Evidence-Based Medicine
Should US physicians use the Canadian head CT rule?

 “Use the rule to cut costs and radiation without litigation exposure.”
Per capita, the US currently spends twice as much as Canada, yet ranks behind on many quality measures. Undoubtedly, the US malpractice environment contributes to this fiscal conundrum. In caring for anyone, anyplace, anytime EM physicians sit at the forefront of the defensive medicine frontier. Rarely, do EM physicians benefit from well-established science to augment clinical gestalt. Deciding whether or not to obtain a head CT on head injury patients is one exception.
Mild traumatic brain injury can be defined as loss of consciousness (LOC), amnesia, or witnessed disorientation with initial GCS ≥ 13. The Canadian Head CT (CHCT) rule was developed to guide decision making about the use of CT-head ordering excluding those less than16 years old, pregnant, on Coumadin, suffering seizures, lacking clear trauma, LOC, obvious skull fracture, or hemodynamic instability. The primary outcome was the need for an intervention (craniotomy, skull fracture elevation, ICP monitoring, intubation for head injury), while the secondary outcome was clinically important brain injury on CT as defined by a consensus of 129 neurosurgeons, radiologists, and EM physicians. Insignificant CT findings included: solitary cerebral contusion < 5mm; localized subarachnoid blood < 1mm; thick smear subdural hematoma <4 mm; isolated pneumocephaly; and closed depressed skull fracture not through the inner table. After being derived and validated on over 3000 Canadian patients, the CHCT rule has subsequently been validated in Germany and England. Ironically, in contrast to the US debate about whether to utilize the CHCT rule to decrease head CT ordering, British physicians debated whether to incorporate it into their guidelines fearing it might increase CT use! Other blunt head injury clinical decision rules (CDR) have been developed, but none explicitly followed the guidelines for CDR development and subsequently none have been externally validated.

While the CHCT rule has not been demonstrated cost-effective or widely acceptable, no US legal literature demonstrates appropriate use of the CHCT rule missing a clinically significant, successfully litigated intracranial injury. Nonetheless, no sane physician wants to be the first. Therefore, the CHCT rule should not replace clinical judgment. Instead, the rule should augment the intuition experience garners. If a seasoned EM physician’s insight suggests that central nervous system imaging is merited despite a negative CHCT rule, the patient should definitely be scanned. If the clinician’s gestalt is that no significant intracranial injury exists, concern about the medical-legal implications of the rare (0.6% neurosurgically significant – all detected by the CHCT Rule) clinically significant injury should not drive decision making. Practicing knee-jerk, counterintuitive, defensive medicine drives up the cost of health care for everyone and negates the value of clinical expertise.

The clinically insignificant CT findings represent the central argument against the CHCT rule. Dr. Klauer has argued that if one of these injuries is missed by one physician only to be found by another physician, a lawyer will portray the patient as a major corporate CEO if only the first physician had ordered the scan even if few surgeons in the world would intervene on the injury. If detecting any abnormality regardless of the cost to the individual or society is top priority, recognize that a brain injury can exist despite a normal head CT. If we are to take a fully defensive posture and toss the evidence out with the tort reform, should we not order more sensitive head injury tests such as MRI, PET scans, serum biomarkers, and formal cognitive testing before any of these patients leave the ED?  Most EM physicians would not advocate this extreme viewpoint, but what distinguishes unnecessary head CT ordering from other unnecessary brain injury testing?

Tort reform is past due to alleviate counterintuitive over-testing. Intracranial injuries, missed or recognized, often have devastating results for individuals and families. The CHCT rule, or future worthy substitutes, ought to continue to be the focus of research until cost-effectiveness and acceptability can be demonstrated. If physicians are being legally persecuted for appropriately applying these rules, the case details and testimony should be reported in venues such as EP Monthly.  Lacking an effective alternative, cost-conscious, evidence-based physicians are left with two choices: augmenting clinical decision making with the CHCT rule or substituting best-evidence practice for uniform testing of all head injury victims at tremendous cost to society with no proven benefit to the individual patient.

Christopher Carpenter, MD, is an Assistant Professor at Washington University School of Medicine.
Should US physicians use the Canadian head CT rule?
“A miss is a miss. Juries don’t care if you’re saving cost!”
by Kevin Klauer, DO
The Canadians have demonstrated a unique expertise in developing evidence-based rules that make sense. However, as a risk manager, board member of our group’s risk retention group and a practicing emergency physician, clinical application is where the rubber meets the road.

At journal club, this is an academic discussion. In the ED, we need to make certain we are practicing safe, appropriate and efficient, risk-averse emergency medicine. Which would you prefer to do? Avoid a lawsuit and all the fun this will add to your life for 18 months; or use an evidence-based guideline to convince a jury of laypersons why it was OK to miss a small subdural.
We could have an academic discussion. However, how does that help the average doc working a shift? Many research-oriented discussions focus on the quality of the research. Unfortunately, they fall short with how or if the included recommendations should be implemented clinically and how to do so.

For most of us, this is not an academic discussion. It’s a matter of how to get through the next shift, practicing high quality emergency medicine, avoiding a lawsuit, walking away with some sense of gratification for the care we provide.

The issues I have with the Canadian CT Head rule are a lack of external validity, the exclusion of “clinically unimportant” lesions and the use of a utilization reduction tool that may result in missed intracranial injuries.

External validity is key. Can the study results be reproduced elsewhere, or is it only valid in the environment in which the original study was conducted. In example, an excellent head CT rule was studied at Louisiana State University (LSU) in 2000. It reduced CT utilization without any additional missed injuries. Unfortunately, attempts to reproduce the same results in other environments have been unsuccessful. Thus, rendering application of their rule to LSU only.

Some might say that the Canadian CT head rule has stood the test of external validity. Well, the devil is in the details. Boyle, et al., British investigators, noted a substantial increase in utilization of CTs, without any additional benefit. In Australia, the New Orleans criteria and the Canadian CT head rule were tested. Neither performed well. Rosengren, et al., from down under, reported that if the high-risk Canadian criteria were utilized, 74% of the normal scans could have been avoided, and the one patient requiring “neurosurgical intervention” would have been identified. However, with addition of the medium-risk (2 additional) criteria were used, only a 46.7% reduction in scans would be noted. Unfortunately, two “clinically significant” abnormalities would have been missed. Their conclusion was that such studies have no applicability outside the environment in which they were developed. Hence, they lack external validity. I agree with them!

If you feel external validity has been proven by these studies, conducted in Europe and Australia, how does that apply to apply to those of us practicing in the U.S.? When’s your next shift in Canada, England or Australia?

A fundamental challenge applying the many utilization-driven guidelines, generated by Canadian researchers, to EDs in the United States is that our medical malpractice climate does not afford any forgiveness for medical error. The Canadian CT head rule was 92% sensitive for identifying traumatic neurosurgical lesions. CT utilization could have been reduced to 32%. Unfortunately, only 320 of the 348 patients with positive CT scans would have been identified. Stiell was proud to report that they did not miss any of the lesions requiring neurosurgical intervention. However, they don’t talk much about the 28 “clinically important” and “clinically unimportant” lesions the rule would have missed.

In the malpractice climate in the United States, there is no such thing as a “clinically unimportant” intracranial injury. Any such miss is an opportunity for a lawsuit. If a claim is made, convincing a jury of laypersons that the lesion was “unimportant” will be nearly impossible. Furthermore, average defense costs will be at least $100,000, and that’s just to get you to your first day of trial.

Great concept. But, it falls way short on practical application. Plaintiffs and jurors don’t commend a physician for reducing utilization and saving healthcare costs when they have an injured patient in front of them. I agree that we spend far too much on unnecessary tests in this country. However, until our legal system accepts misses in exchange for controlling costs, we can’t put ourselves at risk.

Kevin Klauer, DO, is the Director of Quality and Clinical Education at EMP & the Director of the Center for Emergency Medicine Education


# John Frey, MD 2008-01-13 16:25
I think many new grads quickly realize that some of what they were taught in residency concerning cost savings and resource utilization is merely opinion and will leave them more vulnerable in the face of a claim.
# Mike Klevens, MD FAAEM 2008-01-14 17:12
It is not just cost savings that we are talking about. Defensive medicine, which is the sole basis of the "con" argument undermines the emergency department and the entire hospital system. By needless and time-consuming testing, we delay patients who are in the waiting room or delay patients who are in the ED waiting for a bed upstairs. Those patients are often the ones at risk of a catastrophic life threat. I don't want to miss those patients. Resource utilization and knowledge translation is validated and the time has come to put away the hubris that anecdotal knowledge is somehow superior.
# Douglas Char 2008-01-14 22:10
Thanks to Dr Carpenter and Dr Klauer to providing two opposing perspectives. As with much of medicine the "real world" is somewhere in the middle. While I don't subscribe to the "test everyone or risk getting sued" school of practice, there are times when testing in the face of low yeild is appropriate - esp if the outcome being tested for is a righ risk diagnosis. The bigger question is how do I communicate with my patients the risks and benefits of testing (or not testing)for these infrequent but serious conditions? Depending on the patient, family and situtation a frank discussion that is documented may be more helpful than "preemptive testing to protect my butt". Well written clinical guidelines and rules help me feel more comfortable in choosing one path or the other. Sometimes I'm comfortable knowing that the patient and family understand the inherent risks of "wait and see" and will come back if things change. On the other hand there are times when I just order the test because I can't trust the patient or the social situation is concerning. I'd like to be a purist but our patients live in a gray (not black or white) world.
# MDDaniel Theodoro 2008-01-15 16:43
As an academic physician at a research institution, I'm not immune to malpractice lawsuits, overcrowding, or administrative pressure to see more patients in less time. In fact I would argue that these pressures are likely to affect my thinking in my last few hours of a shift and it's an area where a CDR may enhance my judgement. What irks me the most is the 'fear of litigation' argument against CDRs. Tell me, as Dr. Klauer elegantly does, that the CDR lacks external validity and requires further development. Don't argue that the malpractice climate has no room for a CDR and they should all be abandoned in favor of nebulous clinical gestalt. That argument implicitly agrees that physicians can have zero bad outcomes 100% of the time if they really try hard enough and hone their 'gestalt,' whatever that means. What I think our field needs more than anything for tort reform is a set of reasonable expectations we all agree upon. Tort reformers should champion wide acceptance of these CDRs by our national organizations and begin to establish differences between malpractice and mal-occurrence. Only when we, as physicians, identify where the limitations of medical practice exist can we define and reform what 'malpractice' is. CDRs are a step in that direction.
# Larry Lewis 2008-01-18 21:49
First of all, I do not believe you can get a CDR that over the long haul has 100% sensitivity. Just as pre-marketing drug testing on a couple of thousand people will miss rare, but serious side effects that become evident when the drug has been used on hundreds of thousands of people, a CDR derived and validated on a relatively small population will likely overestimate sensitivity, especially when it approaches 100%.
That being said, CDRs, especially those that have been evaluated in more than one setting can be very useful to support a decision to not test any further.
Even though the relationship between negligence and malpractice suits is tenuous, with a 1991 NEJM article which reviewed about 30,000 hospital discharges and 3500 malpractice claims in New York, showing that only 2% of negligent injuries resulted in claims and only 17% of claims appeared to involve a negligent injury.
Having served as a defense expert in malpractice cases both before and after the implementation of CDRs, I can tell you that when a physician follows a CDR, it truly shifts the burden of proof to the plaintiff team. I recently aborted a lawsuit when I told a plaintiff lawyer that a doctor who did not get a head CT in someone with blunt head injury had followed evidence based rules (which they had). I also was able to use the PORT score to help defend the decision to not admit a person with pneumonia. These experiences lead me to believe that not ordering a test or admitting a patient when there is good evidence to support you is less likely to result in lawsuits. A CDR helps to objectify the ethereal "standard of care", and the plaintiff attorneys as well as the defense attorneys recognize this.
# Thomas Green 2008-02-07 01:51
I do not believe that there is one correct answer here. We must look at these studies for what they are: instrument decisions - and folks like Bill Mower and Gerry Hoffman go the extra step to purposely state that they are not "rules" but instead decision instruments - maybe its semantics but these are instruments to help decide which patients are at low risk for a clinically significant (usually requiring some sort of neurosurgical intervention) intracranial lesions.
When looking at the New Orlean's Criteria (by Haydell et al) their rule seems to be overly sensitive and very non-specific with regards to one of their criteria - any trauma above the clavicles - using just this one part of the criteria would subject many patients to CTs that I would never normally order a CT on. Stiell's CCT rule had over 10% of patients that were excluded because there was no follow-up and he comments in a response that since they were at the referral neurosurgery center that they wouldn't have missed any significant injuries. Then there is the NEXUS-II which is a good instrument - it misses some (1.7%) but then again, what question are we asking, "Who is look risk for a clinically significant injury?" They excluded those that never even got a CT - so the real minor cases - which would have increased their sensitivity.
That being said, when I go into my next shift, I will use the available literature, evidence as well as clinical judgment to make my informed decision as well as my discussion with the patient.
One issue that has not been addressed is the issue of radiation exposure and the lifetime risk of cancer and malignant transformation. According to the National Academy of Sciences- BEIR VII report - The risk of developing cancer is 1 in 1000 after exposure to 10 mSv of radiation (Head CT = 2 and Abd/Pelvis = 9) and this number is 1 in 550 in kids. In a study by Lee et al. in Radiology in 2004 - 5 out of 76 pts (7%) were told about the risks of the CT scan they were going to undergo. 22% of ED physicians had reported giving such information to their patients. Less than 50% of radiologists believed that there was an increase risk of cancer. In addition, all patients and most ED physicians and radiologists were unable to accurately estimate the dose of radiation for one CT scan compared to one chest x-ray (CT brain = 120 CXR's, CT Chest = 350 CXR's and CT Abd/Pelvis = 500 CXR's).
As its been stated previously in the literature - technology is a good thing but the end-point can be VOMIT (Victims Of Modern Imaging Technology) - CT is so available, so easy to obtain and order and we are going to start seeing more and more cases of litigation resulting from lack of informed consent as well as malignancies.
So, while no decision instrument or rule should be blindly followed nor will it necessarily protect one from a lawsuit - using the available literature and evidence along with common sense and clinical judgment will allow us to practice good, sound medicine.

Add comment

Security code

Popular Authors

  • Greg Henry
  • Rick Bukata
  • Mark Plaster
  • Kevin Klauer
  • Jesse Pines
  • David Newman
  • Rich Levitan
  • Ghazala Sharieff
  • Nicholas Genes
  • Jeannette Wolfe
  • William Sullivan
  • Michael Silverman

Subscribe to EPM