A literature search identified articles on correlated audio-electric cardiography (CAC). For example, Marcus and colleagues reported on a prospective study of 90 patients undergoing elective left-sided catheterization who also underwent CAC, transthoracic echocardiography and electrocardiogram (EKG) within a four hour period. The main outcome measures were diagnostic test characteristics of the phonocardiographic recording using markers of the left ventricular (LV) function (i.e., left ventricular end diastolic pressure, left ventricular ejection fraction, and serum levels of the beta natriuretic peptide [BNP]) as a gold standard. The authors reported that diastolic heart sounds (i.e., third-heart sound or S3 and fourth-heart sound or S4) had poor sensitivity for LV dysfunction. However the phonocardiographic S3 was found to be specific for LV dysfunction, ranging from 87% to 92%, suggesting that if present, S3 can be useful to rule in a diagnosis of ventricular dysfunction.
Additionally, the search did not identify any published data regarding the diagnostic performance of acoustic heart sound recordings or CAC and its impact on patient management in clinical situations, specifically as an alternative or adjunct to physical exam and heart auscultation. It is suggested that CAC might primarily be used in emergency room settings as an aid in evaluating patients with acute chest pain. For example, myocardial infarction (MI) is frequently associated with S3 or S4 heart sounds. The acoustical recording and computer analysis might facilitate recognition of these heart sounds as well as recognition of serious heart murmurs.
This technology is not supported by evidence in the peer-reviewed medical literature that permits conclusions on the effect of acoustic heart sound recordings on health outcomes. Through fall 2004, no literature was found in MedLine or within any Internet web sites addressing clinical trials or studies for acoustic heart sound recordings, phonocardiogram with computer analysis, or CAC.
A MedLine search was done for period of 2005 through July 2007. No additional published studies were identified that would prompt reconsideration of the policy statement, which remains unchanged.
A search of peer reviewed literature from July 2007 through June 2009 identified several studies demonstrating improved clinical outcomes as a result of this technique. One study reported its potential usefulness in distinguishing supraventricular tachycardia (SVT) from ventricular tachycardia (VT). This study from Europe evaluated S1 heart sound intensity and variability in 17 episodes of VT and 22 episodes of SVT among 57 patients. Another study reported on results from this testing in 100 patients with LV dysfunction who were undergoing cardiac catheterization. The clinic evidence for this test is limited.
A large prospective study was done on 995 dyspneic patients, 40 years and older, median age of 63, not dialysis dependent, and presenting to the emergency room. The treating physicians, blinded to the CAC results, performed a visual analog scale to estimate the acute decompensated heart failure probability from 0% to 100%. This was repeated after the CAC was provided to the treating physician. The physician’s initial sensitivity, specificity, and accuracy for acute decompensated heart failure as a diagnosis were 89.0%, 58.2%, and 71.0%, respectively. Acoustic cardiography had sensitivity of 40.2%, specificity of 88.5%, and accuracy of 68%. The CAC did not improve the diagnostic accuracy for acute decompensated heart failure due to the low sensitivity.
Collins et al. has reported two studies. In 2006, the published report explains that an electronic S3 heart sound has a higher sensitivity than that obtained by auscultation (34% vs. 16%), with similar specificity in detecting patients with heart failure. In the 2008 published report, ECG, as the gold standard, was used to diagnosis and classify subtypes of left ventricular hypertrophy (LVH) of 352 patients with suspected heart failure. The diagnostic accuracy of CAC was compared to BNP, and the Cornell Voltage Criteria (CVC, a method of evaluating aspects of an ECG, using specified calculations). The battery of testing included a 12-lead ECG, CAC, BNP samplings, and echocardiography. According to the authors, the BNP combined with either CAC (95% confidence interval) or CVC (95% confidence interval) had the best outcome performance over all other models (CVC, BNP, or CAC) alone.
The impact of this testing on outcomes still remains uncertain. Therefore, the coverage position for this medical policy remains unchanged.
Heart Failure and Cardiac Resynchronization Therapy
The majority of studies of acoustic cardiography or CAC, identified in the peer reviewed literature search through July 2011, evaluated the technology in two general areas. First, it has been used as an aid for the diagnosis of systolic function, primarily using S3 heart sound detection, or S3 strength, compared to auscultation alone. Other studies have used acoustic cardiography for optimization of cardiac resynchronization therapy (CRT), utilizing the parameter of electromechanic activation time (EMAT) and comparing this method of optimization to Doppler echocardiography. These two categories of evidence are examined separately as follows.
1. In patients with suspected heart failure, does acoustic cardiography improve the ability to diagnose systolic dysfunction, compared with auscultation only?
Michaels et al. evaluated whether acoustic cardiography improved detection of S3 and S4 in 90 patients referred for angiography. A total of 35 subjects at various levels of clinical experience, from medical student to attending, listened to recordings of each patient’s heart sounds using auscultation alone, and then using both auscultation and acoustic cardiography. The gold standard for the presence or absence of heart sounds was the consensus of two experienced readers who were blinded to other aspects of the study.
There was improvement in the ability to detect S3 in each cohort of clinical training, with overall accuracy improving by between 2-18%. The improvement in accuracy was statistically significant for more experienced trainees but not for medical students. For example, using auscultation alone, residents detected an S3 correctly in 68% of patients. This improved to 85% when auscultation was combined with acoustic cardiography. For attending physicians, the accuracy of S3 detection was 72% with auscultation alone, and this was improved to 80% (p<0.01) with the addition of acoustic cardiography.
Maisel et al. evaluated the predictive ability of acoustic cardiography for acute heart failure in 995 patients greater than 40 years old who presented to the emergency department with dyspnea. The main parameter used was the strength of the S3 heart sound graded on a 0-10 scale. The gold standard for the diagnosis of acute heart failure was consensus by two cardiologists who were blinded to the results of acoustic cardiography. For the entire population, the S3 strength was predictive of acute heart failure in univariate analysis, but was not an independent predictor in multivariate analysis. For the subpopulation of patients who were labeled as ‘gray zone’ patients based on an intermediate level of BNP (100-499 pg/mL), the information from acoustic cardiography improved the diagnostic accuracy of acute heart failure from 47-69%. Another potentially problematic subgroup examined was obese patients (body mass index >30), in whom auscultation is often more difficult. In this population, the sensitivity of S3 detection improved from 14-28% with the addition of acoustic cardiography, but the specificity decreased from 99-88%.
A second study compared the diagnostic accuracy of acoustic cardiography with BNP in 433 patients who had results from acoustic cardiography, BNP, and echocardiography. Echocardiography was used as the gold standard to diagnosis of systolic dysfunction. When compared to BNP alone, acoustic cardiography was more accurate in diagnosing systolic dysfunction (area-under-the-curve [AUC] 0.88 vs. 0.67, respectively; p<0.0001). When confined to patients with BNP levels in the indeterminate range, acoustic cardiography also outperformed BNP in diagnosing systolic dysfunction (AUC 0.89 vs. 0.64, respectively; p<0.0001).
Thus, acoustic cardiography may improve the accuracy of detection of an S3 heart sound, although this finding has not been consistent in all subgroups examined. Acoustic cardiography has not been demonstrated to be an independent predictor of the diagnosis of acute heart failure when combined with other relevant clinical information. In order to demonstrate an incremental benefit in the diagnosis of heart failure, the improvement in diagnostic accuracy with and without acoustic cardiography must be in the context of the entire spectrum of clinical information collected routinely in the workup of a patient with suspected heart failure. For example, two studies report that acoustic cardiography improves the accuracy of heart failure diagnosis for patients with a “gray zone” BNP. However, a gray zone BNP does not necessarily mean the diagnosis of heart failure is uncertain when all clinical information is considered; therefore, this type of evidence is not sufficient to conclude that acoustic cardiography improves the diagnosis of heart failure.
2. In patients treated with a CRT device, does optimization of hemodynamic parameters by acoustic cardiography improve outcomes, compared to optimization by Doppler echocardiography?
Toggweiler et al. reported that optimization of CRT settings by EMAT resulted in improved measures of clinical and hemodynamic factors such as work capacity, maximum oxygen uptake, ejection fraction, and end-systolic volume. However, this study did not compare EMAT with Doppler echocardiography and thus does not offer relevant data on this question.
Zuber et al. reported the correlation of optimal atrioventricular (AV) and interventricular (VV) intervals, as determined by echocardiography and acoustic cardiography in 43 patients with a CRT device. There was a high correlation for the optimal AV delay intervals (r=0.86, p<0.001) and a moderate correlation for the VV delay intervals (r=0.58, p<0.05). These authors also reported that the test-retest reproducibility was higher for the EMAT method (r=0.91) than for echocardiography (r=0.35) and that the intraobserver variability was similar for EMAT versus echocardiography (9.9% vs. 8.5%, respectively).
In a similar study, Hasan et al. reported the correlation of acoustic cardiography and echocardiography for optimization of CRT in 22 subjects. The correlation between the overall values as determined by each method was high (r=0.90, p<0.001). In the majority of patients (77.3%), the values obtained from echocardiography and acoustic cardiography were within 20 msec of each other. The authors also reported that acoustic cardiography took less time to perform and was easier to interpret.
Taha et al. also evaluated the correlation of acoustic cardiography with echocardiography for optimization of CRT settings, using the parameter of S3 signal strength rather than EMAT. There was a high correlation between the two parameters for optimization of AV delay (r=0.86, p<0.001) and a somewhat lower correlation for optimization of VV delay (r=0.64, p<0.001). For VV delay, the optimal intervals were identical in 56% of patients, and for VV delay the optimal intervals were identical in 75% of patients.
Zuber et al. evaluated the agreement in optimal AV and VV values using a number of different optimization methods in 20 patients treated with a CRT device. The different methods included various sequencing of Doppler echocardiography and EMAT parameters. There was poor agreement between the different methods of optimization, and there was not one method that was clearly preferable to the others.
Thus, there is a high correlation between optimization of CRT settings by Doppler echocardiography and acoustic cardiography using EMAT values. EMAT may be simpler and easier to perform compared to Doppler echocardiography. However, it is extremely unlikely that clinical centers performing CRT optimization would not have expertise in performing echocardiography for this purpose. There is no evidence that health outcomes are improved when using acoustic cardiography for optimization compared to echocardiography.
Acoustic cardiography has been explored for other conditions.
Atrial Fibrillation: A Switzerland study studied 194 patients for detection of left ventricular (LV) dysfunction with or without atrial fibrillation. Patients underwent acoustic cardiography and cardiac catheterization, including measurement of angiographic ejection fraction (EF) and maximum LV. The authors reported acoustic cardiography detected systolic dysfunction with high specificity and moderate sensitivity, with similar results to EF. The literature is scant and more studies are needed.
Mitral Stenosis: Twenty seven patients underwent computerized acoustic cardiography (measuring the QRS onset and the first heart sound [QS1]) comparing with Doppler echocardiography and invasive hemodynamics prior to and after percutaneous transvenous mitral commissurotomy (opening of the narrowed mitral opening) to relieve mitral stenosis. The study was designed to evaluate mitral stenosis severity. By echocardiography, the mitral area increased from 0.82 to 1.50 cm2. The QS1 interval decreased from 101.7 to 93.2 ms. This small study confirmed acoustic cardiography could be used as an adjunct non-invasive diagnostic tool to assess mitral severity. Additional randomized controlled clinical trials were not located.
Morbid Obesity: In 2010, comparison of using audiocardiography to conventional auscultation by stethoscopes in 190 morbidly obese patients, with overall body mass indexes of 47.3. Of those with an S3 heart sound by audiocardiography (n=7), one had a history of coronary artery disease (CAD), none had a history of heart failure, and one had a LVEF. In contrast, of those (n=6) with an S3 by stethoscope, one had a history of CAD, two had histories of heart failure, and three had LVEF. There were 40 patients with S4 identified by acoustic cardiography, while there were 42 by stethoscope. S4 was heard with both methods in nine patients. The data suggested traditional cardiac auscultation using a stethoscope in a quiet room remains the gold standard.
Asymptomatic Patients: One hundred twenty eight asymptomatic patients were studied after wearing an ambulatory monitor with acoustic cardiography. The recordings spanned a mean duration of 14 hours, including sleep. Data was analyzed for the presence of S3 and S4 heart sounds and for systolic time intervals. S3 was more prevalent in those aged lesser than 40 years of age, and significantly more pronounced during sleep. For S4, the results were the opposite for the older than 40 group. The authors concluded that the nocturnal increase of S4 in the elderly (greater than 40 years of age) reflects diastolic impairment, likely a result of changes in diastolic filling patterns with increasing age. An S3 heart sound after the age of 40 is a relatively uncommon finding and therefore should be a specific sign of cardiac disease. Utilization of acoustic cardiography is possible for asymptomatic individuals.
Acoustic cardiography is a technique that integrates electric and acoustic information in order to enhance the ability to detect and characterize heart sounds. Published literature has evaluated the use of acoustic cardiography in two areas: 1) as an aid in the diagnosis of heart failure, and 2) for optimization of hemodynamic parameters in patients with a CRT device.
A number of published articles support that acoustic cardiography improves the detection of an S3 compared to auscultation alone. However, there is no evidence that acoustic cardiography contributes independent predictive information when added to a standard clinical workup for heart failure that includes physical exam findings, laboratory testing, and routine imaging studies.
When used to optimize CRT settings, several studies report that acoustic cardiography has a high correlation with Doppler echocardiography. No studies have demonstrated that acoustic cardiography is superior to echocardiography for this purpose, and therefore there is no evidence that acoustic cardiography improves outcomes when used for optimization of CRT therapy.
There is very little evidence demonstrating a benefit in clinical outcome. Moreover, the evidence remains primarily limited to short-term effects; the long-term durability of benefit has not yet been determined. As far as screening utilization for asymptomatic patients, there is insufficient evidence to determine if this technique provides any clinical utility and improvement of patient outcomes. Thus, acoustic cardiography is considered experimental, investigational and unproven when used for any condition.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.