Article type
Year
Abstract
Background: The Cochrane Collaboration has decided to include systematic reviews of diagnostic test accuracy studies within its remit. We have completed a pilot review on the accuracy of magnetic resonance imaging (MRI) versus computed tomography (CT) for the diagnosis of acute stroke to test the new Cochrane methods for systematic reviews of diagnostic test accuracy.
Objectives: To compare MRI with CT for the detection of acute ischaemic stroke. To assess the accuracy of MRI for the detection of acute haemorrhagic stroke.
Methods: We searched major electronic databases and perused reference lists of included studies. Studies were selected if they compared MRI and/or CT versus a valid reference standard and if they provided sufficient data to construct a 2x2 table. Data extraction and quality assessment of selected studies were performed independently by three reviewers, two neurologists and a methodologist. Data were initially analysed by simple descriptive statistics (sensitivity and specificities, likelihood ratios) and subsequently meta-analysed using Summary Receiving Operating Characteristics (SROC) methodology. In particular three methods to meta-analyse data from diagnostic.
Results: Results will be available at the Cochrane Colloquium.
Conclusions: The findings of this pilot systematic review will inform the Cochrane Reviews of Diagnostic Test Accuracy Working Group on the practicality of performing systematic reviews of accuracy according to the new Cochrane methods, and will provide empirical evidence of the reliability of current methods for meta-analyses of diagnostic studies.
Objectives: To compare MRI with CT for the detection of acute ischaemic stroke. To assess the accuracy of MRI for the detection of acute haemorrhagic stroke.
Methods: We searched major electronic databases and perused reference lists of included studies. Studies were selected if they compared MRI and/or CT versus a valid reference standard and if they provided sufficient data to construct a 2x2 table. Data extraction and quality assessment of selected studies were performed independently by three reviewers, two neurologists and a methodologist. Data were initially analysed by simple descriptive statistics (sensitivity and specificities, likelihood ratios) and subsequently meta-analysed using Summary Receiving Operating Characteristics (SROC) methodology. In particular three methods to meta-analyse data from diagnostic.
Results: Results will be available at the Cochrane Colloquium.
Conclusions: The findings of this pilot systematic review will inform the Cochrane Reviews of Diagnostic Test Accuracy Working Group on the practicality of performing systematic reviews of accuracy according to the new Cochrane methods, and will provide empirical evidence of the reliability of current methods for meta-analyses of diagnostic studies.