SAN DIEGO—There is great variability in the number and types of tests ordered when a patient presents with acute ischemic stroke, a survey of 58 North American stroke experts revealed. Although the reasons for the variability in test-ordering practices is not yet clear, the identification of those patients and tests that are most controversial may be important in directing clinical research and the development of practice guidelines.

"While I was being trained, I found it frustrating and very interesting that doctors had very different styles when it came to testing," said Timothy G. Lukovits, MD, a neurologist from Springfield, Massachusetts. Dr. Lukovits presented survey results at the 52nd Annual Meeting of the American Academy of Neurology. "I think this is a big problem area not only for the stroke neurologist, but for those who look to us for guidance."

TESTING 1-2-3

To assess the test-ordering practices of stroke experts, Dr. Lukovits and his colleagues devised three hypothetical cases in which a wide range of diagnostic tests might be utilized and had the experts indicate different imaging or blood tests they would perform on admission if presented with such a case. A test was considered controversial if less than two thirds of the experts agreed on ordering it.

Case One involved a previously healthy 36-year-old carpenter who developed sudden, severe left neck pain while lifting a heavy beam at work, followed by signs and symptoms of a vertebral artery dissection. "We found that there was a significant range in the number of tests 'performed'—both imaging and blood tests—despite it being very clear what the mechanism was," Dr. Lukovits said. The average total number of tests ordered was 5.3 (2.2 blood tests and 3.0 imaging/cardiac tests). Just over 12% of the respondents ordered transesophageal echocardiography (TEE), and 31% ordered a conventional angiogram. However, there was no consensus on whether to obtain magnetic resonance imaging (MRI) without contrast, magnetic resonance angiography (MRA) of the brain, erythrocyte sedimentation rates (ESR), or rapid plasma reagent (RPR).

Case Two described a previously healthy 40-year-old woman who developed fluent aphasia and mild right-sided weakness 24 hours before admission. "We also found a wide range of testing for this case, which I sort of expected. It's on the other end of the spectrum, compared to the other cases, in the sense of not knowing what was going on and the younger age of the patient," Dr. Lukovits noted. The average total number of tests ordered was 10.1 (5.9 blood tests and 4.2 imaging/cardiac tests). On the other hand, nearly 75% ordered a TEE, but only 19% a conventional angiogram. In addition, respondents did not agree on ordering MRI without contrast, MRA of the brain, MRA of the neck, carotid duplex, homocysteine, protein C/S, antithrombin III deficiency, or activated protein C resistance.

The third vignette involved a 65-year-old man with a history of hypertension who presented with sudden persistent numbness and paresthesia of the left face, arms, and legs starting five hours before admission. There was no history of heart disease for this pure sensory stroke. "We thought this case would not be very controversial; however, there was a lot of variability," Dr. Lukovits said. The average total number of tests ordered was 4.8 (1.8 blood tests and 3.0 imaging/cardiac tests). As with the first case, about 12% of the respondents ordered a TEE. In contrast, though, just under 2% ordered a conventional angiogram. In addition, there was no consensus on whether to obtain MRI without contrast, MRA of the brain, MRA of the neck, carotid duplex, ESR, or RPR.

A MATTER OF STYLE?

Overall, the least agreement among the three scenarios was found with MRI and MRA of the brain, as well as certain blood tests, including sedimentation rates and RPRs. In addition, there was no correlation found between a respondent's years in the American Academy of Neurology and the number of tests ordered.

However, "this study doesn't necessarily tell us what the sources of these variabilities are," Dr. Lukovits pointed out. Reasons for such discrepancies are unclear but may include differences in style (shotgun versus algorithmic), view of pathophysiology (interactionists versus reductionists), or an opinion about the relevance of specific tests in certain stroke syndromes. "I've noticed that there are some practitioners who try to initiate as many tests as possible in a short period of time so that the patient can leave the hospital. They then analyze all the abnormalities found after the patient is discharged. This is a shotgun approach," Dr. Lukovits said. On the other hand, an algorithmic approach targets "the most likely problem areas first. This is how I was trained. But sometimes this logical approach is much more expensive because it may prolong hospitalization. So, there are downsides to both approaches."

"We are far from agreeing on the appropriate workup for many stroke patients," Dr. Lukovits said, and added that he hopes that this study will stimulate discussion about this important issue.

—Bob Kronemyer

Return to table of contents