Dial M for Misdiagnosis

Mild traumatic brain injury (TBI) is second only to migraine headache as the most common neurological disorder worldwide, yet we lack reliable methods to rapidly differentiate those patients who have suffered intracranial pathology from those who have merely taken a bump to the head.

The result is that healthy persons often receive unnecessary imaging, hospitalization, and restrictions on work and leisure activity. These patients are often left with an altered view of themselves as a brain injury “survivor” and told to expect certain symptoms. In fact, a full recovery in six months or less is nearly universal from a single, mild TBI when there is a loss of consciousness of less than 20 minutes, a Glasgow Coma Scale of 13 or higher, and no neuroimaging evidence of intracranial pathology. In the vast majority of cases, the patient should be told to expect a prompt and complete recovery.

Unfortunately, in the context of the stressful events that often accompany a concussion (e.g., a motor vehicle accident or emergency medical care), many of these patients attribute otherwise benign emotional, physiological and memory symptoms to their injury. Such selective attention causes even more stress and more symptoms, which confirms their perception that they have experienced the expected symptoms and further increases their anxiety. Then the insurer is left with a demand for compensation for brain injury in a case without objective medical evidence of intracranial pathology.

Even worse are the small numbers of patients, around 8%, who have intracranial pathological injury but display no focal neurological signs. These patients often don’t get the medical care they require or the financial compensation to which they are legally entitled. That outcome is contrary to the purpose of insurance and our concepts of justice. As we recently saw in the death of actress Natasha Richardson, the results of this mistake can be fatal. The low base rate of mild TBI patients with intracranial pathology and a lack of focal neurologic signs make both of these mistakes easy to make.

Currently, there is no blood-based test to assist physicians in reaching a decision about which mild TBI patients are likely to require brain imaging and further observation and which can be safely discharged from the hospital. We might soon have a solution to this problem.

At the 2009 Military Health Research Forum, Gerald A. Grant, M.D., associate professor of neurosurgery and director of neurotrauma at Duke University Medical Center, reported success using three biomarkers to predict which mild TBI patients had intracranial pathology.

From June 2006 to September 2007, researchers at Duke University Medical Center, a Level 1 trauma center, drew blood from 119 mild TBI patients (defined as having a Glasgow Coma Scale score of 14-15). Blood draws were performed an average of 8 hours and 14 minutes after the time of injury. A predictive model incorporating three biomarkers (D-Dimer, BNP, and S100B) was compared to non-contrast CT images to see if these biomarkers could identify which patients had intracranial pathology. The combination of the three biomarkers was found to be highly sensitive (92%) to intracranial pathology and was found to clearly indicate when the patient was healthy (also 92%).

Unfortunately, this combination of markers was not very specific (32%) in distinguishing brain pathology from other pathological processes. That might have been because of the large range of times over which blood was drawn. S100B begins to lose its sensitivity after about three hours. S100B changes are also associated with other diseases and disorders, which might have caused some false positives. In addition, BNP is elevated in patients with congestive heart failure, renal insufficiency, or advanced age. Further studies with larger patient numbers and blood draws done sooner after the head injury and at a more consistent time are needed to validate this approach. Still, this three-marker approach proved much more useful than past approaches utilizing only a single biomarker.

This approach offers hope that, after further study, we might have a practical test that could easily be performed in the field to determine which patients require further care and which can safely be released to home. This biomarker panel might reduce unnecessary imaging, hospitalization and activity restrictions. Soon a simple blood draw might resolve a problem that has perplexed neuropsychology, neurology and emergency room physicians for many years.