Multi-slice computed tomography (CT) could provide similar measurements of left ventricular ejection fraction (LVEF) to magnetic resonance imaging and two-dimensional transthoracic echocardiography, and multi-slice CT was valid for the combined evaluation of LVEF and coronary artery disease. The first conclusion appears reasonable, but no data were presented to support the second conclusion.

To assess the accuracy of multi-slice computed tomography (CT) for the measurement of the left ventricular ejection fraction (LVEF), compared with cardiac magnetic resonance imaging (MRI) or two-dimensional transthoracic echocardiography (TTE).

PubMed, EMBASE and The Cochrane Library were searched to October 2011, without language restrictions, and the search terms were reported. In-press online available published articles were screened, and the bibliographies of retrieved articles were screened for additional studies. Only published studies were included.

Studies were eligible for inclusion if they measured the LVEF using multi-slice CT (64 slices or more) and compared this with the reference standard: MRI, TTE, or both. Included studies had to report the mean and standard deviation for the LVEF, as measured by multi-slice CT and the reference standard. Included studies had to perform multi-slice CT according to standard procedures, which required patients to be in sinus rhythm, without any arrhythmias, able to hold their breath for 10 to 12 seconds during scanning, and to have no allergy to contrast media and no renal dysfunction. Studies of patients not fulfilling these criteria were excluded.

In studies comparing multi-slice CT with MRI, the mean age of participants was 57.5 years (SD 12) and 77% were male. In studies comparing multi-slice CT with TTE, the mean age of participants was 58 years (SD 12.5) and 67% were male. The indications for imaging varied, including: known or suspected coronary artery disease; previous or suspected ischaemic heart disease; previous heart transplant; acute myocardial infarction; mitral regurgitation; aortic regurgitation; aortic stenosis; and cardiomyopathy. Most of the included studies used 64-slice CT and one third used dual-source CT.

Three reviewers independently assessed studies for inclusion and any disagreements were resolved by discussion.

The authors did not state that they assessed study quality.

Mean and standard deviation LVEF measurements were extracted for each test and study. Weighted mean differences in LVEF, with 95% confidence intervals, between multi-slice CT and the reference standard, were calculated for each study.

Three reviewers independently extracted the data and any disagreements were resolved by discussion.

Pooled estimates of the weighted mean difference and 95% confidence interval were calculated using a Mantel–Haenszel fixed-effect model, unless there was evidence of statistically significant between-study heterogeneity, where a DerSimonian and Laird random-effects model was used.

Between-study heterogeneity was assessed using Χ² and Ι².

The Bland-Altman method comparison was used to assess agreement between multi-slice CT and MRI or TTE measurements of LVEF. Agreement limits were reported as ±1.96 standard deviations and 95% confidence intervals were calculated.

Twenty-seven studies, with 1,791 patients, were included in the review.

Multi-slice CT versus MRI: Twelve studies, with 472 patients were included. There was no significant difference between LVEF measured by multi-slice CT and LVEF by MRI; this did not change when single- and dual-source multi-slice CT were considered separately. Bland-Altman analysis showed excellent agreement between multi-slice CT and MRI.

Multi-slice CT versus TTE: Fifteen studies, with 1,319 patients were included. There was no significant difference between LVEF measured by multi-slice CT and LVEF by TTE; this did not change when single- and dual-source multi-slice CT were considered separately. Bland-Altman analysis showed excellent agreement between multi-slice CT and TTE.

Multi-slice CT provided similarly accurate measurements of LVEF, compared with MRI or TTE, and it was a valid technique for the evaluation of LVEF and coronary artery disease together.

This review stated its research objective to be a comparison of the accuracy of multi-slice CT with two alternative reference standards. The studies included in the review assessed the agreement between multi-slice CT and the two alternatives in the measurement of LVEF. A range of sources was searched for relevant studies and no language restrictions were applied, decreasing the likelihood of missing relevant studies. Only published studies were included, which introduces the possibility of publication bias. Measures to minimise error and bias were applied throughout the review process, but the methodological quality of the studies was not assessed. Some possible sources of bias and potential limits to generalisability were discussed by the authors, but it was not possible to assess the effects of any methodological limitations on the findings of the review. The meta-analysis methods were broadly appropriate.

The authors' conclusions on the similarity between LVEF measured by multi-slice CT and MRI or TTE reflected the data presented. But no data were presented to support their conclusion that multi-slice CT was a valid technique for the combined evaluation of LVEF and coronary artery disease. The conclusions of this review should be interpreted with caution.

Practice: The authors stated that, due to radiation and contrast risks, the available multi-slice CT techniques could not be recommended for the routine evaluation of LVEF in all patients. They stated that TTE was the first choice for the overall evaluation of cardiac structures and function in patients with suspected or known coronary artery disease or heart failure. Multi-slice CT could be used for the simultaneous evaluation of coronary artery disease and LVEF,particularly for patients with an equivocal TTE.

Research: The authors recommended further research on the radiation dose reduction strategies in multi-slice CT.

Not stated.