The prevalence of non-cardiac findings (NCFs) identified during routine diagnostic cardiac imaging with computed tomography (CT), magnetic resonance imaging (MRI), and nuclear perfusion scanning has been described to be as high as 23% to 43%. However, in spite of transthoracic echocardiography (TTE) being by far the most commonly used cardiac imaging modality worldwide, the prevalence of NCFs with TTE remains poorly defined. Khosa et al. set out to assess the prevalence and significance of NCFs on clinically indicated TTEs from a single institution.
The study design was a retrospective review of an electronic TTE database consisting of all complete TTEs, including both inpatients and outpatients, obtained over a one calendar month time period. The authors elected to focus on subcostal TTE images that consisted of the 4-chamber view, color flow Doppler across the interatrial septum, the mitral and tricuspid valves, and inferior vena cava (IVC) flow during quiet respiration followed by the sniff maneuver to assess right atrial pressure. In addition, limited views of the suprarenal abdominal aorta were included. The rationale for confining the images to the subcostal view was that it provides the best TTE imaging planes for visualization of extracardiac organs. Nonetheless, in a subset of patients, parasternal and apical views were examined, as well to investigate whether these views uncovered other NCFs.
All subcostal TTE digital clips were reviewed for the presence of NCFs by an experienced radiologist with expertise in both cardiac and abdominal imaging. Identified NCFs were confirmed by a second experienced radiologist. NCFs were classified as benign, indeterminate or worrisome. Benign NCFs were not expected to alter patient management or require follow up; indeterminate NCFs may require some form of follow-up, and worrisome NCFs could have a major impact on patient management. Electronic medical records including discharge summaries, outpatient records, and prior imaging studies were retrieved for all indeterminate and worrisome NCFs and examined by a reviewer blinded to the NCFs in order to establish whether these NCFs were previously known or unknown. In addition, a downstream analysis was performed for indeterminate and worrisome NCFs discovered on TTE review but not reported on previous imaging studies.
The study involved 922 patients; more than half (61%) were outpatients and 39% were inpatients; a total of 77 NCFs were detected. Not surprisingly, NCFs were more frequent in inpatients who were likely sicker than routine outpatients. Twenty-six percent of all NCFs were benign, 67% were indeterminate and 7% were worrisome. Of the total NCFs, two-thirds were already known, having been previously recognized by other imaging modalities and documented in the electronic medical records. A similar proportion of the more concerning indeterminate and worrisome NCFs had been previously documented. The majority of benign NCFs were simple hepatic cysts, while approximately half of the indeterminate and worrisome NCFs consisted of pleural effusions and ascites. The majority of these latter patients had a referral diagnosis of heart failure.
There were comparatively few NCFs overall. Small prior TTE studies that focused on NCFs in pre-operative or intensive care patients showed a similarly low prevalence of NCFs. However, there was a consistent major shortfall in the detection of NCFs by TTE as compared to MRI and CT. The magnitude of these discrepancies is difficult to reconcile but is most likely related to the standardized cross-sectional nature of CT and MRI with much more extensive systematic tomographic imaging of extracardiac structures. Furthermore, pulmonary pathology that is so easily visualized by CT and MRI is not demonstrable with TTE because of ultrasound attenuation in lung tissue.
Importantly, in their review of a randomly selected subset of 300 patients for detection of additional NCFs from the parasternal long axis, apical 4-chamber and apical two chamber views, the authors identified an additional 2%, none of which seriously altered patient management.
A particular strength of this study was the care taken in image review. Other studies examining NCFs using TTE relied on analysis of reports and did not review the actual images. This aspect, in addition to the fact that radiologists read the images in the current study, led to detection of a substantial number of NCF. However, the question arises as to whether all of the NCFs found were actually “noncardiac.” Two of the worrisome findings were in the proximal portion of the IVC—arguably a structure that is described in standard reports of routine echocardiograms. If abnormalities of the IVC are “noncardiac”, does the same apply to dissections, intramural hematoma, enlargement, aneurysms, or plaque noted in the portions of the aorta that are visualized in standard echocardiographic imaging planes?
The image review also allowed the estimation of a “miss” rate of 78% by the echocardiographers who generated the official reports based on the images. While potentially disturbing, this statistic is not particularly surprising. There is little emphasis placed on the detection of NCFs by TTE since TTE is acknowledged to be inferior to CT and MRI for extracardiac pathology. Most echocardiographers restrict interpretation of TTEs to the heart and great vessels and are probably uncomfortable recognizing or reporting NCFs, even if something looks “funny.” There is no ASE recommendation in consensus statements or guidelines for “standard of care” regarding the reporting of NCFs, and there have been no educational sessions on NCFs at ASE meetings (personal correspondence with Andrea M. Van Hoever, Vice President of Research, American Society of Echocardiography, March 1, 2012). Alternatively, the failure to record NCFs may be due to an unintended oversight or a communication breakdown with sonographers who imaged the findings.
NCF Detection: A New Vista in Echocardiography?
Thus Khosa et al. uncovered two important aspects of NCF detection by TTE: (1) worrisome NCFs are uncommon and often confirm prior diagnosis, and (if this study is representative) NCFs do not seriously impact patient management; and (2) less than one-quarter of all NCFs are recognized or reported by echocardiographers.
The question arises whether sonographers and echocardiographers should pay increased attention to extracardiac structures in order to avoid “missing” NCFs. On the one hand, it may not require much effort for sonographers. Though only 2% additional NCFs were discovered in a review of imaging planes other than the subcostal, the more common NCFs involved the pleural space. It is not excessively time-consuming for sonographers to increase the depth in the left parasternal view to check the posterior pleural space where effusions are usually easily identifiable. The standard subcostal view usually includes images of ascites and near-field liver pathology. On the other hand, it is not always easy to size pleural effusions by traditional echocardiographic views. Right pleural effusions are far field in the subcostal view and may be difficult to detect, much less quantify. It is difficult to diagnose the etiology of anterior echolucent spaces, although Cardello et al. , suggested that identification of the falciform ligament on TTE enables reliable diagnosis of ascites. Most of the liver is not well seen, and in our experience, it is not routinely scanned systematically for NCFs.
Even though increased attention to NCFs on the part of sonographers may not be onerous, the fact that so few previously unknown NCFs were found in this study suggest that the extra work may not be worth the effort. The extra work involved would include training to recognize pathology and distinguish it from artifact. Should the ASE provide educational sessions (e.g. “Imaging Non Cardiac Findings” or “Recognizing and Reporting Non cardiac Findings”)? Should JASE provide online continuing medical education training or computer tutorials on the topics of NCFs? Should the National Board of Echocardiography (and review courses for the examinations of special competence in echocardiography) include recognition of these findings in its official content syllabus? Individual echocardiographers would have to expend extra effort looking for NCFs and include a description in the formal echo reports.
On the other hand, the fact that NCFs are found in standard echocardiographic images raises the question whether ignoring them in a report could be construed as negligence and grounds for medical malpractice litigation, should a patient be harmed by a late or missed diagnosis. At present, the recognition of NCFs probably cannot be considered to be within the scope of practice for echocardiographers. However, there is arguably a moral/ethical duty to seize every opportunity to make a potentially important diagnosis. Granted this duty exists, does it apply to NCFs and to which NCFs? Is it discharged by simply reporting that something appears to be amiss, and/or does it necessitate appropriate additional training to diagnose NCFs? Does it involve recommending a definitive imaging modality for clarification of a NCF? Presuming not everyone has to be crosstrained in abdominal radiology (and that radiologists will not be over-reading the subcostal views of all echocardiograms), should there be a standard created for recognizing the pathological processes which may show up in routine echocardiographic views? Should this study by Khosa et al. , be used as a guide to what that standard should be?
In an era of cost-consciousness, duty may also encompass a fiscal responsibility to minimize the costs of further diagnostic testing and the costs of educating sonographers and echocardiographers. The authors of this study were unable to assess the rate of “overcalls” by echocardiographers, since they only looked at the electronic medical records of patients with NCFs. The chances are that overcalls would have been few and far between, but they might have generated extra costs in follow up imaging to investigate NCFs that are unimportant or artifactual. It is conceivable that potentially harmful interventions could also unnecessarily follow on the heels of such a diagnosis, let alone patient anxiety. Future studies should track the overcall rate in order to determine the “real world” specificity and positive predictive value, because the costs (financial and otherwise) of unnecessary follow up testing may be substantial.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
