This review concluded that coronary computed tomography angiography appears to have high accuracy for the diagnosis of coronary artery disease, but its effects on patient management and outcomes were unclear. These conclusions reflect the data presented, but should be interpreted cautiously. Limitations in the literature searches mean that relevant studies, particularly in relation to patient outcomes, may have been missed.

To evaluate the diagnostic accuracy and effects on clinical decision making of coronary computed tomography (CT) angiography in patients with suspected coronary artery disease.

MEDLINE, EMBASE and The Cochrane library were searched for studies in English from January 2005 (the first year of published studies of 64-slice scanners) to February 2010. The search strategy was reported and included methodological terms for test accuracy studies. Health technology assessments and systematic reviews were also sought. The bibliographies of included studies were screened for additional articles. Authors were contacted to identify any additional unpublished studies and to check possible overlap between publications.

Test accuracy studies and studies that assessed the effect of 64-slice or higher resolution coronary CT angiography (alone or in combination with other tests) on clinical decision-making and/or rates of major cardiovascular events, were eligible for inclusion.

Test accuracy studies were required to: use invasive catheter angiography as the reference standard in all, or a random sample, of participants; evaluate accuracy in native coronary arteries only; interpret coronary CT angiography and invasive catheter angiography blind to the results of the other test; conduct coronary CT angiography and invasive catheter angiography within three months of each other; report per patient estimates of test performance.

Studies of patient outcomes were limited to those that only included patients with low-to intermediate pre-test probability of coronary artery disease (10-30%).

Most (over 80%) of the included studies were conducted in an out-patient setting and the remainder were conducted in emergency departments. Most studies were of patients already scheduled for invasive catheter angiography and hence the prevalence of coronary artery disease was relatively high (15 to 91%). Approximately two thirds of studies excluded patients with a history of coronary artery disease or previous revascularisation and many excluded small or heavily calcified vessels form their analyses. The mean age of study participants ranged from 46 to 69 years and 63% were male. Details of the CT angiography techniques used in individual studies were not reported.

The methodological quality of included diagnostic accuracy studies was assessed using a modified version of the QUADAS tool. Items relating to reporting quality (adequate description of the index test and reference standard) were removed. Items considered particularly relevant to coronary CT angiography were added: use of an established threshold for a positive test; presentation of data on inter-observer variation with acceptable results; appropriate patient subgroups; reporting of true disease prevalence on invasive catheter angiography (or data to derive this).

The authors did not state that they assessed the quality of patient outcomes studies. A single, trained reviewer conducted quality assessment.

For test accuracy studies, data were extracted to calculate sensitivity and specificity, with 95% confidence intervals (CIs). Data were analysed on the "intention-to-diagnose" principle (indeterminate or non-diagnostic results on CT angiography were assumed to be false positive). For studies of patient outcomes of coronary CT angiography, the principal findings of individual studies were extracted, as reported.

The authors did not state how many reviewers performed the data extraction.

Pooled estimates of sensitivity and specificity were calculated using a bivariate model. Assessment of between-study heterogeneity and meta-regression to investigate possible sources of heterogeneity were undertaken using a random-effects model; factors investigated included sample size, mean age, proportion male, coronary artery disease prevalence and proportion with a history of coronary artery disease.

Forty-two diagnostic accuracy studies (5,390 participants) and 11 patient outcome studies (1,483 participants) were included in the review. Most test accuracy studies were small (less than 100 participants), range 30 to 1,372. Spectrum bias (due to the high prevalence of coronary artery disease) was considered to be present in two thirds of the included test accuracy studies. Other frequently failed criteria were information on study withdrawals and observer variation.

Intention-to-diagnose analyses resulted in pooled estimates of sensitivity 98% (95% CI 96 to 99%), Ι²=59.8% and specificity 85% (95% CI 81 to 89%), Ι²=82,5%. When non-diagnostic coronary CT angiography results were excluded from the analysis, the pooled estimate of specificity was 88% (95% CI 85 to 91%) and sensitivity did not change.

There was no evidence of a threshold effect. Age was the only significant variable in the meta-regression. Subgroup analyses indicated that sensitivity was unaffected by age and specificity decreased with age (91% in studies with a mean participant age under 59 years and 77% in studies with a mean participant age over 62 years).

Of the 11 patient outcome studies, only two provided a comparison of coronary CT angiography with standard care; most were case series or retrospective cohort studies that compared major cardiovascular event rates in patients with positive and negative coronary CT angiography results. There was some indication (mainly from emergency department studies) that triage of low-risk patients using coronary CT angiography produced no serious adverse outcomes and was time-saving compared to usual care.

Coronary CT angiography appears to have high diagnostic accuracy in patients with suspected coronary artery disease, but its potential impact on clinical decision-making and patient outcomes was less well-understood, particularly in non-emergent settings.

The review addressed a clearly stated research objective, defined by appropriate inclusion criteria. Several sources were searched for relevant articles and attempts were made to identify unpublished studies. The restriction of searches to English language studies may have resulted in the omission of relevant data and raised the possibility of language bias. The inclusion of test accuracy terms in the search strategy was likely to reduce the sensitivity of searches and may also have resulted in the omission of relevant studies. The EMBASE search, as reported, may have resulted in the exclusion of clinical trials which reported patient outcomes, if such studies did not also report test accuracy. No measures to reduce error and/or bias in the review process were reported. The methodological quality of included studies was assessed and fully reported, but quality assessment results were not incorporated in the analyses/synthesis. No details of the coronary CT angiography technologies used in individual studies were reported or considered in the analyses. The meta-analytic methods used were appropriate.

The authors' conclusions reflected the data presented, but should be interpreted cautiously (particularly in relation to the lack of evidence on patient management and outcomes) given the limitations outlined.

Practice: The authors stated that, on current evidence, the role of coronary CT angiography remains unclear. Where data on cardiac function were needed, SPECT or other testing functional imaging techniques followed by invasive catheter angiography may be preferred. For patients with a high risk of coronary artery disease, immediate invasive catheter angiography without non-invasive testing was currently recommended. CT angiography may be useful, for lower risk patients, in excluding coronary artery disease as a cause of chest pain.

Research: The authors stated that further trials were needed to compare outcomes between patients evaluated with and without coronary CT angiography. In addition, the impact of coronary CT angiography on care and outcomes of patients with mild-to-moderate stenoses should be evaluated.

Part funded by the Washington state Health Technology Assessment Program.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.