Author Insights: Neuroimaging for Headache is Overused and Provides Little Additional Benefit

Brian C. Callaghan, MD, MS, an assistant professor of neurology at the University of Michigan Health System in Ann Arbor, and colleagues suggest headache neuroimaging is common, costly, and overused. (Image: University of Michigan)

Brian C. Callaghan, MD, MS, an assistant professor of neurology at the University of Michigan Health System in Ann Arbor, and colleagues suggest headache neuroimaging is common, costly, and overused. (Image: University of Michigan)

Although most headaches are caused by benign conditions, sometimes they signal the presence of a more dangerous condition, such as a brain tumor or aneurysm. To determine if such a condition is present, a physician, often at the request of the patient, will order neuroimaging with computed tomography (CT) or magnetic resonance imaging (MRI).

But a research letter appearing Monday in JAMA Internal Medicine suggests neuroimaging for headache generally goes against current recommendations from multiple guidelines; is overused, costing the health system hundreds of millions of dollars; increases patient radiation exposure; and can detect incidental findings that lead to other tests and procedures for often benign conditions.

Using the National Ambulatory Medical Care Survey, the study authors found that from 2007 through 2010, there were 51.1 million adult headache visits, mostly to primary care physicians (54.8%); 88% were by patients younger than 65 years and 78% were by female patients. Neuroimaging occurred for 12.4% of all headache visits during that period, costing a total of $3.9 billion, and the use of neuroimaging for headaches has increased substantially in recent years, from 5.1% in 1995 to 14.7% in 2010.

Lead author Brian C. Callaghan, MD, MS, an assistant professor of neurology at the University of Michigan Health System in Ann Arbor, discusses his team’s findings.

news@JAMA: Why has the use of neuroimaging for routine headaches increased in recent years?

Dr Callaghan: My speculation is that imaging use in general has been increasing across the board as MRI and CT become more accessible. But the number one reason physicians give scans for headache is patient reassurance, not to detect a bad intracranial condition. The next reason is legal—physicians not willing to miss the very rare condition.

news@JAMA: Do reassurance and concerns about potential legal consequences have societal costs?

Dr Callaghan: The major problem is that we’re ordering lots of tests, and it’s a huge amount of money. Headache neuroimaging is one of the big-ticket items where we spend a lot of money, and we don’t get much bang for our buck.

news@JAMA: Why do patients seek reassurance, and is there a cost to seeking such reassurance?

Dr Callaghan: If you ask most patients who have had bad headaches why they want a scan, they would say that they worry about a brain tumor. But they don’t think about the other things that can harm them. They don’t appreciate some of the downstream consequences, like radiation exposure from CT or undergoing an MRI scan and getting a false-positive for something else that leads to more tests and procedures.

news@JAMA: How should physicians reassure patients who request imaging because they worry their headaches signal a serious condition?

Dr Callaghan: There are some circumstances where neuroimaging is warranted, but if the physician feels this is not one of those circumstances, he or she has to have a conversation with the patient to explain why the headaches are occurring and why an imaging test, with its potential side effects, is not warranted.

Author Insights: Top 5 List of Diagnostic Tests That Emergency Physicians Can Forgo

Jeremiah D. Schuur, MD, MHS, department of emergency medicine at Brigham and Women’s Hospital in Boston, and colleagues created a top-five list of diagnostic tests they believe offer little value when assessing patients in the emergency department. (Image: Brigham and Women's Hospital)

Jeremiah D. Schuur, MD, MHS, department of emergency medicine at Brigham and Women’s Hospital in Boston, and colleagues created a top-five list of diagnostic tests they believe offer little value when assessing patients in the emergency department. (Image: Brigham and Women’s Hospital)

Diagnostic testing, treatments, and hospitalizations ordered by emergency physicians account for up to 10% of US health expenditures. From 2003 through 2011, the mean cost of an emergency department visit increased about 240%, from $560 to $1354. In an effort to rein in unnecessary and costly activities, an expert panel identified a top 5 list of tests, treatment, and disposition decisions (whether to admit patients to the hospital or discharge them) that they believe are of little value and can be controlled by emergency medicine physicians and avoided for most patients.

The list appears today in JAMA Internal Medicine.

The expert panel considered 64 tests, treatments, and disposition decisions before producing its top 5 list:

• Don’t order computed tomography (CT) of the cervical spine for patients after trauma who do not meet certain criteria
• Don’t order CT to diagnose pulmonary embolism without first determining a patient’s risk for pulmonary embolism
• Don’t order magnetic resonance imaging of the lumbar spine for patients with lower back pain without high-risk features
• Don’t order CT of the head for patients with mild traumatic head injury who do not meet certain criteria
• Don’t order tests to assess blood clotting times for patients without bleeding or unless you suspect they are on anticoagulation therapy or have a clotting disorder

The expert panel consisted of 7 emergency physicians practicing at 6 emergency departments in the Partners Healthcare health system in eastern Massachusetts; these 6 departments account for more than 320 000 annual patient visits. The panel members said their process provides a method that other emergency departments can use to identify targets of overuse that can be controlled by emergency physicians to improve the value of health care services delivered.

Lead author and expert panel member Jeremiah D. Schuur, MD, MHS, department of emergency medicine at Brigham and Women’s Hospital in Boston, discusses his team’s list.

news@JAMA: Why did you create a top 5 list of low-value clinical actions in the emergency department?

Dr Schuur: I read an article that really struck a chord with me on how physicians have a responsibility to be stewards of resources as part of our missions to improve health. So I thought it was important for emergency medicine physicians to do this. The article made a strong case for the idea that physician specialties largely control the knowledge and guidelines around much of what we do, and it was within our power to identify tests, treatments, and procedures that are not uniformly beneficial to patients. I thought the narrative of a top 5 list was a good way to get the message across.

news@JAMA: You emphasize “actionability.” What is it, and how does it pertain to your list?

Dr Schuur: Actionability is important in emergency medicine because we practice in close collaboration with many other specialties and rely on other practitioners and institutions to continue care. For example, ordering preoperative testing is something emergency physicians often order when a patient will be admitted. And while we may place the orders, ultimately the surgical admitting service is making the decisions of what needs to be ordered. So, we thought, the average practicing emergency physician is not going to be able to change that dynamic. On the other hand, for example, many who present with head injuries get treated and released, and the emergency physician is responsible for all the decisions.

news@JAMA: Might emergency physicians be hesitant to embrace the recommendations in your top 5 list for fear of being sued for not doing everything possible when treating a patient?

Dr Schuur: I understand physicians fear malpractice, but our recommendations are structured such that you can avoid doing those things safely. The best way to avoid medical liability is to practice high-quality medicine, and the top 5 items we chose are strongly supported by clinical evidence.

Improved end-of-life care, limited use of imaging, and reduced drug costs could reduce the increasing costs of cancer care, according to 2 oncology experts. (Image: fluxfoto/iStock.com)

Improved end-of-life care, limited use of imaging, and reduced drug costs could reduce the increasing costs of cancer care, according to 2 oncology experts. (Image: fluxfoto/iStock.com)

Oncologists could cut the ever-increasing cost of cancer care by improving end-of-life care, limiting the use of expensive imaging, and using cancer drugs more cost-effectively, according to experts at the Johns Hopkins Medical Institutions in Baltimore, Maryland.

“We need to find the best ways to manage costs effectively while maintaining the same, if not better, quality of life among our patients,” Thomas Smith, MD, coauthor of the article released online today in The Lancet Oncology, said in a statement.

Smith and coauthor Ronan Kelly, MD, noted that the number of new cancer cases diagnosed annually is projected to increase to 21.4 million in 2030. In the United States the cost of care is expected to rise 39% by 2020, to $173 billion. But that cost isn’t driven only by the aging population and an increased demand for services. “In the [United States], 91% of the rise in costs was due to price increases,” they wrote.

The best opportunities for safe, ethical cost reductions are in caring for patients with metastatic cancer, not new surgical or radiation treatments, clinical trials, curative care, or pediatric care, according to Smith and Kelly, both oncologists at Hopkins’ Sidney Kimmel Comprehensive Cancer Center.

Their article notes that among Medicare’s cancer patients, 60% are hospitalized in the last year of their life, 30% die in a hospital, and only 54% ever use hospice care. However, “hospice and palliative care provide better overall care at a smaller cost than hospital care,” they wrote. “Hospice care also improves symptoms, reduces caregiver distress, and saves $2700 to $6500 per person as compared with care that does not actively involve a hospice.”

They cited a 2012 study indicating that of 608 deaths at a tertiary care center, 229 of the patients were eligible for hospice care at their second to last visit, but only 17 were informed about it. Additional studies show that hospice or palliative care may reduce costly hospital readmissions, they added.

The US cost of cancer imaging has increased by about 5% to 10% annually between 1996 and 2006 “without attendant changes in mortality from metastatic disease,” the authors wrote. They noted that the cure rate from salvage chemotherapy for lymphoma is the same regardless of whether a recurrence is diagnosed clinically or by a routine surveillance positron emission tomography scan.

The American Society of Clinical Oncology addressed appropriate imaging use in its Choosing Wisely recommendations. “Clinicians explaining the truth to patients about curability [and imaging] will allow most patients to choose wisely,” Smith and Kelly wrote.

The high cost of new drugs—sipuleucel-T, used to treat advanced prostate cancer, costs $93 000 per course of treatment and prolongs survival for a median 4.3 months—should motivate clinicians to clearly define “meaningful clinical benefit” when they make treatment decisions, the pair added. “There does seem to be some recent enthusiasm to establish thresholds,” they wrote.

Their article explains how several approaches, including accountable care organizations, competitive bidding for drug prices, and shared decision making with patients may reduce cancer care costs.

“The oncology community should do its part to limit itself to use of evidence-based treatments, standardize care with pathways, transition away from fee-for-service to alternative payment methods, and change communication with patients about reasonable end-of-life care,” Smith and Kelly wrote.

“We need political leadership working together with oncologists and patients to explicitly address the issue of what value is needed to justify expensive treatments.”

Waiting and Watching Can Reduce Use of Brain Scans for Kids in the Emergency Department

Monitoring a child with a minor head injury in the emergency department may help reduce unnecessary brain scans. Image: meshaphoto/iStockphoto.com

Monitoring a child with a minor head injury in the emergency department may help reduce unnecessary brain scans. Image: meshaphoto/iStockphoto.com

Allowing a longer period of observation for a child with a head injury in the emergency department before deciding whether a brain scan is necessary reduces the use of such scans and may protect children from unnecessary radiation exposure, according to a study published in the Annals of Emergency Medicine yesterday.

The growing use of computed tomography (CT) scans among children has raised concerns that such scans may be exposing children to harmful levels of radiation. A recent study highlighted by news@JAMA found that CT scan rates had increased by 2 to 3 times among various age groups between 1996 and 2005. These concerns have led to calls for more judicious use of CT scans in children.

In the latest study, researchers studied whether choosing to wait and observe a child with a head injury would reduce use of CT scans. The team enrolled 1381 patients who presented with head injuries at Boston Children’s Hospital and found that every additional hour of observation reduced the likelihood that a child would receive a CT scan. Ultimately, about half the children were observed in the emergency department, about a third were sent home without observation, and 17% received an immediate CT scan.

The patients who were observed were monitored in the emergency department for an average of 4 hours, but the authors note that the optimal time for observation is not clear. They said that large multicenter trials are necessary to confirm their results.

Recent recommendations from the American Academy of Pediatrics (AAP) suggest that observation or another evaluation tool should be used in the initial evaluation of a minor head injury in a child.

“Clinical observation prior to CT decision-making for children with minor head injuries is an effective approach,” according to the AAP.

Author Insights: Use of CT Scans on Children Remains High, May Increase Cancer Risk

 Diana L. Miglioretti, PhD, a senior investigator at the Group Health Research Institute and professor of biostatistics at the University of California–Davis School of Medicine, and her colleagues found high rates of CT use among children may contributed to increased cancer risk. Image: UC Regents©


Diana L. Miglioretti, PhD, a senior investigator at the Group Health Research Institute and professor of biostatistics at the University of California–Davis School of Medicine, and her colleagues found high rates of CT use among children may contributed to increased cancer risk. Image: UC Regents©

Children continue to undergo many more computed tomography (CT) scans than in the past, and the doses of radiation they are receiving may increase the risk of certain types of cancer, according to a study published in JAMA Pediatrics today.

To better understand the trends in pediatric CT and the risks posed to children by such scans, Diana L. Miglioretti, PhD, a senior investigator at the Group Health Research Institute and professor of biostatistics at the University of California–Davis School of Medicine, and her colleagues evaluated the use of CT scans between 1996 and 2010 in children younger than 15 years in 7 US health systems. They found that the use of CT doubled among children younger than 5 years (from 11 to 20 scans per 1000 children) and almost tripled for older children (from 10.5 to 27 per 1000 children) between 1996 and 2005. But by 2010, CT use had decreased to 15.8 scans per 1000 children younger than 5 years and 23.9 per 1000 children children aged 5 years and older.

The team found that the radiation doses that children were exposed to varied considerably and that the cancer risks associated with exposure were estimated. For example, exposure to abdominal CT scans posed the greatest cancer risk, with an estimated 1 case of abdominal cancer among every 300 to 390 girls scanned or among every 670 to 760 boys scanned.

But the team suggested that reducing the highest 25% of radiation doses might reduce 43% of CT-related cancers.

Dr Miglioretti discussed her team’s findings with news@JAMA:

news@JAMA: Why do you think the use of CT scans in children has increased?

Dr Miglioretti: One thing is that CT is a very sensitive and accurate diagnostic tool. It is very useful in children, especially because other diagnostic technology like magnetic resonance imaging [MRI] requires children to be still for a long time. CT is very fast.

news@JAMA: You found that CT use remains elevated compared to the 90s, but has dipped somewhat since 2005. Why?

Dr Miglioretti: I personally believe the decrease is due to campaigns to reduce radiation exposure in children, such as the Image Gently campaign.

news@JAMA: Despite some improvements in CT use, you still found that current rates of CT use in children may raise cancer risk. How did you estimate cancer risk associated with CT use in children?

Dr Miglioretti: First, we calculated actual radiation exposure for a random sample of hundreds of CT scans of each of the 4 most frequent types (head, chest, abdomen, and spine) by extracting the technical parameters used for each scan associated with exposure. We used a new modeling approach from the National Cancer Institute to estimate organ doses from these exams, and used models from the BEIR-VII report to project the cancer risks due to radiation exposures of those levels.

news@JAMA: Were the exposure levels you found consistent with what other studies have found?

Dr. Miglioretti: We found exposure levels were highly variable for exams of the same type. In terms of risk, we found the cancer risks associated with CT exposure in our study were higher, which was due to the variation in radiation exposure we found. Prior studies used mean doses reported in the literature, rather than actual exposure levels.

news@JAMA: Why do you think the doses varied so much?

Dr Miglioretti: We suspect it is due to differences in radiologists’ preferences and differences in technologists’ knowledge of the importance of reducing radiation exposure in children. Differences in the type of scanner and differences in patients’ sizes might also contribute.

news@JAMA: What do you think physicians should learn from your study?

Dr Miglioretti: We can drastically reduce the number of cancers caused by CT if we only use CT scans on children when they are medically necessary. We need to be judicious in our use to make sure they are medically necessary and that there is not another diagnostic modality we can use, such as ultrasound. If CT is used, we need to optimize the dose for children. The Image Gently website is a great resource for both physicians and parents.

news@JAMA: What should parents know about your results and their implications for CT use on their child?

Dr Miglioretti: It is important to know that if the CT is medically necessary, the benefits likely outweigh any individual cancer risks. The increased risks become a concern when we look at the whole population, but the risk to the individual child is very low. Parents should also feel empowered to discuss their concerns with their physician. They should ask the physician whether CT is medically necessary and how it will change the clinical care for their child. They should ask whether there are alternatives; could MRI or ultrasound be used—or other approaches, such as waiting? If CT is medically necessary, they should ask whether the radiation dose will be optimized for their child to keep the dose as low as possible.

Author Insights: Mammography Every Other Year for Older Women Doesn’t Increase Risk That Breast Cancer Will Be Advanced When Detected

Karla Kerlikowske, MD, of the University of California, San Francisco, and colleagues found that older women who undergo biennial screening mammography are less likely than comparable women who undergo annual screening to have false-positive results and are no more likely to have a cancer detected at an advanced stage. (Image: Susan Merrell, UCSF)

Karla Kerlikowske, MD, of the University of California, San Francisco, and colleagues found that older women who undergo biennial screening mammography are less likely than comparable women who undergo annual screening to have false-positive results and are no more likely to have a cancer detected at an advanced stage. (Image: Susan Merrell, UCSF)

Women aged 50 to 74 years who undergo screening mammography every other year, even those with high breast density or a history of hormone therapy after menopause, appear no more likely to be diagnosed with advanced-stage breast cancer than comparable women who are screened annually. They also have a lower risk of receiving false-positive results, which means they are less likely to undergo unnecessary tests, such as additional imaging or a biopsy, according to findings appearing today in JAMA Internal Medicine.

In addition, the study found that for women aged 40 to 49 years with extremely dense breasts who are considering mammography, annual examinations may minimize their risk of advanced-stage disease, but they have a greater chance of experiencing false-positive results. Continue reading

New Approach May Prevent Some Unnecessary Emergency CT Scans in Children

Seven clinical criteria reported in a new study can help emergency physicians rule out which children with abdominal injuries do not need a diagnostic computed tomography scan. (Image: JAMA, ©AMA)

Seven clinical criteria reported in a new study can help emergency physicians rule out which children with abdominal injuries do not need a diagnostic computed tomography scan. (Image: JAMA, ©AMA)


A new study shows how emergency department physicians can avoid needless diagnostic computed tomography (CT) scanning of children with abdominal injuries. Researchers who developed a decision support tool say their findings could prevent many vulnerable children from being unnecessarily exposed to medical radiation.

Between 5 million and 9 million CT scans are performed on US children annually, according to the National Cancer Institute. The scans can save lives, but with annual growth in the number of CT scans estimated at 10%, they’re now the top contributor to medical radiation exposure in the United States.

Published online today in the Annals of Emergency Medicine, the study describes 7 clinical criteria that accurately predicted which children arriving in emergency departments with blunt torso trauma didn’t have injuries severe enough to require a diagnostic CT scan. Continue reading