Description

Wong et al identified a prolonged QTc as a risk factor for cardiac death in patients with a history of stroke. Patients with a prolonged QTc may benefit from closer monitoring and more specific diagnostic testing. The authors are from the University of Dundee.


 

NOTE: As indicated in the next section, prolongation of the QTc may be seen in subarachnoid hemorrhage.

 

Time from stroke: median time was 12 months between stroke and first evaluation

 

Exclusions because of difficulty measuring QTc: atrial fibrillation, atrial flutter, bigeminy, paced rhythm, bundle branch block.

 

 

maximum QTc in 12 lead ECG

Cutoff for QTc in ms

Sensitivity

Specificity

>= 400 ms

100%

6%

>= 410 ms

100%

19%

>= 420 ms

100%

31%

>= 430 ms

90%

40%

>= 480 ms

35%

82%

from Table 3, page 379

 

 

QTc in Lead V6

Cutoff for QTc in ms

Sensitivity

Specificity

>= 400 ms

100%

23%

>= 410 ms

85%

39%

>= 420 ms

70%

53%

>= 430 ms

65%

64%

>= 480 ms

10%

94%

from Table 3, page 379

 

 

 

QTc in Lead III

Cutoff for QTc in ms

Sensitivity

Specificity

>= 400 ms

75%

47%

>= 410 ms

65%

62%

>= 420 ms

55%

73%

>= 430 ms

45%

86%

>= 480 ms

5%

98%

from Table 3, page 379

 

Interpretation:

• A QTc >= 430 ms for any of the 3 measures above gave the maximum Youden index.

 

Limitations:

• The threshold used to define a prolonged QTc in Table 1 is not stated (at least I could not locate it). The leads with significant p values were II, III, aVL, V4, V6 and max QTc.

• The sensitivity and specificity for the different cut-off points indicate a significant numbers of either false positives and/or false negatives. The cutoff of >= 480 ms for V6 mentioned in the abstract only has a sensitivity of 10% despite a respectable specificity of 94%.

• The areas under the ROC curves for max QTc, QTc in V6 and QTc in lead III was about 0.70, which is not terribly good.

• The authors appear to have used the test to identify patients who may benefit from more definitive tests.

 


To read more or access our algorithms and calculators, please log in or register.