Diagnosis, not risk stratification, advised for TIA patients

By Eleanor McDermid

A meta-analysis suggests that the ABCD² score predicts short-term stroke risk in patients with transient ischemic attack (TIA), but has only minimal value in its recommended role as a risk stratification tool.

Researchers have begun to improve on the ABCD² by incorporating brain and cerebrovascular imaging.

But Jonathan Edlow (Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA), author of an editorial accompanying the paper in Neurology, says that although this approach "makes sense," it "renders the concept of risk stratification meaningless."

He says: "My view is that physicians who diagnose a TIA should not bother risk stratifying but should initiate an immediate diagnostic evaluation, administer antiplatelet agents (absent a contraindication), begin risk factor modifications, and educate the patient and family about stroke symptoms that should precipitate a 911 call.

"A patient thus treated is not going to benefit from better risk stratification."

The meta-analysis, by Velandai Srikanth (Monash University, Clayton, Australia) and colleagues, included 33 studies with 16,070 patients.

With ABCD² scores of 0-3 and 4-7 representing low and high risk, respectively, the pooled positive likelihood ratio (PLR) for stroke within 7 days was 1.43; no single study had a PLR of at least 10 and only four had a PLR between 5 and 10, and all of these used an ABCD² cutoff of 5 to denote high stroke risk. Assessment of the moderate heterogeneity for PLR indicated that the ABCD² score was more reliable in the hands of neurologists than emergency physicians.

The pooled negative likelihood ratio (NLR) was 0.40, with just three studies achieving a NLR below 0.1.

Use of the ABCD² score at the recommended cutoff of 3 points resulted in an absolute increase in stroke risk of just 0.8% if a patient had a baseline (pretest) stroke risk of 2.0%, and of 2.0% if the baseline probability was 5.0%.

Edlow concedes that sending all TIA patients for immediate diagnostic workup has cost implications, but asks what is actually saved by delaying evaluation of some patients for 48 hours, as per current guidelines. And he says: "Prevention of a single stroke will pay for quite a few TIA workups."

This is particularly the case when patients are seen in a specialist TIA clinic, rather than in hospital, notes Edlow, concluding: "In areas where this does not already exist, neurologists and emergency physicians should work together to create highly efficient, low-cost systems of care to evaluate TIA patients."

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.