Medicare's Efficiency Measure for Head CT for Atraumatic Headache Is Profoundly Flawed
A multicenter review of medical records shows that the measure is wildly unreliable, invalid, and inaccurate.
As part of their initiative to publically report and eventually pair reimbursement with specific quality and efficiency measures, the Centers for Medicare and Medicaid Services (CMS) developed an efficiency measure to evaluate use of brain computed tomography (CT) for emergency department (ED) patients who present with atraumatic headache. The measure (CMS OP-15) uses administrative billing data for patients with a final diagnosis of nonspecific headache. A CT scan is not included in the calculation (i.e., considered appropriate) if the patient has any of the following exclusions, which were derived based on guideline review and expert opinion: headache associated with lumbar puncture, dizziness, paresthesia, lack of coordination, subarachnoid hemorrhage, complicated or thunderclap headache, focal neurologic deficit, pregnancy, HIV, tumor or mass, or CT scan related to reason for admission. The National Quality Forum (NQF), which typically reviews and approves all measures for the CMS, rejected the proposed measure because it lacked scientific validity. The CMS implemented the measure despite the NQF finding.
These authors assessed the reliability, validity, and accuracy of the measure in a retrospective review of medical records for a convenience sample of 748 patients deemed by the CMS as having received inappropriate head CT because no exclusions were documented. The authors determined that exclusions were documented in the medical record for 489 patients (reliability, 35%) and that universally accepted indications for head CT (according to expert consensus, society guidelines, and literature) were present for 123 of 259 patients without exclusions documented in the medical record (validity, 48%). Overall accuracy of the measure was 17%.
Comment: CMS OP-15 should be immediately discontinued because it is fatally flawed: it was not validated with chart review and was implemented against the recommendation of the National Quality Forum, thus discrediting both organizations. Implementation of CMS OP-15 is unjustifiable and inexcusable and represents a complete abandonment of patient safety and quality of care in favor of thoughtless reduction of reimbursement for CT scans, which, according to national guidelines, are indicated and appropriate. In addition, implementation of this ill-conceived measure raises concerns about the ability of the CMS to police itself or anything else.
Dr. Zane and Dr. Walls were both members of Dr. Schuur's department at the time of this multicenter study, but did not participate in the study in any way.
Published in Journal Watch Emergency Medicine September 21, 2012
Schuur JD et al. Assessment of Medicare's imaging efficiency measure for emergency department patients with atraumatic headache. Ann Emerg Med 2012 Sep; 60:280.
- Medline abstract (Free)
From time to time, a study is published that is of such importance to physicians that it warrants special attention and should lead to an immediate reappraisal of a current practice or to the adoption of a new approach. When such articles are identified by the editorial boards of the Journal Watch publications, they will receive the designation "Landmark Article." We encourage our readers to obtain the full text of these articles and review them carefully.
- Appropriate head ct
s p, 24 Sep 2012 9:41 AM EST
Specialty: Internal Medicine
This is going to be controversial but the CMS OP15 is actually addressing a need to reduce unnecessary testing. I... [more]
- CT in headache
J Brown, 3 Jan 2013 12:33 PM EST
Specialty: General Practice
I see no raw data on which to draw any conclusion. The only conclusion I can draw from the information... [more]
- headache in ED
ROBERT A. MCWILLIAM, NORTH WEST AUSTRALIA, 9 Jan 2013 10:57 PM EST
Specialty: Emergency Medicine
In my experience the most common atraumatic headache in ED is migraine,and CT for migraine is unnecessary
To ensure that your Reader Remark is not formatted as one long paragraph, precede new paragraphs with either a blank line or an indentation.