Diagnostic Accuracy of 64-Slice Computed Tomography Angiography in Patients with Chest Pain vs. SPECT in the Assessment of Significant Coronary Artery Disease: A Systematic Review and Meta-analysis

Maryam Jamali; Rajabali Daroudi; Masih Tajdini; Ali Akbari Sari; Sajad Alaei; Sahar Salehi

doi:10.18502/htaa.v4i2.6230

Maryam Jamali Department of Health Economics and Management, School of Public Health Tehran University of Medical Sciences, Tehran, Iran
Rajabali Daroudi Department of Health Economics and Management, School of Public Health Tehran University of Medical Sciences, Tehran, Iran
Masih Tajdini Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran
Ali Akbari Sari Department of Health Economics and Management, School of Public Health Tehran University of Medical Sciences, Tehran, Iran
Sajad Alaei Department of Health Services Management, School of Health Management ad Information Sciences, Iran University of Medical Sciences, Tehran, Iran
Sahar Salehi Department of Health Economics and Management, School of Public Health Tehran University of Medical Sciences, Tehran, Iran

DOI: https://doi.org/10.18502/htaa.v4i2.6230

Keywords: Coronary Artery Disease; Computed Tomography Angiography; Single-photon Emission Computed Tomography (SPECT); Sensitivity; Specificity

Abstract

Context: This systematic review and meta-analysis intended to investigate the diagnostic accuracy of computed tomography angiography (CTA) in comparison with single-photon emission computed tomography (SPECT) for the diagnosis of coronary artery disease (CAD) in chest pain patients with no history of cardiovascular diseases (CVDs).

Methods: Invasive angiography was considered as the reference test with a stenosis threshold of ≥ 50%. Cochrane, Scopus, Science Direct, PubMed, and Embase databases were comprehensively searched from the time of inception of these databases to May 15, 2018. A manual search in Google Scholar, a reference review of the obtained studies, and a review of gray literature (including those presented in conferences and congresses) regarding diagnostic performances of CTA and SPECT techniques were performed independently by two researchers. A meta-analysis was performed to determine pooling estimates of sensitivity, specificity, diagnostic odds ratio, and positive as well as negative likelihood ratios in CTA and SPECT tests. According to the 2 × 2 contingency table of each study, at 0.95 confidence interval, the diagnostic accuracy of CTA and SPECT was meta-analyzed by pooling estimates of sensitivity, specificity, diagnostic odds ratio (DOR), and positive and negative likelihood ratios based on DerSimonian-Laird’s random-effects model. Heterogeneity was assessed by calculating I2. Analyses were performed using MetaDiSc version 1.4 and Stata version 11. The qualities of the selected studies were assessed independently by two researchers according to the quality assessment of diagnostic accuracy studies (QUADAS) questionnaire. Sensitivity analyses were performed by the Jackknife method. Publication bias was evaluated by Deeks’ funnel plot.

Results: Fourteen studies related to CTA (1206 individuals) and 15 related to SPECT (1638 individuals) were eligible for meta-analysis. The pooled sensitivity and the specificity of CTA for CAD diagnosis were 91% (95% CI, 88% - 94%) and 87% (95% CI, 84% - 98%), respectively. The pooled positive and negative likelihood ratios, the diagnostic odds ratio, and the area under the ROC curve for CTA were 7.93 (95% CI, 5.11 - 12.29), 0.1 (95% CI, 0.06 - 0.17), 95.71 (95% CI, 59.81 - 153.15), and 0.96, respectively. The pooled sensitivity and the specificity of SPECT for CAD diagnosis were 81% (95% CI, 79% - 83%) and 74% (95% CI, 71% - 78%), respectively. The pooled positive and negative likelihood ratios, the diagnostic odds ratio, and the area under the ROC curve for SPECT were 3.03 (95% CI, 2.34 - 3.91), 0.25 (95% CI, 0.21 - 0.30), 13.56 (95% CI, 10.60 - 12.34), and 0.86, respectively. According to the sensitivity analyses, the removal of any single study at a time did not change the effect size of the remaining studies. We observed symmetry in the Deeks’ funnel plot, indicating that there was ignorable publication bias for CTA and SPECT studies.

Conclusions: This study demonstrated that the diagnostic accuracies of CTA and SPECT tests lie in the ‘excellent’ and the ‘very good’ ranges, respectively. CTA is stronger evidence, than SPECT, to rule out CVDs in patients with low and intermediate risks of CAD with no history of cardiovascular diseases.