For individual patients, that means unnecessary exposure to radiation. It increases costs, and it extends ED length of stay. The last point “affects not only the patient in particular, but all the other patients who need to be seen,” Dr. Korley says. “The longer a patient occupies an ED bed, the more difficult it is to bring someone from the waiting room back to the ED.” Moreover, he notes, most U.S. hospitals are struggling with overcrowding; the lack of available beds when patients are admitted often leads to boarding in the ED.
That makes these biomarkers a tiny but mighty member of the larger ecosystem. “A test like this enables us to improve our throughput” at a time when it has become pressing to do so, Dr. Korley says.
As noted, the retrospective study was done on a plasma-based test. Eventually, as the laboratory completed its study for diagnostic accuracy and began its discussions on how to implement the test, the FDA approved a whole blood test, which is what Michigan now uses.
“Whole blood was a game-changer,” Dr. Gherasim says, since the test could be performed in the ED laboratories (they use Abbott’s i-Stat Alinity point-of-care system) without additional processing. “So we ended up evaluating the whole blood test with a smaller [n=100] subset of samples.”
One difference is stability of the specimens, she says. “Whole blood has to be tested within an hour of collection.” But turnaround time with the whole blood test is approximately 24 minutes. “Which is really fast,” she says. The test has a strong negative predictive value. The positive predictive value is slightly lower, so anyone who has a positive test result may or may not have intracranial hemorrhage; a CT scan would be needed to resolve that, she says.
While the implementation has, without doubt, been a team effort, the initial push came from the clinical side.
That was a wise move, says Dr. Gherasim. “I find that it works the best.” When the laboratory takes the lead in bringing on a new test, she says, physicians can be slower to feel comfortable using the test.
It’s no secret that influencing how physicians practice is not for the faint of heart. Michigan does not yet have data to track changes in CT orders, but it should be available in the next several months. From early observations, says Dr. Korley, it’s been apparent that CT orders are dropping in cases where the biomarker test is negative. “People are becoming more comfortable with not getting a CT.”
Perhaps not surprisingly for a test adoption that was clinician-driven, there is still heavy clinician say-so in making decisions about whether to order a CT.
While the FDA approved the test with cutoffs, there are data to suggest that the extent of injury might influence biomarker levels. “Therefore, we believe that especially for borderline elevations,” says Dr. Gherasim, “it’s important to know that it may be a patient who presents to the [ED] too soon,” whose biomarker levels may not have peaked. “So if there is high clinical suspicion that a patient still has traumatic intracranial hemorrhage even with a negative biomarker result, a provider can still order a CT scan,” she says. Similarly, with a case involving a borderline elevated result with low clinical suspicion, “It’s going to be a clinical call on the next steps.”
The real surprise might be the lack of so-called drift in how the biomarker test is used, even as the number of clinicians ordering the test has grown. “We brag a little bit about saying we’ve had more than a hundred providers order the test,” Dr. Korley says. “We’re excited about that.” But even as the group of users grows, the test continues to be ordered for only a specific subgroup of patients.
How did Michigan succeed?
Dr. Korley offers several insights. “Clinicians are very data driven, so we did a lot of education presenting the data, both good and bad.” This allowed them to form a complete—nuanced, realistic—picture of the test’s utility.
“We also did audit and feedback,” he says. Clinicians had questions about the target population, naturally. “So as we observed usage of the test, I would send them emails calling out specific instances of where the test was used appropriately or inappropriately.”