A study was recently published in “Academic Emergency Medicine,” a medical journal for emergency room doctors, which said that CT Scans were being done way too frequently for patients with head injuries. The study’s abstract is available here, although the full version requires a subscription, so I can’t post the whole thing.

The medical researchers did a computerized review of 27,240 medical records from patients who had been seen at 14 emergency departments between 2008-2013, and then did an actual review of 100 of those records to see if they lined up with what the computer said. Based on those reviews, the researchers concluded that “36.8% of trauma head CTs were truly likely avoidable.”

As a medical malpractice lawyer, I’m always suspicious when a new study comes out that claims doctors should perform fewer tests. These studies are often used during medical malpractice trials to help the defendant doctors and hospitals avoid responsibility for their failure to do a test. That is, for example, what the American College of Obstetricians and Gynecologists has done for years with its studies on brain injury at birth. Is the same true for this study of emergency room practices when it comes to CT scans of the brain after a head injury?

Let’s back up for a moment and review what the study looked at.

The Canadian CT Head Rule criteria (CCHR) are designed to identify patients at risk for serious intracranial injury that requires neurosurgical intervention, like a surgery to release pressure from a brain bleed. If a patient meets any of these criteria, the CCHR classifies the patient as high risk and recommends that a head CT scan be performed. The five main criteria are:

  1. Glasgow Coma Score <15 at 2 hours after injury
  2. Suspected open or depressed skull fracture
  3. Sign of basal skull fracture (hemotympanum, ‘raccoon’ eyes, cerebrospinal fluid otorrhea/rhinorrhea, or Battle’s sign)
  4. Vomiting ≥2 episodes
  5. Age ≥65 years

The researchers admitted that their records didn’t easily tell them if the patient had a specific type of skull fracture, so all they really looked at was the Glasgow Coma Score, vomiting, and age.

It doesn’t take much for a patient to get a 15 on the Glasgow Coma Score: they just need to be able to use their eyes properly, respond to questions appropriately, and be able to move their arms and legs in response to pain. Anyone who can’t do any of those things after a head injury likely has a serious brain injury and definitely needs a CT Scan. So all the “CCHR” is really looking at is vomiting and age.

Do we really want to have emergency medicine physicians who, after a head injury, ask only if the patient has vomited? In many ways, the most important part of a trip to the emergency room is when the doctor takes the “clinical history” so they can get a whole picture of what’s going on with the patient. Do we want to discourage doctors from ordering a CT Scan when that clinical history – which might include nausea, memory issues, and sleep disturbances, none of which show up on the CCHR criteria – suggests the possibility of a brain injury?

But perhaps the most disturbing part of the study was the “outcomes” they looked for:

Our primary outcome was discordance with the CCHR among adult patients receiving a CT for head injury (likely avoidable imaging). Our secondary outcome was the number of patients in the discordant cohort who ultimately had a neurosurgical intervention within 30 days of the index ED visit.

In other words, the researcher did not look at all for patients who didn’t get a head CT scan but then ended up with a subsequent brain injury. This study looked only at patients who actually got a CT scan, and didn’t look at all for patients who did not get a head CT scan but should have. Unfortunately, that’s exactly the situation which is the most dangerous for the patient, and the situations we see regularly in our medical malpractice work.

Thus, unfortunately, this looks like another study with a questionable design from the beginning. The researchers didn’t go looking for whether there were too many, or too few, brain CT scans after a head injury. They looked only for evidence there could be too many, and, big surprise, found what they were looking for. So when are they going to go back through those files and review the patients who, after a head injury, did not get a CT scan? That would tell us the whole story.

(For more information, Emergency Medicine Literature of Note has a few words about it, too.)