Authors' objectives:

Considerable controversy exists over the appropriate use of imaging procedures to target stroke treatments, such as carotid endarterectomy (CEA) and anticoagulant therapy, to those most likely to benefit. This report discusses the effectiveness and cost-effectiveness of various imaging strategies for evaluating and managing new stroke patients: transthoracic echocardiography (TTE), transesophageal echocardiography (TEE), carotid ultrasound (CUS), magnetic resonance imaging (MRA), and cerebral angiography.

Authors' recommendations: Echocardiography Results and Conclusions: Available evidence is insufficient to allow conclusions regarding whether and to what degree most echocardiographically identifiable lesions are associated with increased risk of future stroke. Moreover, insufficient data exist regarding the efficacy of treatment for reducing the risk of future stroke associated with intracardiac thrombus or other lesions identifiable with echocardiography. Under current estimates of echocardiographic accuracy and the prevalence of intracardiac thrombus, testing all stroke patients with echocardiography likely results in false positives at least as often as true positives. Assuming that anticoagulation reduces the risk of recurrent stroke from intracardiac thrombus by 33% over one year, both TEE and TTE cost over $290,000 per quality-adjusted life year (QALY) saved at thrombus prevalences of 5% or below. Cost-effectiveness ratios dropped below $50,000 per QALY if the relative risk reduction with anticoagulation was 86 percent and the prevalence of thrombus at least 6%. More information is needed on the risk of recurrent stroke among those with potential sources of cardioembolism, and the efficacy of anticoagulation in reducing that risk. Carotid Imaging Results and Conclusions: The accuracy of CUS appears to vary substantially across centers. MRA may be more accurate than CUS, but few high-quality studies have addressed its accuracy. The combination of CUS and MRA has high reported sensitivity, but all relevant studies to date have been affected by verification bias and were of fair to poor methodological quality. In cost-effectiveness analyses varying sensitivities and specificities of noninvasive tests over a wide range, all testing strategies cost at least $250,000 per QALY when the prevalence of severe (70-99%) stenosis was assumed to be 15 percent. However, two testing strategies - initial CUS with angiographic confirmation and CEA for those with severe stenosis, and MRA with direct referral to CEA for those with severe stenosis - had cost-effectiveness ratios below $75,000 per QALY when the prevalence of severe stenosis increased above 20%, and below $50,000 per QALY as the prevalence exceeded 25-30%. High-quality assessments of CUS, MRA, and cerebral angiography are needed to better inform clinical decision-making about the appropriate use of these imaging strategies.

Authors' methods: Systematic review

Project Status: Completed

URL for project: http://www.ahrq.gov/clinic/tp/strokemantp.htm

Year Published: 2002

English language abstract: An English language summary is available

Publication Type: Not Assigned

Country: United States

Organisation Name: Agency for Healthcare Research and Quality

Contact Address: Center for Outcomes and Evidence Technology Assessment Program, 540 Gaither Road, Rockville, MD 20850, USA. Tel: +1 301 427 1610; Fax: +1 301 427 1639;

Contact Name: martin.erlichman@ahrq.hhs.gov

Contact Email: martin.erlichman@ahrq.hhs.gov

Copyright: Agency for Healthcare Research and Quality (AHRQ)