Risk Score May Be Better ED Stroke Screen

Once again this is being looked at wrong, trying to make a subjective measure better, rather than figuring out an objective measurement like urine testing, blood testing, impedance spectroscopy.

The risk scoring here;
http://www.medpagetoday.com/Cardiology/Strokes/33106
Using a modified ABCD² risk score may be a better approach than the decades-old emergency department strategy of using subjective measures to assess dizzy patients for stroke, researchers suggested.
The modified ABCD² score -- which includes Age, Blood pressure, Clinical features, Diabetes, and Duration of symptoms -- was significantly higher for those presenting with dizziness who had cerebrovascular events than for those who did not have such events, reported Anthony S. Kim, MD, from the University of California in San Francisco, and colleagues.
Specifically, age over 60, blood pressure above 140/90 mmHg, and clinical features such as unilateral weakness or speech disturbance without weakness, but not diabetes were found significantly more often in those with cerebrovascular events, according to the study published in the June issue of Stroke: Journal of the American Heart Association.
The assessment of dizziness in the emergency department is currently based on a "symptom quality" approach that looks at vertigo, presyncope, disequilibrium, or nonspecific dizziness, a strategy which has "limited clinical utility in the ED setting," the researchers noted in their introduction.
The ABCD² score is well validated to predict stroke after transient ischemic attack (TIA), but it is unknown whether the application of this tool would work with those presenting with dizziness in the emergency department.
Dizziness is a common patient symptom, accounting for 3.3% of emergency department visits in the U.S. each year, and although it is usually attributable to benign etiologies such as peripheral vertigo, up to 5% of acute dizziness cases may be from cerebrovascular disease, they noted.
The modified ABCD² risk predictor, with a possible range of 0-7 points, is scored as follows:

• 1 point for age over 60
• 1 point for blood pressure over 140/90 mmHg
• 2 points for unilateral weakness, and 1 point for speech disturbance without weakness
• 1 point for diabetes
• 1 point for symptom duration of 10 to 59 minutes, and 2 points for symptom duration ≥60 minutes

The duration of symptoms was not readily available to investigators for this study, so they categorized all patients as having symptoms greater than an hour (2 points), "which effectively removes the contribution of this item to the overall score."
For the study, Kim and colleagues retrospectively identified 907 patients presenting to the UCSF emergency department complaining of dizziness, vertigo, or imbalance from 2007 to 2009.
In total, 4.1% of these patients had a cerebrovascular cause of dizziness: of those, 65% were ischemic strokes, 22% were TIAs, and 14% were intracerebral hemorrhages.
The mean age of those with a cerebrovascular event was 73, compared with 58 for those without such an event. Most of those with an event were men (75%) and were white (59%). Those with a cerebrovascular event had more risk factors than those without, and hypertension was the most prevalent risk factor (86%).
In both adjusted and unadjusted analysis, the individual components of the ABCD² except diabetes were significantly more common in those with cerebrovascular events.
Kim and colleagues reported that the c-statistic for the ABCD² score for predicting cerebrovascular event as the cause of dizziness was 0.79 (95% CI 0.73 to 0.85).
The median modified ABCD² score was 3. Five of 512 patients with an ABCD² score of 3 or less had a cerebrovascular event. However, 25 of 369 patients with a score of 4 or 5, and seven of 26 patients with a score of 6 or 7, all had cerebrovascular events.
The investigators concluded that the modified ABCD² score is effective at identifying those at low risk of cerebrovascular events who present with dizziness.
Given that there is room for clinical judgment, researchers said that their results "suggest that it may be possible to develop a risk score for dizziness to help streamline evaluations and target costly testing, consultation, and observation to the most appropriate patients."
The results may not be generalizable to other emergency departments, especially those that do not have 24/7 access to neruoimaging, researchers said. Also, they did not always know the specific features of dizziness, and follow-up data were not always complete.
Another limitation was the incomplete availability of duration of symptoms, which could have changed the performance of the model, Kim and colleagues said. Finally, they said the modified ABCD² risk score should be tested prospectively.