CT coronary angiography vs. invasive coronary angiography in CHD

SCIENTIFIC BACKGROUND Various diagnostic tests including conventional invasive coronary angiography and non-invasive computed tomography (CT) coronary angiography are used in the diagnosis of coronary heart disease (CHD). RESEARCH QUESTIONS The present report aims to evaluate the clinical efficacy, diagnostic accuracy, prognostic value cost-effectiveness as well as the ethical, social and legal implications of CT coronary angiography versus invasive coronary angiography in the diagnosis of CHD. METHODS A systematic literature search was conducted in electronic data bases (MEDLINE, EMBASE etc.) in October 2010 and was completed with a manual search. The literature search was restricted to articles published from 2006 in German or English. Two independent reviewers were involved in the selection of the relevant publications. The medical evaluation was based on systematic reviews of diagnostic studies with invasive coronary angiography as the reference standard and on diagnostic studies with intracoronary pressure measurement as the reference standard. Study results were combined in a meta-analysis with 95 % confidence intervals (CI). Additionally, data on radiation doses from current non-systematic reviews were taken into account. A health economic evaluation was performed by modelling from the social perspective with clinical assumptions derived from the meta-analysis and economic assumptions derived from contemporary German sources. Data on special indications (bypass or in-stent-restenosis) were not included in the evaluation. Only data obtained using CT scanners with at least 64 slices were considered. RESULTS No studies were found regarding the clinical efficacy or prognostic value of CT coronary angiography versus conventional invasive coronary angiography in the diagnosis of CHD. Overall, 15 systematic reviews with data from 44 diagnostic studies using invasive coronary angiography as the reference standard (identification of obstructive stenoses) and two diagnostic studies using intracoronary pressure measurement as the reference standard (identification of functionally relevant stenoses) were included in the medical evaluation. Meta-analysis of the nine studies of higher methodological quality showed that, CT coronary angiography with invasive coronary angiography as the reference standard, had a sensitivity of 96 % (95 % CI: 93 % to 98 %), specificity of 86 % (95 % CI: 83 % to 89 %), positive likelihood ratio of 6.38 (95 % CI: 5.18 to 7.87) and negative likelihood ratio of 0.06 (95 % CI: 0.03 to 0.10). However, due to non-diagnostic CT images approximately 3.6 % of the examined patients required a subsequent invasive coronary angiography. Using intracoronary pressure measurement as the reference standard, CT coronary angiography compared to invasive coronary angiography had a sensitivity of 80 % (95 % CI: 61 % to 92 %) versus 67 % (95 % CI: 51 % to 78 %), a specificity of 67 % (95 % CI: 47 % to 83 %) versus 75 % (95 % CI: 60 % to 86 %), an average positive likelihood ratio of 2.3 versus 2.6, and an average negative likelihood ratio 0.3 versus 0.4, respectively. Compared to invasive coronary angiography, the average effective radiation dose of CT coronary angiography was higher with retrospective electrocardiogram (ECG) gating and relatively similar with prospective ECG gating. The health economic model using invasive coronary angiography as the reference standard showed that at a pretest probability of CHD of 50 % or lower, CT coronary angiography resulted in lower cost per patient with true positive diagnosis. At a pretest probability of CHD of 70 % or higher, invasive coronary angiography was associated with lower cost per patient with true positive diagnosis. Using intracoronary pressure measurement as the reference standard, both types of coronary angiographies resulted in substantially higher cost per patient with true positive diagnosis. Two publications dealing explicitly with ethical aspects were identified. The first addressed ethical aspects regarding the principles of beneficence, autonomy and justice, and the second addressed those regarding radiation exposition, especially when used within studies. DISCUSSION The discriminatory power of CT coronary angiography to identify patients with obstructive (above 50 %) coronary stenoses should be regarded as "high diagnostic evidence", to identify patients without coronary stenoses as "persuasive diagnostic evidence". The discriminatory power of both types of coronary angiography to identify patients with or without functionally relevant coronary stenoses should be regarded as "weak diagnostic evidence". It can be assumed that patients with a high pretest probability of CHD will need invasive coronary angiography and patients with a low pretest probability of CHD will not need subsequent revascularisation. Therefore, CT coronary angiography may be used before performing invasive coronary angiography in patients with an intermediate pretest probability of CHD. For identifying or excluding of obstructive coronary stenosis, CT coronary angiography was shown to be more cost-saving at a pretest probability of CHD of 50 % or lower, and invasive coronary angiography at a pretest probability of CHD of 70 % or higher. The use of both types of coronary angiography to identify or to exclude functionally relevant coronary stenoses should be regarded as highly cost-consuming. WITH REGARD TO ETHICAL, SOCIAL OR LEGAL ASPECTS, THE FOLLOWING POSSIBLE IMPLICATIONS WERE IDENTIFIED: under-provision or over-provision of health care, unnecessary complications, anxiety, social stigmatisation, restriction of self-determination, unequal access to health care, unfair resource distribution and legal disputes. CONCLUSION From a medical point of view, CT coronary angiography using scanners with at least 64 slices should be recommended as a test to rule out obstructive coronary stenoses in order to avoid inappropriate invasive coronary angiography in patients with an intermediate pretest probability of CHD. From a health economic point of view, this recommendation should be limited to patients with a pretest probability of CHD of 50 % or lower. From a medical and health economic point of view, neither CT coronary angiography using scanners with at least 64 slices nor invasive coronary angiography may be recommended as a single diagnostic test for identifying or ruling out functionally relevant coronary stenoses. To minimise any potential negative ethical, social and legal implications, the general ethical and moral principles of benefit, autonomy and justice should be considered.