The most common diagnostic error in neurology over the past three decades is the misdiagnosis of headache, often the failure to diagnose brain tumor or other intracranial structural disease. This recurring error is partially attributable to the guidelines set forth by the American Academy of Neurology and six other professional societies under the United States Headache Consortium (USHC), which state that neuroimaging is not warranted in patients with migraine and a normal examination.
These guidelines were based on a meta-analysis of 11 small retrospective studies with serious flaws, all antedating 1985 and almost half using first generation CT scans. This outdated data suggested a very low incidence of intracranial abnormalities in patients with headache (0.2%), and led to the improper conclusion that a normal examination excludes any brain tumor or other intracranial disease. Both of these presumptions are flatly wrong.
For example, proper imaging demonstrates intracranial abnormalities in 6-8% of people, far higher than the outdated figure of 0.2%. Some of these abnormalities are incidental, but many warrant monitoring (aneurysm), further investigation (stroke) or treatment (arachnoid cyst). Additionally, over 10% of patients with brain tumors present with isolated headache and a normal examination, so following the guidelines would deny imaging to 3-7% of patients with brain tumor. Earlier imaging reduces the cost of headache care in patients that rank high on anxiety or depression scales.
Unfortunately, these outdated guidelines were parroted by the American Headache Society and the American College of Radiology which bluntly stated “Don’t do imaging for uncomplicated headache.”
This type of intransigent adherence to flawed, outdated guidelines falls below an acceptable standard of care. Additionally, the failure to discuss the likelihood of finding an abnormality on imaging, and giving the patient an option of having an MRI, eviscerates any meaningful informed consent.
Dr. James C. Johnston discussed these concerns in several recent articles including Neurologic Clinics 2016; 34:747-773, and recommended deleting the guidelines until further research properly correlates intracranial abnormalities with individual patient data, headache patterns, underlying disease, imaging protocols, associated conditions and a host of other factors to provide rational evidence based guidelines. (https://www.researchgate.net/publication/305409777_Neurological_Fallacies_Leading_to_Malpractice_A_Case_Studies_Approach).
In the meantime, he advised it is prudent to consider imaging all patients presenting with a new headache, or a headache increasing in frequency or pattern, or with any other warning sign.
Dr. Johnston presented these findings at the 2017 World Association for Medical Law conference, and the concerns were discussed by Dr. Mehila Zebenigus of Addis Ababa, Ethiopia who noted that in her clinic most patients with headache are referred for a baseline imaging study. Attendees from other countries echoed her recommendation. And, yet, in the United States about 12% of patients with headache have an MRI, and some neurologists are calling for more restrictions, even stating that limiting MRI should be a “major national priority.” It is past time for the USHC and Choosing Wisely societies to review the evidence and move forward.