Study blasts CMS rule to limit CT scans as unreliable, inaccurate

An imaging efficiency measure developed by the Centers for Medicare & Medicaid Services (CMS) to reduce CT scans in emergency departments (EDs) does not accurately determine which hospitals are performing CT scans inappropriately, according to a study published Feb. 23 online in Annals of Emergency Medicine.

The CMS created the measure, referred to as Outpatient Measure 15 (OP-15) in response to the increasing use of brain CT and the low diagnostic yield for the procedure in U.S. EDs. The measure uses Medicare billing data to determine the appropriateness of scans ordered for ED patients with a traumatic headache according to whether the patients had any clinical exclusions. The measure was not validated against chart review or endorsed by the National Quality Forum due to lack of supporting scientific evidence, according to Jeremiah Schuur, MD, of the department of emergency medicine at Brigham and Women’s Hospital in Boston, and colleagues.

To determine whether clinical appropriateness of imaging can be established from Medicare claims data, Schuur et al conducted a retrospective record review of patient visits from 2009 at 21 U.S. EDs. The authors identified 769 patient visits labeled by the CMS measure for including an inappropriate brain CT. Medical record reviews were then used to identify the reliability, validity and accuracy of OP-15 when measured against testing protocol from the American Medical Association’s Physician Consortium for Performance Improvement.

Of the total number of patient visits with an “inappropriate” brain CT, 748 had available records. Of those patients, 83 percent should not have been labeled as inappropriate based on clinical policy guidelines or expert consensus standards. Sixty-five percent of the CT scans labeled inappropriate actually complied with the CMS measure, and 18 percent had valid reasons for the CTs documented on their charts.

"The measure, OP-15, was only 17 percent accurate in assessing which patients should receive a CT scan," Schuur said in a statement. "Furthermore, hospitals' performance on the new measure as reported by the CMS did not match the proportion of CTs with a documented clinical indication. By using it, Medicare runs the risk of publicizing inaccurate information about clinical performance and rewarding hospitals based on unreliable data."

To underscore the findings, the authors included a collection of clinical vignettes that provided examples of patients whose brain CTs were incorrectly labeled inappropriate. One was of a 67-year-old man with a known aneurysm who was sent to the ED by his primary care physician because of an unusually severe headache lasting four days. Another case dealt with a middle-aged man with a gradually worsening headache who had a recent history of neurosurgical interventions. Despite brain CTs being justified in these cases, they were labeled as inappropriate under OP-15.

“Using OP-15 to report hospitals’ performance could discourage emergency providers from ordering CTs in older adults, a group for whom there is little evidence-based guidance. Such pressure could result in missed and delayed diagnoses in this high-risk group,” wrote the authors, who said policymakers should find alternative strategies to improve imaging efficiency.

The CMS plans to publish the data from measure OP-15 on its website, Hospital Compare.