goes to the referring provider’s pager or network email inbox—as opposed to “passive” internal EHR notifications, whereby providers have to log into the EHR to see and acknowledge alerts.

“Providers may not accept a system that requires them to actively seek test results, especially if that system does not fit with their workflow,” the authors note, adding that their findings support the hunch that results-management applications work best when designed to support enterprise-wide policies.