Georg M. Schuetz

Meta-analysis: Noninvasive Coronary Angiography Using Computed Tomography Versus Magnetic Resonance Imaging

BACKGROUND Two imaging techniques, multislice computed tomography (CT) and magnetic resonance imaging (MRI), have evolved for noninvasive coronary angiography. PURPOSE To compare CT and MRI for ruling out clinically significant coronary artery disease (CAD) in adults with suspected or known CAD. DATA SOURCES MEDLINE, EMBASE, and ISI Web of Science searches from inception through 2 June 2009 and bibliographies of reviews. STUDY SELECTION Prospective English- or German-language studies that compared CT or MRI with conventional coronary angiography in all patients and included sufficient data for compilation of 2 x 2 tables. DATA EXTRACTION 2 investigators independently extracted patient and study characteristics; differences were resolved by consensus. DATA SYNTHESIS 89 and 20 studies (comprising 7516 and 989 patients) assessed CT and MRI, respectively. Bivariate analysis of data yielded a mean sensitivity and specificity of 97.2% (95% CI, 96.2% to 98.0%) and 87.4% (CI, 84.5% to 89.8%) for CT and 87.1% (CI, 83.0% to 90.3%) and 70.3% (CI, 58.8% to 79.7%) for MRI. In studies that included only patients with suspected CAD, sensitivity and specificity of CT were 97.6% (CI, 96.1% to 98.5%) and 89.2% (CI, 86.0% to 91.8%). Covariate analysis yielded a significantly higher sensitivity for CT scanners with more than 16 rows (98.1% [CI, 97.0% to 99.0%]; P < 0.050) than for older-generation scanners (95.6% [CI, 94.0% to 97.0%]). Heart rates less than 60 beats/min during CT yielded significantly better values for sensitivity than did higher heart rates (P < 0.001). LIMITATIONS Few studies investigated coronary angiography with MRI. Only 5 studies were direct head-to-head comparisons of CT and MRI. Covariate analyses explained only part of the observed heterogeneity. CONCLUSION For ruling out CAD, CT is more accurate than MRI. Scanners with more than 16 rows improve sensitivity, as do slowed heart rates. PRIMARY FUNDING SOURCE None.

Use of 3x2 tables with an intention to diagnose approach to assess clinical performance of diagnostic tests: meta-analytical evaluation of coronary CT angiography studies.

Objective To determine whether a 3×2 table, using an intention to diagnose approach, is better than the “classic” 2×2 table at handling transparent reporting and non-evaluable results, when assessing the accuracy of a diagnostic test. Design Based on a systematic search for diagnostic accuracy studies of coronary computed tomography (CT) angiography, full texts of relevant studies were evaluated to determine whether they could calculate an alternative 3×2 table. To quantify an overall effect, we pooled diagnostic accuracy values according to a meta-analytical approach. Data sources Medline (via PubMed), Embase (via Ovid), and ISI Web of Science electronic databases. Eligibility criteria Prospective English or German language studies comparing coronary CT with conventional coronary angiography in all patients and providing sufficient data for a patient level analysis. Results 120 studies (10 287 patients) were eligible. Studies varied greatly in their approaches to handling non-evaluable findings. We found 26 studies (including 2298 patients) that allowed us to calculate both 2×2 tables and 3×2 tables. Using a bivariate random effects model, we compared the 2×2 table with the 3×2 table, and found significant differences for pooled sensitivity (98.2 (95% confidence interval 96.7 to 99.1) v 92.7 (88.5 to 95.3)), area under the curve (0.99 (0.98 to 1.00) v 0.93 (0.91 to 0.95)), positive likelihood ratio (9.1 (6.2 to 13.3) v 4.4 (3.3 to 6.0)), and negative likelihood ratio (0.02 (0.01 to 0.04) v 0.09 (0.06 to 0.15); (P<0.05)). Conclusion Parameters for diagnostic performance significantly decrease if non-evaluable results are included by a 3×2 table for analysis (intention to diagnose approach). This approach provides a more realistic picture of the clinical potential of diagnostic tests.

The Revised QUADAS-2 Tool

TO THE EDITOR: It was with great interest that we read the recent article by Whiting and colleagues (1) introducing the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies) tool, a revision of the tool from 2003 (2). It is important that such pivotal methodological work about diagnostic review research is made widely available through publication in Annals of Internal Medicine. We would like to discuss some issues related to QUADAS-2, as some of them attracted our attention when reading the otherwise impressive article. First, in the original QUADAS tool (2), items 1 and 2 related to variability with the potential to affect clinical generalizability (external validity) of study results. The term “applicability” is commonly used as a synonym (3). But, in QUADAS-2, “applicability” refers to whether certain aspects of an individual study are matching or not matching the review question (1). During a systematic review, the review question (according to Participants, Interventions, Comparisons, Outcomes, and Setting [PICOS] [4]) and eligibility criteria are prespecified before systematically searching for relevant studies (5). Therefore, all studies reaching the stage of quality assessment in a systematic review should be applicable to the review question because studies not matching the review question and eligibility criteria had already been sorted out during the selection process. The use of applicability in QUADAS as the “degree to which the results of a study can be applied to patients in practice” (2) may thus be preferable. Second, some aspects of QUADAS-2 seem to be more timeconsuming (that is, drawing a flow chart for each study). What is the average time needed to evaluate a study using QUADAS-2 versus QUADAS? Third, the examples of study assessment that would facilitate understanding of QUADAS-2 are not yet available from www.quadas .org. In addition, an explanation paper further describing use in detail would be very helpful for researchers conducting diagnostic accuracy meta-analyses. Is this planned? Lastly, calculating the interrater reliability only for agreement on the domain level when piloting the tool may not be appropriate to detect variability on the level of signaling questions, which, however, are used for the important decision on whether high or low risk of bias is present in individual studies.

Individual patient data meta-analysis for the clinical assessment of coronary computed tomography angiography: protocol of the Collaborative Meta-Analysis of Cardiac CT (CoMe-CCT)

BackgroundCoronary computed tomography angiography has become the foremost noninvasive imaging modality of the coronary arteries and is used as an alternative to the reference standard, conventional coronary angiography, for direct visualization and detection of coronary artery stenoses in patients with suspected coronary artery disease. Nevertheless, there is considerable debate regarding the optimal target population to maximize clinical performance and patient benefit. The most obvious indication for noninvasive coronary computed tomography angiography in patients with suspected coronary artery disease would be to reliably exclude significant stenosis and, thus, avoid unnecessary invasive conventional coronary angiography. To do this, a test should have, at clinically appropriate pretest likelihoods, minimal false-negative outcomes resulting in a high negative predictive value. However, little is known about the influence of patient characteristics on the clinical predictive values of coronary computed tomography angiography. Previous regular systematic reviews and meta-analyses had to rely on limited summary patient cohort data offered by primary studies. Performing an individual patient data meta-analysis will enable a much more detailed and powerful analysis and thus increase representativeness and generalizability of the results. The individual patient data meta-analysis is registered with the PROSPERO database (CoMe-CCT, CRD42012002780).Methods/DesignThe analysis will include individual patient data from published and unpublished prospective diagnostic accuracy studies comparing coronary computed tomography angiography with conventional coronary angiography. These studies will be identified performing a systematic search in several electronic databases. Corresponding authors will be contacted and asked to provide obligatory and additional data. Risk factors, previous test results and symptoms of individual patients will be used to estimate the pretest likelihood of coronary artery disease. A bivariate random-effects model will be used to calculate pooled mean negative and positive predictive values as well as sensitivity and specificity. The primary outcome of interest will be positive and negative predictive values of coronary computed tomography angiography for the presence of coronary artery disease as a function of pretest likelihood of coronary artery disease, analyzed by meta-regression. As a secondary endpoint, factors that may influence the diagnostic performance and clinical value of computed tomography, such as heart rate and body mass index of patients, number of detector rows, and administration of beta blockade and nitroglycerin, will be investigated by integrating them as further covariates into the bivariate random-effects model.DiscussionThis collaborative individual patient data meta-analysis should provide answers to the pivotal question of which patients benefit most from noninvasive coronary computed tomography angiography and thus help to adequately select the right patients for this test.

Extracardiac findings on coronary CT angiography: A systematic review

UNLABELLED Coronary computed tomography angiography (CCTA) is of growing importance in noninvasive diagnosis of coronary artery diseases. The CT data allow evaluation not only of coronary arteries but also of adjacent anatomical territories. Our objective was to review, to analyze, and to quantify the spectrum and the prevalence of extracardiac findings (ECF) in CCTA. Therefore, we searched MEDLINE, EMBASE, and ISI Web of Science. Prior to quantitative analysis, we categorized the ECF of all included studies into clinically significant and clinically non-significant findings. First, we calculated the average prevalences of ECF and clinically significant ECF performing a meta-analysis for proportions using the double arcsine transformation. Second, we analyzed the spectrum and location of clinically significant ECF. Third, we identified ECF of acutely life-threatening potential as well as malignancies and calculated their prevalences. Thirteen studies with a total of 11,703 patients were found to meet the inclusion criteria. The average prevalence of overall ECF was 41.0% (95% confidence interval [95% CI]: 27, 56; P < .0001) and 16.0% (95% CI (9, 24; P < .0001) for clinically significant ECF. Clinically significant ECF were most commonly detected in the lungs (50.2%), the abdomen (26.7%), the vessels (13.1%), the mediastinum (3.6%), and in other adjacent anatomical territories (6.4%). The prevalence of acutely life-threatening and malignant ECF accounted for 2.2% (95% CI: 1.9, 2.5; P < .0001) and 0.3% (95% CI: 0.2-0.4; P < .0001), respectively. In conclusion, clinically significant and acutely life-threatening ECF are common. Reading CCTA for ECF may lead to earlier detection of relevant disease. CONCLUSION Clinically significant and acutely life-threatening ECF are common. Reading CCTA for ECF may lead to earlier detection of relevant disease.

Noninvasive evaluation of global and regional left ventricular function using computed tomography and magnetic resonance imaging: a meta-analysis.

ObjectivesTo compare the diagnostic accuracy of computed tomography (CT) in the assessment of global and regional left ventricular (LV) function with magnetic resonance imaging (MRI).MethodsMEDLINE, EMBASE and ISI Web of Science were systematically reviewed. Evaluation included: ejection fraction (EF), end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV) and left ventricular mass (LVM). Differences between modalities were analysed using limits of agreement (LoA). Publication bias was measured by Egger’s regression test. Heterogeneity was evaluated using Cochran’s Q test and Higgins I2 statistic. In the presence of heterogeneity the DerSimonian-Laird method was used for estimation of heterogeneity variance.ResultsFifty-three studies including 1,814 patients were identified. The mean difference between CT and MRI was -0.56 % (LoA, -11.6–10.5 %) for EF, 2.62 ml (-34.1–39.3 ml) for EDV and 1.61 ml (-22.4–25.7 ml) for ESV, 3.21 ml (-21.8–28.3 ml) for SV and 0.13 g (-28.2–28.4 g) for LVM. CT detected wall motion abnormalities on a per-segment basis with 90 % sensitivity and 97 % specificity.ConclusionsCT is accurate for assessing global LV function parameters but the limits of agreement versus MRI are moderately wide, while wall motion deficits are detected with high accuracy.Key Points• CT helps to assess patients with coronary artery disease (CAD).• MRI is the reference standard for evaluation of left ventricular function.• CT provides accurate assessment of global left ventricular function.

Applicability and accuracy of pretest probability calculations implemented in the NICE clinical guideline for decision making about imaging in patients with chest pain of recent onset

ObjectivesTo analyse the implementation, applicability and accuracy of the pretest probability calculation provided by NICE clinical guideline 95 for decision making about imaging in patients with chest pain of recent onset.MethodsThe definitions for pretest probability calculation in the original Duke clinical score and the NICE guideline were compared. We also calculated the agreement and disagreement in pretest probability and the resulting imaging and management groups based on individual patient data from the Collaborative Meta-Analysis of Cardiac CT (CoMe-CCT).Results4,673 individual patient data from the CoMe-CCT Consortium were analysed. Major differences in definitions in the Duke clinical score and NICE guideline were found for the predictors age and number of risk factors. Pretest probability calculation using guideline criteria was only possible for 30.8 % (1,439/4,673) of patients despite availability of all required data due to ambiguity in guideline definitions for risk factors and age groups. Agreement regarding patient management groups was found in only 70 % (366/523) of patients in whom pretest probability calculation was possible according to both models.ConclusionsOur results suggest that pretest probability calculation for clinical decision making about cardiac imaging as implemented in the NICE clinical guideline for patients has relevant limitations.Key Points• Duke clinical score is not implemented correctly in NICE guideline 95.• Pretest probability assessment in NICE guideline 95 is impossible for most patients.• Improved clinical decision making requires accurate pretest probability calculation.• These refinements are essential for appropriate use of cardiac CT.