Clinical Context

Headache is one of the most common pain disorders seen in outpatient practice, and the authors of the current study provide a brief overview of the epidemiology and larger consequences of headache. Nearly everyone experiences headache at some point, and the prevalence of recurrent, severe headaches approaches 25%.

The high prevalence of headache is associated with substantial societal costs. There are approximately 12 million clinician office visits for headache per year in the United States, and headaches are estimated to cost the US economy $31 billion per year. The evaluation of headache accounts for some of this financial cost, but it can also have severe medical consequences. One study estimated that 4000 additional cancers were promoted by the 18 million computed tomography (CT) scans of the head performed in the United States in 2007. Another study found that the majority of all CT imaging of the head and brain were inappropriate, based on current recommendations. Most inappropriate CT imaging was ordered for headache.

The inappropriate use of resources in the management of headache has led to recommendations to reduce the use of brain imaging and physician referral, as well as limited use of opioids and barbiturates in the treatment of headache. But are clinicians following these recommendations? The current study by Mafi and colleagues explores this issue.

Study Synopsis and Perspective

Contrary to practice guidelines, clinicians treating patients with headache are increasingly ordering costly imaging tests and referring patients to other physicians, and they are doing less counseling on lifestyle changes, authors of a new review suggest.

The researchers found an almost doubling of the use of CT and magnetic resonance imaging (MRI) in a recent 10-year period.

Although given the nature of the study they could not determine which referrals or imaging studies were not appropriate, the trend toward a doubling of these tests is concerning, said lead study author John N. Mafi, MD, fellow, general internal medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts.

"We have no reason to suspect that headache is a disease that epidemiologically or pathophysiologically has changed over the past decade, so this, we think, is inappropriate changes in physician practice patterns."

The study was published online January 8 in the Journal of General Internal Medicine.

Inappropriate Changes?

Researchers used data from the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS) from January 1, 1999, to December 31, 2010. Both surveys obtain nationally representative samples of ambulatory patient visits across the United States.

From these surveys, researchers identified ambulatory visits (excluding emergency department visits) with a chief complaint and/or a primary diagnosis of headache. They also included those with secondary complaints and diagnoses of headache but unrelated primary reasons for the visit (eg, hyperlipidemias).

In the group of patient visits with a chief complaint of headache (80.8%), the surveys provide duration or context of symptoms in 5 categories: new onset (<3 months), acute-on-chronic flare-up, chronic routine, routine/preventive, and preoperative or postoperative visit.

Researchers studied 4 types of measures: use of advanced imaging, including CT and MRI; referrals to other physicians; counseling on diet, nutrition, exercise, mental health, and stress management; and use of medications.

The analysis included 9362 visits related to headache, which represented approximately 144 million visits during the study period. Almost 75% of the patients were women, and their mean age remained the same during the study, at approximately 46 years.

The study showed that advanced imaging, including CT or MRI, increased from 6.7% of visits in 1999-2000 to 13.9% in 2009-2010 (P < .001), as did referrals to other physicians, which rose from 6.9% to 13.2% (P = .005).

Although rare, there are "real concerns" about kidney injury due to use of contrast dye and radiation during imaging tests, commented Dr Mafi.

The study showed that during the same study period, counseling for headache prevention declined from 23.5% to 18.5% (P = .041).

But there were also somewhat encouraging results. For example, use of opioids and barbiturates, which are not recommended for headache, remained the same at approximately 18%.

"That's still quite a high number for a discouraged medication, so there's lots of room for improvement," commented Dr. Mafi.

Also, preventive medication for migraine, such as verapamil, amitriptyline, propranolol, and topiramate, nearly doubled — from 8.5% to 15.9% (P = .001). Abortive therapies, such as triptans and ergot alkaloids, rose from 9.8% to 15.4% (P = .022).

Use of nonsteroidal anti-inflammatory drugs and acetaminophen remained stable at roughly 16%.

The trends remained after adjustment for age, gender, race, geographic region, insurance status, symptom duration, urban location, and whether the physician was identified as the primary care practitioner. As well, there were similar trends between visits related to migraine vs nonmigraine headache, although patients with migraine received triptans/ergot alkaloids and preventive therapies more often, and their use of opioid or barbiturates was higher.

Use of CT/MRI rose more rapidly for those with nonmigraine headache, as did referrals to other physicians.

Acute vs Chronic

As for acute vs chronic headache, again the trends were similar. Use of CT/MRI appeared to rise more rapidly among patients with acute symptoms, although this difference did not achieve significance.

In addition, patients with chronic headache had lower adjusted odds of receiving referrals to other physicians (odds ratio [OR], 0.59; 95% confidence interval [CI], 0.42 - 0.82) or undergoing advanced imaging (OR, 0.47; 95% CI, 0.35 - 0.63) compared with other presentations. However, there were no significant differences across time as trends in referrals and imaging still nearly doubled for both groups.

Numerous factors are driving these trends, including a growing number of assertive and demanding patients, the authors suggest. "They come in and say 'I need an MRI for my headache' as opposed to 'I have a headache'," said Dr Mafi, adding that MRI and CT are more available now than ever before.

As well, physicians have an incentive to make patients happy because they are increasingly being rated on satisfaction surveys, said Dr Mafi.

There are also financial incentives for physicians to do more testing and concerns about legal liability if tests are not ordered.

Time constraints are another contributing factor. "There is simply not enough time during the visit to counsel" patients on making lifestyle modifications to help treat their headache, said Dr Mafi.

Reversing the trends will require a change in approach to reimbursement that relies less on the physical clinician visit and more on new technologies, said Dr Mafi.

"Rather than lengthen the visit or pay doctors more, I think we need to move away from that 20th-century mindset and really think about reimbursing care across a continuum of time," he said. "We need to think about reimbursing high-quality care at any time, whether it's electronic or through a secure email message between the patient and physician, or an electronic visit or telemedicine visit where patients can contribute to their records online and have the doctor provide feedback."

Appropriateness Unclear

For a comment, Medscape Medical News reached out to Elizabeth Loder, MD, chief, Division of Headache and Pain, Brigham and Women's Hospital, Boston, Massachusetts, and immediate past president, American Headache Society (AHS), who headed the AHS "Choosing Wisely" project that looked at opioid prescribing.

The new study was well done, has "some fascinating findings," and is important in light of the fact that headache treatment and management trends are a neglected medical problem, said Dr Loder.

However, she raised several issues about interpretation. For one thing, she pointed out that the NAMCS and NHAMCS do not provide information that would help determine whether imaging studies were appropriately or inappropriately ordered.

"In clinical practice, we see both underuse and overuse of imaging studies. Patients with complex, refractory chronic headache problems frequently have multiple — usually unnecessary — imaging studies. In contrast, it remains common for us to see in a headache clinic patients with worrisome presentations of headache who have not been imaged."

The data suggest that much of the increase in imaging may be due to an increase in acute as opposed to chronic headache, said Dr Loder. "This is exactly the situation where imaging is most likely to be appropriate."

As well, the results show that imaging is more common in patients with nonmigraine headache rather than migraine headache. "Again, this is a situation where imaging may well be appropriate."

Dr Loder stressed that when the AHS formulated its "Choosing Wisely" recommendations, the committee felt that evidence to discourage imaging was strong only for patients with stable headache who met criteria for migraine. According to the recommendations, imaging in patients without migraine is not necessarily inappropriate.

"The proportion of scan abnormalities is higher in patients with nonmigraine headaches and we did not feel evidence was sufficient to discourage imaging in those cases."

Dr Loder also noted that the increase in imaging and referrals has occurred alongside an "enormous downward pressure" on the time physicians can devote to patient concerns during an office visit. Physicians have to spend more time on paperwork, electronic medical record "meaningful use" requirements, and other things, she said.

"Ordering tests and making referrals are both relatively quick ways to demonstrate concern for a patient's symptoms and maintain patient satisfaction."

She pointed out that although some guidelines or recommendations encourage physicians to counsel about diet, nutrition, and lifestyle, these are based only on expert opinion. "The evidence that this type of counseling actually improves patient outcomes is thin to nonexistent."

She also noted that the "Choosing Wisely" recommendations were developed on the basis of processes that are less rigorous and detailed than those used to create actual guidelines (eg, those developed by the American College of Physicians in 2000), and the two probably should not be confused.

Low Value

"'Choosing Wisely' recommendations are intended to identify practices that often — although not always — represent low-value care and that patients and physicians should discuss and question," said Dr Loder. "That is not the same thing as saying they are always inappropriate."

According to these recommendations, situations exist where the use of opioids and barbiturates may be appropriate, although they should not be first-line treatments in most situations, said Dr Loder. Similarly, the recommendations encourage the use of advanced imaging with MRI rather than CT in nonemergent headache for which imaging was thought to be appropriate.

"I think the authors are overstating the case for limiting imaging studies when they say that there is 'broad agreement' on these issues and characterize these things as 'low-value services.' They are not inherently low value — they are low value only in specific contexts."

Dr Loder found it "very interesting" that among those presenting with headache, women were significantly less likely than men to have imaging studies.

Dr Mafi and Dr Loder have disclosed no relevant financial relationships. Dr Loder is acting director of research at the BMJ but noted that although she is paid for that work, her comments do not necessarily reflect the views of the BMJ, the AHS, or Brigham and Women's Hospital.

J Gen Intern Med. Published online January 8, 2015. Abstract

Clinical Implications

• The prevalence of recurrent, severe headaches approaches 25%. There are approximately 12 million clinician office visits for headache per year in the United States. The inappropriate application of head CT imaging for headache has been implicated in promoting higher rates of cancer. The most common indication for inappropriate CT imaging is headache.
• The current study by Mafi and colleagues suggests that physicians are actually doing worse across time in offering best practices to patients with headache, particularly regarding patient referrals and counseling. Rates of preventive medications for headache did increase across time in the current study.

CME Test