SGEM #8: ABCD2 Not as Simple as 1,2,3 (CVA, TIA)

Case Scenario:

68yo woman with unilateral weakness lasting 20 minutes. Blood pressure 165/95.

Question:

Reference:

Populations:

Control:

Outcome:

Results:

Authors Conclusions:

“This multicentre prospective study involving patients in emergency depart- ments with transient ischemic attack found the ABCD2 score to be inaccurate, at any cut- point, as a predictor of imminent stroke. Furthermore, the ABCD2 score of more than 2 that is recommended by the American Heart Association is nonspecific.”

BEEM Comments (Dr. Chris Carpenter):

 This prospective, ED-based multicenter Canadian trial represents the highest quality evidence to assess the ABCD₂ thus far.  Although an ABCD₂ score of 0 or 1 is 100% sensitive for 30-day stroke risk, very few ED patients with a suspected TIA fall within this low-risk category.  Therefore, the ABCD₂ is not an accurate post-TIA stroke risk prognostic instrument and should NOT be used to risk-stratify ED patients.   Instead, clinical gestalt and local resource availability should be used to determine the disposition and timing for the diagnostic work-up of post-TIA ED patients.

EBM Information:

• Sensitivity and Specificity
• Likelihood Ratio
• Evidence-Based Emergency Care: Diagnostic Testing and Clinical Decision Rules
• TIA Risk Stratification
• Further TIA Risk Stratification

BEEM Bottom Line:

ABCD2 score is not a reliable way to risk stratify patients presenting to the ED with TIAs on who will go on to have a stroke.