Approximately 400 abdominal imaging studies are performed daily across Penn Medicine. For patients with indeterminate or suspicious lesions identified in abdominal imaging reports, a missed follow-up appointment can result in delayed cancer diagnosis or other serious implications.
The Department of Radiology at HUP implemented standardized categorization for focal abdominal mass reporting in July 2013. However, Penn Medicine does not currently have an automated system in place to monitor follow-up recommendations from providers and determine whether the patient has scheduled, completed or missed the appointment.
Implementing an automated system to monitor follow-up recommendations and patient action would increase the likelihood that patients complete clinically indicated follow-up instructions, and that radiologists receive feedback about lesions diagnosed on abdominal imaging.
We worked with Hanna Zafar, MD, MHS, and Tessa Cook, MD, PhD to build an informatics application that uses the lesion categorization system to identify all patients with indeterminate or suspicious lesions. The Automated Radiology Recommendation Tracking Engine (ARRTE) uses the Radiology Information System (RIS) to identify patients and requires radiologists to specify the modality and timing of follow-up for any indeterminate and suspicious lesions.
When fully implemented, ARRTE will mine data from the RIS daily to determine if recommended follow-up is scheduled, completed or neither scheduled nor completed within the specified time frame. If a patient does not have a scheduled or completed follow-up within the specified time frame, ARRTE will generate an automated electronic message to Penn Medicine providers or a flag to call providers outside of the system.