In this era of cost control and limited resources available to meet the growing demand for cardiovascular imaging testing, quality control initiatives that reduce the need for alternative testing are highly relevant. The implementation of continuous quality improvement (CQI) through the analysis of variations in clinical or test performance at an organizational or individual level is termed ‘clinical outcomes research’ (COR). At the most recent American Heart Association Scientific Sessions held November 12-16, 2011, in Orlando, Florida, scientists stressed the rapidly growing importance of this broad field of study, which, in addition to CQI, includes assessment of the appropriateness of therapy and/or testing and analysis of the clinical effectiveness of therapy and/or testing.
Implementing CQI in the imaging lab can be extremely challenging, due to the high-volume nature of testing and the fact that most labs employ large numbers of interpreters and sonographers who have varied levels of training and experience. Assessing quality at the several stages of testing simultaneously is impracticable for most laboratories. Targeting a particular imaging parameter for a quality control initiative is therefore a reasonable starting point.
Assessing left ventricular ejection fraction (LVEF) is one of the most important clinical investigations driving the use of cardiovascular imaging in medicine. Echocardiography is one of the most common and cost-effective methods to provide this. Its advantages over other methods include low cost, portability, safety, and ubiquity. However, these same factors can also contribute to a lowering of the study quality at a number of stages, including test ordering (inappropriate ordering), image acquisition (poor sonographer technical performance), clinical interpretation (inter-observer variability [IOV]), and report dissemination. It is our professional imperative to analyze lapses in quality at each stage and devise methods to remedy any issues in accordance with COR principles.
In this brief article, we discuss approaches to further enhance the quality of LVEF assessment in the echocardiography lab, including the use of contrast-enhanced imaging, of teaching interventions to reduce IOV, and of newer investigative methods such as 3D echocardiography (3DE) and speckle-tracking.
Contrast-enhanced Imaging
Intravenous injection of a microbubble contrast agent to opacify the left ventricle has been shown to markedly improve intra-observer variability and IOV in the assessment of LVEF.1 Galema et al. showed that, in a cohort of patients who had recently undergone primary angioplasty for the treatment of acute myocardial infarction, both intra-observer variability and IOV in the assessment of LVEF were significantly reduced after left ventricular (LV) opacification through the administration of a bolus of SonoVue® (Bracco, Milan, Italy) contrast agent.1 Comparison with the mean LVEF results of experienced cardiologists demonstrated that intra-observer variability had decreased from 12.5 ± 11.5 % to 7.0 ± 7.0 % (p<0.001) with opacification through administration of a contrast agent. Additionally, IOV decreased from 16.9 ± 9.9 % to 7.0 ± 6.2 % (p<0.001).
Due to the improved image quality offered by the high echogenicity of contrast agents, the US Food and Drug Administration (FDA) has approved their use, although currently only the microbubble contrast agents Definity® (Lantheus Medical Imaging, North Billerica, MA) and Optison™ (GE Healthcare, Princeton, NJ) have received approval—with some provisions and warnings. Regardless of these provisions, the Intersocietal Commission for the Accreditation of Echocardiography Laboratories (ICAEL) released a report in 2010 requiring the use of contrast agents in all patients with suboptimal image quality as part of its accreditation process.2
The concerns raised by the FDA about the use of contrast agents in acutely ill patients resulted in several studies assessing the safety of these agents. After the release of a report entitled ‘Acute mortality in hospitalized patients undergoing echocardiography with and without an ultrasound contrast agent: results in 18,671 consecutive studies’,3 which found no statistical difference in acute mortality within a population of over 18,500 patients undergoing echocardiography with and without Definity contrast agent enhancement, the FDA revised its previous black box warning on the use of contrast agents in acutely ill patients.
The approval of contrast agents has represented a major step toward improving echo imaging quality. In patients with known impairment of LV functioning, contrast enhancement may be an effective way to monitor changes in function, due to the increased reproducibility of results.4
The accuracy and reproducibility of the images obtained have become especially relevant, allowing to make therapeutic decisions based on the results of contrast-enhanced imaging in this patient population. The technique should therefore be considered as an important method of CQI that can be integrated into the imaging laboratory.
Because suboptimal image quality is a common issue during stress echocardiography, investigators have studied methods to improve cost-effectiveness using contrast enhancement. In one institution, during a one-year period, 2,594 patients underwent stress echocardiography. Of these 2,594 patients, 12 % had suboptimal baseline images and therefore required further testing using an additional modality, such as myocardial perfusion or single-photon emission computed tomography (SPECT) imaging.5 This prompted an investigation into the ability of contrast enhancement to improve image quality in these patients. The results of that investigation demonstrated that, in 267 of 277 (96 %) patients who received the intravenous contrast enhancement agent Optison during stress echocardiography, endocardial border definition at both rest and peak stress was significantly improved when compared with a control group in which patients did not receive contrast enhancement. Furthermore, 53 % of control patients required further testing due to suboptimal image quality, compared with only 3 % of patients who underwent contrast-enhanced stress testing. The authors calculated that the use of contrast enhancement under conditions of suboptimal baseline image quality allowed to save $238 per patient in the contrast enhancement group, because of the cost of additional testing for patients in the control group.
Laboratory Teaching Interventions
Periodic teaching interventions employed at an organizational level have been shown to significantly improve the quality (accuracy and precision) of LVEF assessment in a sustained manner. Investigators at the Massachusetts General Hospital echocardiography laboratory chose to study the reduction in IOV in the assessment of LVEF after implementation of a carefully devised training course for interpreters.6 In a group of 25 participants of various training and experience levels, IOV in the assessment of LVEF was reduced from ± 14 % ejection fraction (EF) before the training course to ± 8 % EF after the training course over a six-month period.
For the baseline assessment of IOV, participants reviewed 14 2D transthoracic echocardiograms and were asked to provide an estimation of LVEF. They were unaware of the reasons for referral and were blinded to the interpretations of their fellow participants. Each participant was assigned an anonymous identifier in order to track individual results relative to the group. Following the baseline assessment of IOV, the teaching intervention consisted of three one-hour sessions conducted over a three-month period. During the sessions, the cases were reviewed in a group format led by a facilitator. For each case, the estimated LVEFs were displayed using the anonymous identifiers, so that each participant could track their own performance relative to the group. This also allowed the investigators to determine whether there were interpreters who were outliers in the group. For each case, discussion points were centered around some of the common observations made during data collection (such as wall motion abnormalities, abnormal septal motion or foreshortening of the apex) and the effect of these observations on LVEF assessment. A series of reference cases in which LVEF had been derived by Simpson’s biplane method (quantitative method) was also reviewed and made available for self-directed learning in a digital format. Three months after the teaching intervention, participants assessed the LVEF in a new set of cases with similar pathologies and image quality. IOV was determined again and found to be reduced by 40 %. Furthermore, the effect was found to be sustained when reassessed in a subgroup of participants more than six months after the entire exercise.
Currently, there are no other studies looking at the use of teaching interventions to enhance LVEF assessment, but this one study lays the groundwork for further teaching interventions that may be employed by other laboratories or at a national level.
Newer Investigative Methods
Three-dimensional Echocardiography
One major benefit of 3DE is higher-quality images, regardless of contrast enhancement, when compared with conventional 2D methods. 2D methods are often limited by the acquisition of off-axis images, making the assessment of LV volume and EF more difficult—if not impossible. The 3DE method has become increasingly used in stress testing for the assessment of LV function, due to its rapidity of acquisition and ability to produce multiple views of the left ventricle. Administration of a chronotropic oral vasopressin agent such as dobutamine is used to increase the patient’s heart rate (up to 85 % of the maximal predicted heart rate) in order to detect the presence of wall motion abnormality as an indicator of myocardial ischemia.
When compared with conventional 2D stress echocardiography, realtime 3D peak dobutamine stress echocardiography (DSE) has recently demonstrated greater inter-observer agreement in the detection of ischemia.7 Additionally, the mean scanning time required for the assessment of patients was 27.4 ± 10.7 seconds with peak DSE, as opposed to 62.4 ± 20.1 seconds with the 2D method (p<0.0001).
In a recent investigation, 3DE was compared with multi-gated acquisition (MUGA) scanning for the assessment of EF.8 While MUGA is an established method of serial EF measurement, radiation exposure is a concern with this technique, especially in patients undergoing chemotherapy or presenting symptoms of a weakened immune system. Three investigators, blinded to one another’s results, interpreted a 3D echocardiogram immediately following a MUGA scan in 65 patients. Fifty of the 3D echocardiograms were of acceptable quality for image analysis, with Bland–Altman scatter plots demonstrating good agreement, both between 3DE EF and MUGA-scan EF and between observers. In this very small study, the 3D assessment of LVEF was comparable to the assessment using MUGA scanning. Further work is required to support the use of 3DE for LVEF assessment, especially in light of its ease of use and safety, compared with the cost and safety profile of nuclear imaging and magnetic resonance imaging (MRI). Currently, the use of contrast enhancement in the assessment of LVEF by 3DE is not supported by published data, which offers an avenue for further investigation.
Speckle-tracking
The speckle-tracking method for LVEF assessment has been shown to be rapid and reproducible: two important factors that should promote its future inclusion in the assessment of cardiac function.9 In this novel method, the operator first defines the endocardium at both sides of the matrix ring and the apex, and then uses a computer software algorithm to complete and track the endocardial contour throughout a complete cardiac cycle, with manual adjustment of the contour possible if necessary. Based on the difference in highest and lowest cavity volumes, corresponding to end-diastolic volume (EDV) and end-systolic volume (ESV), respectively, EF is calculated and displayed, along with the baseline characteristics used for its assessment.
The tracking takes slightly longer to perform (54 seconds) than a visual estimation (25 seconds), but is quicker than the modified Simpson’s biplane method (104 seconds). Importantly, IOV and intra-observer variability with the speckle-tracking method are significantly reduced, compared with visual estimation or Simpson’s method (using MRI as the gold standard). This semi-automated method holds promise, once its feasibility in various clinical contexts and study quality can be demonstrated.
Conclusion
LVEF assessment is a key clinical test guiding the diagnosis, management, and prognosis of a majority of cardiology patients. Echocardiography is a portable, safe, and inexpensive modality that is the most widely used method to provide EF assessment. Enhancing the quality of LVEF assessment is our professional imperative, and tools for CQI are available. Among them, contrast-enhanced echocardiography significantly improves image acquisition and has been shown to reduce both IOV and intra-observer variability; teaching interventions target the quality of interpretation at an organizational level and have been shown to significantly reduce IOV regardless of training level and experience; 3DE is an emerging technique that holds promise for the assessment of LVEF by enhancing image acquisition and avoiding the use of geometric assumptions; finally, newer techniques such as speckle-tracking are technical semi-automated advancements that have been demonstrated to be quick, feasible, and reproducible, with particular clinical advantages when reliable follow-up measurements of LVEF are needed. The implementation and documentation of any or all of these tools can be used to begin, or contribute to an existing, program of CQI in the echocardiography laboratory.