Serial ultrasonography did not predict stroke in asymptomatic carotid artery disease
ACP J Club. 1998 Jan-Feb;128:12. doi:10.7326/ACPJC-1998-128-1-012
Lewis RF, Abrahamowicz M, Côté R, Battista RN. Predictive power of duplex ultrasonography in asymptomatic carotid disease. Ann Intern Med. 1997 Jul 1;127:13-20.
To determine whether repeated ultrasonographic measurements of the carotid arteries are useful for predicting cerebrovascular events in asymptomatic carotid artery disease.
Analysis of the predictive ability of repeated ultrasonography in a combined natural history study and randomized controlled trial (Asymptomatic Cervical Bruit Study). Mean follow-up was 3.2 years.
5 tertiary health care centers in Quebec, Canada.
715 outpatients who were 23 to 91 years of age (mean age 65 y, 60% women) and had cervical bruits. Exclusion criteria were previous symptomatic cerebrovascular disease or carotid endarterectomy, valvular heart disease other than mitral valve prolapse, myocardial infarction in the past 3 months, atrial fibrillation, use of anticoagulant agents or aspirin or other antiplatelet agents, nonarterial cervical bruits, intolerance to aspirin, or life expectancy < 5 years.
Description of test and diagnostic standard
At the initial evaluation and at 6-month intervals, patients had duplex ultrasonography on the common and internal carotid arteries. Progression of maximum stenosis was defined as any change to a higher category, a progression to ≥ 50% stenosis, or a progression to ≥ 80% stenosis. Information on transient ischemic attacks (TIAs), strokes, and death was obtained from outpatient clinic visits and telephone interviews and confirmed through hospital records, death certificates, and autopsy reports by an adjudication committee blinded to the ultrasonographic results.
Main outcome measures
Sensitivity and specificity were estimated for progression as a predictor for TIA or stroke within 2 years.
Progression of stenosis to ≥ 80% increased the risk for cerebrovascular events and death. All definitions of progression gave low estimates of sensitivity (Table).
Repeated ultrasonographic measurements of the carotid arteries had limited usefulness in predicting cerebrovascular events and death in patients with asymptomatic carotid artery stenosis.
Sources of funding: National Health Research and Development Program and Natural Sciences and Engineering Research Council of Canada.
For article reprint: Dr. M. Abrahamowicz, Division of Clinical Epidemiology, The Montreal General Hospital, 1650 Cedar Avenue, Montreal, Québec H3G 1A4, Canada. FAX 514-934-8293.
Table. Estimated test characteristics for predicting transient ischemic attack or stroke within 2 years*
|Baseline stenosis measurements||Progression cut point||Sensitivity||Specificity||+LR||-LR|
|0 to 99%||Any change to a higher category||31%||80%||1.55||0.86|
|< 50%||≥ 50%||15%||92%||1.88||0.92|
|< 80%||≥ 80%||22%||92%||2.75||0.85|
|50% to 79%||≥ 80%||45%||80%||2.25||0.69|
*+LR = likelihood ratio for the presence of disease if the test is positive; -LR = likelihood ratio if the test is negative. Both calculated from data in article.
Lewis and colleagues provide valuable insights into the predictive power of duplex ultrasonography in asymptomatic carotid disease. The authors confirm previous observations that more severe stenosis at baseline is associated with substantially increased risk for cerebrovascular events (1). They extend these previous observations, however, and find that progression to stenosis of ≥ 80% was statistically associated with cerebrovascular events or death, whereas progression to only ≥ 50% was not.
Is the statistical association of progression to stenosis ≥ 80% compelling enough to warrant routine and repeated duplex ultrasonography in patients with asymptomatic stenosis? Probably not. First, progression to stenosis ≥ 80% occurred in only 8.9% of patients studied. Second, the predictive power of repeated measurements was not that strong. In fact, the presence of heart disease at enrollment was more predictive of a stroke outcome than was progression to ≥ 80% stenosis. Third, the benefit of asymptomatic carotid endarterectomy is controversial (2), calling into question any hypothesized favorable effect of repeated screening on patient outcomes.
Further, repeated measurement carries a substantial cost. In this study, 3664 duplex ultrasonographic examinations were done. Even if ongoing trials firmly establish the clinical benefits of asymptomatic carotid endarterectomy (3), the strategy of repeated ultrasonographic measurement will need careful and systematic economic assessment to allow better understanding of its added value to patients, providers, and payers.
Robert Holloway, MD
University of RochesterRochester, New York, USA