An algorithm designed to identify patients with a risk of cancer in need of follow-up may help to reduce delays in the diagnostic evaluation of abnormal chest imaging results, according to a recent study.
Delayed cancer diagnoses are one of the most common reasons for ambulatory malpractice claims and can be associated with greater patient anxiety and worse overall outcomes. Despite the availability of new electronic health records (EHR) technologies and diagnostic tools, these delays have persisted at health care institutions around the United States, highlighting the need for better ways to receive, recognize, and process critical patient information.
-----
Related Content
Algorithm Improves Adverse Event Response for Patients Receiving Immunotherapies
Mock Clinical Pathways: A Method for Exploring the Oncology Clinical Pathway Development Process
-----
In a study published in Chest, researchers led by Daniel R Murphy, MD, MBA, Michael E DeBakey Veterans Affairs Medical Center, Houston, Texas, tested a “trigger” algorithm that scans EHR data for clinical and diagnostic clues of malignancy to see if it could help identify patients experiencing delays during care.
To identify patients they deemed “suspicious for malignancy,” researchers refined previously developed “red flag” criteria and excluded patients for whom further evaluation would be unnecessary, such as those who had a terminal illness, a completed biopsy, or a prior history of lung cancer. These data were then plugged into a computerized algorithm to identify trigger-positive (experiencing delays) and trigger-negative (patients with an abnormal imaging result, but no delays) patients. A panel of experts confirmed the presence or absence of delay in both groups.
The algorithm was applied to 208,633 patients seen between January 2012 and December 2012. Of the 40,218 chest-imaging tests performed in these patients, 1847 were deemed suspicious for malignancy, with 655 (35%) classified as trigger-positive. In a randomly selected population of 400 trigger-positive patients, reviewers confirmed delays in 158 patients (40%) and said that another 84 patients (12%) would require additional tracking. The review of 100 trigger-negative patients showed greater consistency with the algorithm; the panel only identified 3 patients (3%) who they felt were in fact experiencing delays.
The study confirmed that delays in the diagnostic evaluation of abnormal chest imaging results are in fact a common problem. While the results of the algorithm were not in agreement with a reviewer panel 100% of the time, researchers still concluded that the trigger algorithm could help identify patients experiencing delays in diagnostic evaluation of chest imaging results.
