- Email: [email protected]
Usefulness of Acoustic Cardiography to Resolve Ambiguous Values of B-Type Natriuretic Peptide Levels in Patients With Suspected Heart Failure
Usefulness of Acoustic Cardiography to Resolve Ambiguous Values of B-Type Natriuretic Peptide Levels in Patients With Suspected Heart Failure Michel Zuber, MDa,*, Peter Kipfer, MDa, and Christine H. Attenhofer Jost, MDb B-type natriuretic peptide (BNP) levels are helpful to diagnose left ventricular (LV) systolic and/or diastolic dysfunction. BNP levels that are only moderately increased have limited diagnostic ability, and an additional test to resolve this problem would be desirable. The hypothesis that acquiring combined electrocardiographic and electronic cardiac acoustical data can improve the detection of LV dysfunction in patients with nondiagnostic values of BNP was tested. Both BNP and combined 12-lead electrocardiograms with electronic heart sound (acoustic cardiographic) recordings were obtained from 164 outpatients referred for echocardiographic evaluation for suspected heart failure. Acoustic cardiographic parameters included the third heart sound (S3) and percentage of electromechanical activation time, measured as the interval from onset of the Q wave of the electrocardiogram to the first heart sound (S1) and expressed as a proportion of the cardiac cycle. Sixty-nine of 164 patients (42%) had BNP values in the “gray zone” of 100 to 500 pg/ml. Sensitivity and specificity for LV dysfunction of BNP in the gray zone were 55% and 75%, with a positive likelihood ratio of 2.3. The use of acoustic cardiographic parameters in these 69 patients increased sensitivity and specificity to 69% and 100%, with a corresponding positive likelihood ratio of 69. In conclusion, easily obtainable acoustic cardiographic data substantially improved the diagnostic evaluation of patients with nondiagnostic BNP values and therefore can increase the confidence with which physicians diagnose and treat LV dysfunction. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;100:866 – 869)
Systolic and/or diastolic left ventricular (LV) dysfunction may result in heart failure, a major cause of disability and death. Although early diagnosis and treatment result in better clinical outcomes and lower cost of care, up to 46% of patients with acute dyspnea caused by heart failure currently are misdiagnosed.1–7 Although echocardiography is useful for detecting underlying LV dysfunction, the procedure is expensive and often not readily available. Therefore, physicians have adopted other tests for heart failure, such as measurement of B-type natriuretic peptide (BNP). However, BNP values showed an inverse relation between sensitivity and specificity for heart failure, and values in the so-called “gray zone” from 100 to 500 pg/ml were especially problematic.8 The third heart sound (S3) was highly specific for heart failure in the appropriate clinical context, but its detection using auscultation, especially in noisy environments, was unreliable.9,10 Therefore, acoustic cardiography, a technique for recording and algorithmically interpreting simultaneous diagnostic digital 12lead electrocardiographic and acoustic data, was developed. In this study, we tested the hypothesis that acoustic cardiography
a
Outpatient Clinic for Cardiology and Internal Medicine, Othmarsingen, Frauenfeld; and bCardiovascular Center, Klinik im Park, Zurich, Switzerland. Manuscript received January 26, 2007; revised manuscript received and accepted April 6, 2007. *Corresponding author: Tel.: ⫹41-62-896-1333; fax: ⫹41-62-8961464. E-mail address: [email protected] (M. Zuber). 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.04.019
can augment the use of BNP in the detection of LV dysfunction. Methods We evaluated 164 ambulatory patients (111 men; mean age 64 years; range 19 to 88) referred for Doppler echocardiographic evaluation for possible heart failure. The local medical ethics commission approved the protocol for the study, and all patients gave written informed consent before enrollment. Within 1 hour of Doppler echocardiography, each patient had BNP and serum creatinine measured and acoustic cardiographic data recorded using the Audicor device (Inovise Medical, Inc., Portland, Oregon). After this evaluation, we selected patients who had BNP of 100 to 500 pg/ml as subjects for additional study. Patients with atrial fibrillation were excluded. Each patient underwent a complete 2-dimensional and Doppler echocardiographic examination according to guidelines of the American Society of Echocardiography. The investigators interpreted Doppler echocardiographic findings blinded to both BNP and acoustic cardiographic data. Doppler echocardiographic findings were used to identify patients with versus without increased LV filling pressures. LV ejection fraction was measured using Simpson’s rule, and systolic LV dysfunction was defined as LV ejection fraction ⱕ50%. Diastolic function was assessed using the LV filling pattern and tissue Doppler examination of the lateral mitral annulus (for E=), including E/A ratio, deceleration time of the E wave, E/E= ratio, and pulmonary venous flow pattern. Diastolic function was graded as norwww.AJConline.org
Heart Failure/Augmenting BNP With Heart Sounds
mal, impaired relaxation, pseudonormal, or restrictive. Significant diastolic dysfunction was defined as either a pseudonormal (19 patients) or restrictive (14 patients) filling pattern. The presence of increased LV filling pressure was determined using the method described by Ommen and Nishimura.11 BNP and creatinine were measured in each patient after 30 minutes in the supine position using the Biosite Triage test (Biosite, Inc., San Diego, California) (normal values ⬍100 pg/ml), and creatinine levels, using the enzymatic UV Roche Diagnostics Corp (Indianapolis, Indiana). Reflotron test (normal values ⬍97 mol/L for men and ⬍80 for women). Acoustic cardiography consists of recording and algorithmically interpreting simultaneous digital 12-lead electrocardiographic and acoustic, for example, acoustic cardiographic, data, including S3, using dual-purpose sensors placed in the V3 and V4 positions with the Audicor system. Acoustic cardiography requires inexpensive equipment, can be used in a wide variety of clinical environments, requires approximately the same time and effort as that needed to record a standard 12-lead electrocardiogram, and shows no significant variability among operators. A 10-second Audicor recording was obtained for each patient and analyzed using the computerized algorithm with measurements generated for S3 and various systolic parameters. The following parameters of LV function were evaluated both individually and in combination with each other. The system evaluated S3 using a computerized measurement of the intensity and persistence of acoustic energy of sounds that had the appropriate frequency and timing for S3 (S3 strength) occurring up to 130 ms after the second heart sound. The range of possible values of S3 strength was 0 to 10 units, and if the value was ⱖ5.0, the Audicor computerized algorithm declared that S3 was present. Percentage of electromechanical activation time (EMAT) is a continuous parameter that indicates the interval in milliseconds from the onset of the Q wave of the electrocardiogram to the mitral component of the first heart sound (S1) divided by the RR interval in milliseconds. EMAT measures the time required for ventricular contraction to generate sufficient force to close the mitral valve. It may be abnormally prolonged partly because of the decreased dP/dT associated with LV systolic dysfunction. Percentage of EMAT expresses the proportion of the total cardiac cycle that EMAT occupies. For the 69 patients with BNP in the gray zone, diagnostic performances of BNP alone and in combination with acoustic cardiographic parameters for detecting Doppler echocardiographic abnormalities compatible with systolic and diastolic LV dysfunction were calculated. For percentage of EMAT and the additive combination of S3 strength and percentage of EMAT, receiver-operating characteristic curves were analyzed to determine values that yielded the highest positive likelihood ratios for each diagnosis. Positive and negative likelihood ratios were calculated using the following formulas: Positive likelihood ratio ⫽ Sensitivity/(1 ⫺ Specificity), and Negative likelihood ratio ⫽ Specificity/(1 ⫺ Sensitivity). To avoid dividing by 0, the denominator of the formula for positive likelihood ratio
867
Table 1 Diagnostic performances of various ranges of B-type natriuretic peptide (BNP) levels for detecting any type of left ventricular dysfunction BNP (pg/ml)
Sensitivity
⬎100 100–500 ⬎500
62/93 (67%) 51/93 (55%) 11/93 (12%)
Specificity
53/71 (75%) 53/71 (75%) 70/71 (99%)
Positive Likelihood Ratio
Negative Likelihood Ratio
3 2 12
2 2 1
was assigned a value of 1.0 in cases in which specificity was 100%. Calculating positive and negative likelihood ratios using these formulas expressed these ratios on the same numerical scale and therefore facilitated the comparison of a test’s rule-in and rule-out power.12 A positive or a negative likelihood ratio ⬎10 shown by a diagnostic test was considered to have excellent rule-in or rule-out power for the condition being tested, respectively. Unlike positive and negative predictive values, diagnostic inferences based on positive and negative likelihood ratios were independent of the prevalence of the abnormality in the population being tested.12 Statistical analyses were performed using SPSS, version 13.0 (SPSS, Inc., Chicago, Illinois). The entire group of patients with LV dysfunction was analyzed. Separate analysis was performed of subgroups with systolic LV dysfunction with normal LV filling pressures, systolic LV dysfunction with increased LV filling pressures, and isolated diastolic LV dysfunction. Results Of 164 patients in the study, BNP was ⬍100 pg/ml in 83 patients, 100 to 500 in 69 patients (gray zone), and ⬎500 in 12 patients. Of 69 patients with BNP in the gray zone, 51 (74%) had LV dysfunction. Of these 51 patients, 25 (49%) had Doppler echocardiographic evidence of systolic LV dysfunction with normal filling pressures, 19 (37%) had systolic LV dysfunction with increased filling pressures, and 7 (14%) had isolated diastolic LV dysfunction. Of 7 patients with isolated diastolic LV dysfunction, 5 (71%) had a pseudonormal and 2 (29%) had a restrictive pattern of LV inflow by Doppler. Of 69 patients with BNP in the gray zone, Doppler echocardiography showed mildly delayed diastolic relaxation in 40 (58%) and 6 (9%) had no abnormality of diastolic function. Table 1 lists diagnostic performances of BNP for detecting LV dysfunction in the 3 different ranges of values of ⬎100, 100 to 500, and ⬎500 pg/ml. Table 1 lists the inverse relation between sensitivity and specificity shown by low versus high BNP. It also confirms that BNP in the gray zone of 100 to 500 pg/ml had neither good rule-in nor good rule-out power for LV dysfunction. As listed in Table 2, acoustic cardiographic parameters augmented values of BNP in the gray zone for detecting LV dysfunction. Both percentage of EMAT alone and in combination with S3 produced excellent rule-in power for LV dysfunction by providing very high positive likelihood ratios. Tables 3 to 5 list similar types of data for the 3 subgroups
868
The American Journal of Cardiology (www.AJConline.org)
Table 2 Using sound parameters to detect any type of left ventricular dysfunction in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⱖ5.0 ⱖ10.2 ⱖ14.4
Sensitivity 51/93 (55%) 20/51 (39%) 34/51 (67%) 35/51 (69%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
53/71 (75%) 17/18 (94%) 18/18 (100%) 18/18 (100%)
2 7 67 69
2 2 3 3
Table 3 Using sound parameters to detect systolic left ventricular dysfunction with normal filling pressures in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⬎5.0 ⬎12.0 ⬎15.9
Sensitivity 25/55 (45%) 5/25 (25%) 13/25 (52%) 14/25 (56%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
57/82 (70%) 22/25 (88%) 25/25 (100%) 25/25 (100%)
2 2 52 56
1 1 2 2
Table 4 Using sound parameters to detect systolic left ventricular dysfunction with elevated filling pressures in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⱖ5.0 ⱖ12.0 ⱖ15.9
Sensitivity 19/27 (70%) 12/19 (63%) 9/19 (47%) 13/19 (68%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
57/82 (70%) 22/25 (88%) 25/25 (100%) 25/25 (100%)
2 5 47 68
2 2 2 3.1
Table 5 Using sound parameters to detect isolated diastolic left ventricular dysfunction with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value
Sensitivity
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
100–500 ⬎5.0 ⬎12.0 ⬎15.9
7/11 (64%) 2/7 (29%) 1/7 (14%) 1/7 (14%)
83/129 (64%) 39/46 (85%) 33/46 (72%) 32/46 (70%)
2 2 1 1
2 4 7 1
of LV dysfunction. As listed in these tables, for each subgroup of LV dysfunction, BNP in the gray zone had poor diagnostic specificity. However, Tables 3 and 4 also show that regardless of level of filling pressures, using acoustic cardiographic parameters to elucidate the significance of BNP in this range provided excellent rule-in power for systolic LV dysfunction. In contrast, Table 5 shows that the diagnostic impact of acoustic cardiographic parameters is much lower in patients with isolated diastolic LV dysfunction than for those with systolic LV dysfunction. Tables 3 to 5 also show that threshold values of the acoustic cardiographic parameters required to obtain maximum positive likelihood ratios were similar for the 3 types of LV dysfunction. Discussion The findings of this study confirmed that the pathologic significance of BNP in the gray zone of 100 to 500 pg/ml
was uncertain. Importantly, 42% of our subjects with suspected heart failure had BNP values in this range, which indicated this is a commonly encountered problem. Our study showed the great diagnostic ability of detection of LV dysfunction by using S3 and percentage of EMAT as an accurate and fast tool for outpatients. The specificity and positive likelihood ratio for LV dysfunction shown by BNP in the gray zone were only 75% and 2.7, respectively. The specificity of BNP in this range was similar in each subgroup of LV dysfunction. Such a limited rule-in power would be unsuitable for a screening test because it would produce too many false-positive results. Using the more stringent BNP threshold of ⬎500 pg/ml produced very high specificity, but would identify only 12% of the truly positive cases of LV dysfunction. As expected, a 100-pg/ml increase in BNP was associated with decreased sensitivity and increased specificity.
Heart Failure/Augmenting BNP With Heart Sounds
In the diagnostic evaluation of dyspneic patients, BNP ⬍100 pg.ml is commonly believed to make the diagnosis of heart failure associated with LV dysfunction highly improbable.13 Conversely, in the absence of acute myocardial ischemia, renal insufficiency, or recent administration of neseritide, high BNP makes the diagnosis of heart failure likely. However, consistent with findings of the present study, BNP in the intermediate or gray zone is especially difficult to interpret. For example, in 452 patients of the B-Type Natriuretic Peptide for Acute Shortness of Breath Evaluation (BASEL) trial, it was considered necessary to use additional diagnostic information before making a definitive diagnosis of heart failure in patients with BNP in the range of 100 to 500 pg/ml.8 The present study showed that acoustic cardiographic parameters can effectively resolve the diagnostic uncertainty associated with BNP in the gray zone. The data showed that acoustic cardiographic parameters increased positive likelihood ratios for detecting LV dysfunction by a factor ⬎30. Our data showed that diagnostic performances of acoustic cardiographic parameters were best in patients with systolic LV dysfunction regardless of whether filling pressures were increased. Conversely, their diagnostic performances were much poorer in patients with isolated diastolic LV dysfunction. However, patients with isolated diastolic LV dysfunction represented few patients in our study, only 10% of the 69 patients with BNP in the gray zone. Our findings on the diagnostic utility of acoustical data were consistent with those of a recent study showing that the electronically recorded S3 is highly specific for heart failure.9 In addition, a study of ⬎2,400 patients found that S3 detected using auscultation showed a high positive likelihood ratio for detecting heart failure.10 Others observed that in contrast to electronic detection, auscultation was unreliable for showing the presence of S3.14 An additional limitation of physical examination compared with use of acoustic cardiographic measurements is that the former could yield measurements of percentage of EMAT. Our data showed that percentage of EMAT, whether used alone or in conjunction with the electronically detected S3, was especially powerful for augmenting the detection of LV dysfunction. The 2 diagnostic modalities, BNP and acoustic cardiographic parameters, reported in this study are noninvasive, readily available point-of-care tests. Although BNP traditionally was used to evaluate dyspneic patients in the emergency department, the present study showed that both BNP and acoustic cardiographic parameters can be used effectively in nonacute clinical settings, such as the physician’s office or clinic. This is especially important because such venues often do not have the wide array of diagnostic equipment available in the hospital environment. The number of patients with isolated diastolic LV dysfunction was small. However, for the overall population, the diagnostic performances of BNP in the range of 100 to 500 pg/ml were similar to those reported in the literature.8 Furthermore, the high specificity of S3 for LV dysfunction was also similar to findings of recent studies.9,10 The presence of
869
diastolic LV dysfunction was inferred from Doppler echocardiographic data and was not corroborated by invasive measurements of LV filling pressure. However, criteria used to determine the presence of diastolic LV dysfunction corresponded to those advocated by other investigators.11 Doppler echocardiographic, BNP, and acoustic cardiographic findings may have been influenced because many patients were receiving cardiac medications at the time they were studied. However, this is also likely to be true in patients in other clinical settings to whom our findings might be extrapolated. Acknowledgment: The authors thank Timothy Wheeler, BSCE, and Patti Arand, PhD, for assistance and Robert A. Warner, MD, for statistical analysis of the data. 1. Wuerz RC, Meador SA. Effects of prehospital medications on mortality and length of stay in congestive LV dysfunction. Ann Emerg Med 1992;21:669 – 674. 2. Williams D, Fitch & Associates. 2004 EMS 200 City Survey. J Emerg Med Sems 2005;Feb:42– 60. 3. Osganian SK, Zapka JG, Feldman HA, Goldberg RJ, Hedges JR, Eisenberg MS, Raczynski JM, McGovern PG, Cooper LS, Pandey DK, Linares AC, Luepker RV. Rapid Early Action for Coronary Treatment. Use of emergency medical services for suspected acute cardiac ischemia among demographic and clinical patient subgroups: the REACT trial. Prehosp Emerg Care 2002;6:175–185. 4. Canto JG, Zalenski RJ, Ornato JP, Rogers WJ, Kiefe CI, Magid D, Shlipak MG, Frederick PD, Lambrew CG, Littrell KA, Barron HV. Use of emergency medical services in acute myocardial infarction and subsequent quality of care: observations from the National Registry of Myocardial Infarction 2. Circulation 2002;106:3018 –3023. 5. McCullough PA, Nowak RM, McCord J, Hollander JE, Herrmann HC, Steg PG. B-Type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: analysis from Breathing Not Properly (BNP) Multinational Study. Circulation 2002;106:416 – 422. 6. Collins SP, Lindsell CJ, Storrow AB, Abraham WT. Prevalence of negative chest radiography results in the emergency department patient with decompensated LV dysfunction. Ann Emerg Med 2006;47:13–18. 7. Pozner CN, Levine BS, Shapiro N, Hanrahan JP. Concordance of field and emergency department assessment in the pre-hospital management of patients with dyspnea. Prehosp Emerg Care 2003;7:440 – 444. 8. Mueller C, Laule-Kilian K, Martina B, Schindler C, Buser P, Pfisterer M, Perruchoud AP. Use of B-type natriuretic peptide in the evaluation and management of acute dyspnea. N Engl J Med 2004;350:647– 654. 9. Marcus GM, Gerber IL, McKeown BH, Vessey JC, Jordan MV, Huddleston M, McCulloch CE, Foster E, Chatterjee K, Michaels AD. Association between phonocardiographic third and fourth heart sounds and objective measures of left ventricular function. JAMA 2005;293: 2238 –2244. 10. Drazner MH, Rame JE, Phil M, Stevenson LW, Dries DL. Prognostic importance of elevated jugular venous pressure and a third heart sound in patients with LV dysfunction. N Engl J Med 2001;345:574 –581. 11. Ommen SR, Nishimura RA. A clinical approach to the assessment of left ventricular function by Doppler echocardiography: update 2003. Heart 2003;89(suppl III):iii18 –iii23. 12. Weissler AM. A perspective on standardizing the predictive power of noninvasive cardiovascular tests by likelihood ratio computation: 2. Clinical applications. Mayo Clin Proc 1999;74:1072–1087. 13. Wang CS, Fitzgerald JM, Schulzer M, Mak E, Ayas NT. Does this dyspneic patient in the emergency department have congestive heart failure? JAMA 2005;294:1944 –1956. 14. Lok CE, Morgan CD, Ranganathan N. The accuracy and interobserver agreement in detecting the ‘gallop sounds’ by cardiac auscultation. Chest 1998;114:1283–1288.
Systolic and/or diastolic left ventricular (LV) dysfunction may result in heart failure, a major cause of disability and death. Although early diagnosis and treatment result in better clinical outcomes and lower cost of care, up to 46% of patients with acute dyspnea caused by heart failure currently are misdiagnosed.1–7 Although echocardiography is useful for detecting underlying LV dysfunction, the procedure is expensive and often not readily available. Therefore, physicians have adopted other tests for heart failure, such as measurement of B-type natriuretic peptide (BNP). However, BNP values showed an inverse relation between sensitivity and specificity for heart failure, and values in the so-called “gray zone” from 100 to 500 pg/ml were especially problematic.8 The third heart sound (S3) was highly specific for heart failure in the appropriate clinical context, but its detection using auscultation, especially in noisy environments, was unreliable.9,10 Therefore, acoustic cardiography, a technique for recording and algorithmically interpreting simultaneous diagnostic digital 12lead electrocardiographic and acoustic data, was developed. In this study, we tested the hypothesis that acoustic cardiography
a
Outpatient Clinic for Cardiology and Internal Medicine, Othmarsingen, Frauenfeld; and bCardiovascular Center, Klinik im Park, Zurich, Switzerland. Manuscript received January 26, 2007; revised manuscript received and accepted April 6, 2007. *Corresponding author: Tel.: ⫹41-62-896-1333; fax: ⫹41-62-8961464. E-mail address: [email protected] (M. Zuber). 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.04.019
can augment the use of BNP in the detection of LV dysfunction. Methods We evaluated 164 ambulatory patients (111 men; mean age 64 years; range 19 to 88) referred for Doppler echocardiographic evaluation for possible heart failure. The local medical ethics commission approved the protocol for the study, and all patients gave written informed consent before enrollment. Within 1 hour of Doppler echocardiography, each patient had BNP and serum creatinine measured and acoustic cardiographic data recorded using the Audicor device (Inovise Medical, Inc., Portland, Oregon). After this evaluation, we selected patients who had BNP of 100 to 500 pg/ml as subjects for additional study. Patients with atrial fibrillation were excluded. Each patient underwent a complete 2-dimensional and Doppler echocardiographic examination according to guidelines of the American Society of Echocardiography. The investigators interpreted Doppler echocardiographic findings blinded to both BNP and acoustic cardiographic data. Doppler echocardiographic findings were used to identify patients with versus without increased LV filling pressures. LV ejection fraction was measured using Simpson’s rule, and systolic LV dysfunction was defined as LV ejection fraction ⱕ50%. Diastolic function was assessed using the LV filling pattern and tissue Doppler examination of the lateral mitral annulus (for E=), including E/A ratio, deceleration time of the E wave, E/E= ratio, and pulmonary venous flow pattern. Diastolic function was graded as norwww.AJConline.org
Heart Failure/Augmenting BNP With Heart Sounds
mal, impaired relaxation, pseudonormal, or restrictive. Significant diastolic dysfunction was defined as either a pseudonormal (19 patients) or restrictive (14 patients) filling pattern. The presence of increased LV filling pressure was determined using the method described by Ommen and Nishimura.11 BNP and creatinine were measured in each patient after 30 minutes in the supine position using the Biosite Triage test (Biosite, Inc., San Diego, California) (normal values ⬍100 pg/ml), and creatinine levels, using the enzymatic UV Roche Diagnostics Corp (Indianapolis, Indiana). Reflotron test (normal values ⬍97 mol/L for men and ⬍80 for women). Acoustic cardiography consists of recording and algorithmically interpreting simultaneous digital 12-lead electrocardiographic and acoustic, for example, acoustic cardiographic, data, including S3, using dual-purpose sensors placed in the V3 and V4 positions with the Audicor system. Acoustic cardiography requires inexpensive equipment, can be used in a wide variety of clinical environments, requires approximately the same time and effort as that needed to record a standard 12-lead electrocardiogram, and shows no significant variability among operators. A 10-second Audicor recording was obtained for each patient and analyzed using the computerized algorithm with measurements generated for S3 and various systolic parameters. The following parameters of LV function were evaluated both individually and in combination with each other. The system evaluated S3 using a computerized measurement of the intensity and persistence of acoustic energy of sounds that had the appropriate frequency and timing for S3 (S3 strength) occurring up to 130 ms after the second heart sound. The range of possible values of S3 strength was 0 to 10 units, and if the value was ⱖ5.0, the Audicor computerized algorithm declared that S3 was present. Percentage of electromechanical activation time (EMAT) is a continuous parameter that indicates the interval in milliseconds from the onset of the Q wave of the electrocardiogram to the mitral component of the first heart sound (S1) divided by the RR interval in milliseconds. EMAT measures the time required for ventricular contraction to generate sufficient force to close the mitral valve. It may be abnormally prolonged partly because of the decreased dP/dT associated with LV systolic dysfunction. Percentage of EMAT expresses the proportion of the total cardiac cycle that EMAT occupies. For the 69 patients with BNP in the gray zone, diagnostic performances of BNP alone and in combination with acoustic cardiographic parameters for detecting Doppler echocardiographic abnormalities compatible with systolic and diastolic LV dysfunction were calculated. For percentage of EMAT and the additive combination of S3 strength and percentage of EMAT, receiver-operating characteristic curves were analyzed to determine values that yielded the highest positive likelihood ratios for each diagnosis. Positive and negative likelihood ratios were calculated using the following formulas: Positive likelihood ratio ⫽ Sensitivity/(1 ⫺ Specificity), and Negative likelihood ratio ⫽ Specificity/(1 ⫺ Sensitivity). To avoid dividing by 0, the denominator of the formula for positive likelihood ratio
867
Table 1 Diagnostic performances of various ranges of B-type natriuretic peptide (BNP) levels for detecting any type of left ventricular dysfunction BNP (pg/ml)
Sensitivity
⬎100 100–500 ⬎500
62/93 (67%) 51/93 (55%) 11/93 (12%)
Specificity
53/71 (75%) 53/71 (75%) 70/71 (99%)
Positive Likelihood Ratio
Negative Likelihood Ratio
3 2 12
2 2 1
was assigned a value of 1.0 in cases in which specificity was 100%. Calculating positive and negative likelihood ratios using these formulas expressed these ratios on the same numerical scale and therefore facilitated the comparison of a test’s rule-in and rule-out power.12 A positive or a negative likelihood ratio ⬎10 shown by a diagnostic test was considered to have excellent rule-in or rule-out power for the condition being tested, respectively. Unlike positive and negative predictive values, diagnostic inferences based on positive and negative likelihood ratios were independent of the prevalence of the abnormality in the population being tested.12 Statistical analyses were performed using SPSS, version 13.0 (SPSS, Inc., Chicago, Illinois). The entire group of patients with LV dysfunction was analyzed. Separate analysis was performed of subgroups with systolic LV dysfunction with normal LV filling pressures, systolic LV dysfunction with increased LV filling pressures, and isolated diastolic LV dysfunction. Results Of 164 patients in the study, BNP was ⬍100 pg/ml in 83 patients, 100 to 500 in 69 patients (gray zone), and ⬎500 in 12 patients. Of 69 patients with BNP in the gray zone, 51 (74%) had LV dysfunction. Of these 51 patients, 25 (49%) had Doppler echocardiographic evidence of systolic LV dysfunction with normal filling pressures, 19 (37%) had systolic LV dysfunction with increased filling pressures, and 7 (14%) had isolated diastolic LV dysfunction. Of 7 patients with isolated diastolic LV dysfunction, 5 (71%) had a pseudonormal and 2 (29%) had a restrictive pattern of LV inflow by Doppler. Of 69 patients with BNP in the gray zone, Doppler echocardiography showed mildly delayed diastolic relaxation in 40 (58%) and 6 (9%) had no abnormality of diastolic function. Table 1 lists diagnostic performances of BNP for detecting LV dysfunction in the 3 different ranges of values of ⬎100, 100 to 500, and ⬎500 pg/ml. Table 1 lists the inverse relation between sensitivity and specificity shown by low versus high BNP. It also confirms that BNP in the gray zone of 100 to 500 pg/ml had neither good rule-in nor good rule-out power for LV dysfunction. As listed in Table 2, acoustic cardiographic parameters augmented values of BNP in the gray zone for detecting LV dysfunction. Both percentage of EMAT alone and in combination with S3 produced excellent rule-in power for LV dysfunction by providing very high positive likelihood ratios. Tables 3 to 5 list similar types of data for the 3 subgroups
868
The American Journal of Cardiology (www.AJConline.org)
Table 2 Using sound parameters to detect any type of left ventricular dysfunction in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⱖ5.0 ⱖ10.2 ⱖ14.4
Sensitivity 51/93 (55%) 20/51 (39%) 34/51 (67%) 35/51 (69%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
53/71 (75%) 17/18 (94%) 18/18 (100%) 18/18 (100%)
2 7 67 69
2 2 3 3
Table 3 Using sound parameters to detect systolic left ventricular dysfunction with normal filling pressures in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⬎5.0 ⬎12.0 ⬎15.9
Sensitivity 25/55 (45%) 5/25 (25%) 13/25 (52%) 14/25 (56%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
57/82 (70%) 22/25 (88%) 25/25 (100%) 25/25 (100%)
2 2 52 56
1 1 2 2
Table 4 Using sound parameters to detect systolic left ventricular dysfunction with elevated filling pressures in patients with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value 100–500 ⱖ5.0 ⱖ12.0 ⱖ15.9
Sensitivity 19/27 (70%) 12/19 (63%) 9/19 (47%) 13/19 (68%)
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
57/82 (70%) 22/25 (88%) 25/25 (100%) 25/25 (100%)
2 5 47 68
2 2 2 3.1
Table 5 Using sound parameters to detect isolated diastolic left ventricular dysfunction with B-type natriuretic peptide (BNP) of 100 to 500 pg/ml compared with BNP alone Parameter BNP alone S3 strength Percent EMAT S3 strength ⫹ percent EMAT
Value
Sensitivity
Specificity
Positive Likelihood Ratio
Negative Likelihood Ratio
100–500 ⬎5.0 ⬎12.0 ⬎15.9
7/11 (64%) 2/7 (29%) 1/7 (14%) 1/7 (14%)
83/129 (64%) 39/46 (85%) 33/46 (72%) 32/46 (70%)
2 2 1 1
2 4 7 1
of LV dysfunction. As listed in these tables, for each subgroup of LV dysfunction, BNP in the gray zone had poor diagnostic specificity. However, Tables 3 and 4 also show that regardless of level of filling pressures, using acoustic cardiographic parameters to elucidate the significance of BNP in this range provided excellent rule-in power for systolic LV dysfunction. In contrast, Table 5 shows that the diagnostic impact of acoustic cardiographic parameters is much lower in patients with isolated diastolic LV dysfunction than for those with systolic LV dysfunction. Tables 3 to 5 also show that threshold values of the acoustic cardiographic parameters required to obtain maximum positive likelihood ratios were similar for the 3 types of LV dysfunction. Discussion The findings of this study confirmed that the pathologic significance of BNP in the gray zone of 100 to 500 pg/ml
was uncertain. Importantly, 42% of our subjects with suspected heart failure had BNP values in this range, which indicated this is a commonly encountered problem. Our study showed the great diagnostic ability of detection of LV dysfunction by using S3 and percentage of EMAT as an accurate and fast tool for outpatients. The specificity and positive likelihood ratio for LV dysfunction shown by BNP in the gray zone were only 75% and 2.7, respectively. The specificity of BNP in this range was similar in each subgroup of LV dysfunction. Such a limited rule-in power would be unsuitable for a screening test because it would produce too many false-positive results. Using the more stringent BNP threshold of ⬎500 pg/ml produced very high specificity, but would identify only 12% of the truly positive cases of LV dysfunction. As expected, a 100-pg/ml increase in BNP was associated with decreased sensitivity and increased specificity.
Heart Failure/Augmenting BNP With Heart Sounds
In the diagnostic evaluation of dyspneic patients, BNP ⬍100 pg.ml is commonly believed to make the diagnosis of heart failure associated with LV dysfunction highly improbable.13 Conversely, in the absence of acute myocardial ischemia, renal insufficiency, or recent administration of neseritide, high BNP makes the diagnosis of heart failure likely. However, consistent with findings of the present study, BNP in the intermediate or gray zone is especially difficult to interpret. For example, in 452 patients of the B-Type Natriuretic Peptide for Acute Shortness of Breath Evaluation (BASEL) trial, it was considered necessary to use additional diagnostic information before making a definitive diagnosis of heart failure in patients with BNP in the range of 100 to 500 pg/ml.8 The present study showed that acoustic cardiographic parameters can effectively resolve the diagnostic uncertainty associated with BNP in the gray zone. The data showed that acoustic cardiographic parameters increased positive likelihood ratios for detecting LV dysfunction by a factor ⬎30. Our data showed that diagnostic performances of acoustic cardiographic parameters were best in patients with systolic LV dysfunction regardless of whether filling pressures were increased. Conversely, their diagnostic performances were much poorer in patients with isolated diastolic LV dysfunction. However, patients with isolated diastolic LV dysfunction represented few patients in our study, only 10% of the 69 patients with BNP in the gray zone. Our findings on the diagnostic utility of acoustical data were consistent with those of a recent study showing that the electronically recorded S3 is highly specific for heart failure.9 In addition, a study of ⬎2,400 patients found that S3 detected using auscultation showed a high positive likelihood ratio for detecting heart failure.10 Others observed that in contrast to electronic detection, auscultation was unreliable for showing the presence of S3.14 An additional limitation of physical examination compared with use of acoustic cardiographic measurements is that the former could yield measurements of percentage of EMAT. Our data showed that percentage of EMAT, whether used alone or in conjunction with the electronically detected S3, was especially powerful for augmenting the detection of LV dysfunction. The 2 diagnostic modalities, BNP and acoustic cardiographic parameters, reported in this study are noninvasive, readily available point-of-care tests. Although BNP traditionally was used to evaluate dyspneic patients in the emergency department, the present study showed that both BNP and acoustic cardiographic parameters can be used effectively in nonacute clinical settings, such as the physician’s office or clinic. This is especially important because such venues often do not have the wide array of diagnostic equipment available in the hospital environment. The number of patients with isolated diastolic LV dysfunction was small. However, for the overall population, the diagnostic performances of BNP in the range of 100 to 500 pg/ml were similar to those reported in the literature.8 Furthermore, the high specificity of S3 for LV dysfunction was also similar to findings of recent studies.9,10 The presence of
869
diastolic LV dysfunction was inferred from Doppler echocardiographic data and was not corroborated by invasive measurements of LV filling pressure. However, criteria used to determine the presence of diastolic LV dysfunction corresponded to those advocated by other investigators.11 Doppler echocardiographic, BNP, and acoustic cardiographic findings may have been influenced because many patients were receiving cardiac medications at the time they were studied. However, this is also likely to be true in patients in other clinical settings to whom our findings might be extrapolated. Acknowledgment: The authors thank Timothy Wheeler, BSCE, and Patti Arand, PhD, for assistance and Robert A. Warner, MD, for statistical analysis of the data. 1. Wuerz RC, Meador SA. Effects of prehospital medications on mortality and length of stay in congestive LV dysfunction. Ann Emerg Med 1992;21:669 – 674. 2. Williams D, Fitch & Associates. 2004 EMS 200 City Survey. J Emerg Med Sems 2005;Feb:42– 60. 3. Osganian SK, Zapka JG, Feldman HA, Goldberg RJ, Hedges JR, Eisenberg MS, Raczynski JM, McGovern PG, Cooper LS, Pandey DK, Linares AC, Luepker RV. Rapid Early Action for Coronary Treatment. Use of emergency medical services for suspected acute cardiac ischemia among demographic and clinical patient subgroups: the REACT trial. Prehosp Emerg Care 2002;6:175–185. 4. Canto JG, Zalenski RJ, Ornato JP, Rogers WJ, Kiefe CI, Magid D, Shlipak MG, Frederick PD, Lambrew CG, Littrell KA, Barron HV. Use of emergency medical services in acute myocardial infarction and subsequent quality of care: observations from the National Registry of Myocardial Infarction 2. Circulation 2002;106:3018 –3023. 5. McCullough PA, Nowak RM, McCord J, Hollander JE, Herrmann HC, Steg PG. B-Type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: analysis from Breathing Not Properly (BNP) Multinational Study. Circulation 2002;106:416 – 422. 6. Collins SP, Lindsell CJ, Storrow AB, Abraham WT. Prevalence of negative chest radiography results in the emergency department patient with decompensated LV dysfunction. Ann Emerg Med 2006;47:13–18. 7. Pozner CN, Levine BS, Shapiro N, Hanrahan JP. Concordance of field and emergency department assessment in the pre-hospital management of patients with dyspnea. Prehosp Emerg Care 2003;7:440 – 444. 8. Mueller C, Laule-Kilian K, Martina B, Schindler C, Buser P, Pfisterer M, Perruchoud AP. Use of B-type natriuretic peptide in the evaluation and management of acute dyspnea. N Engl J Med 2004;350:647– 654. 9. Marcus GM, Gerber IL, McKeown BH, Vessey JC, Jordan MV, Huddleston M, McCulloch CE, Foster E, Chatterjee K, Michaels AD. Association between phonocardiographic third and fourth heart sounds and objective measures of left ventricular function. JAMA 2005;293: 2238 –2244. 10. Drazner MH, Rame JE, Phil M, Stevenson LW, Dries DL. Prognostic importance of elevated jugular venous pressure and a third heart sound in patients with LV dysfunction. N Engl J Med 2001;345:574 –581. 11. Ommen SR, Nishimura RA. A clinical approach to the assessment of left ventricular function by Doppler echocardiography: update 2003. Heart 2003;89(suppl III):iii18 –iii23. 12. Weissler AM. A perspective on standardizing the predictive power of noninvasive cardiovascular tests by likelihood ratio computation: 2. Clinical applications. Mayo Clin Proc 1999;74:1072–1087. 13. Wang CS, Fitzgerald JM, Schulzer M, Mak E, Ayas NT. Does this dyspneic patient in the emergency department have congestive heart failure? JAMA 2005;294:1944 –1956. 14. Lok CE, Morgan CD, Ranganathan N. The accuracy and interobserver agreement in detecting the ‘gallop sounds’ by cardiac auscultation. Chest 1998;114:1283–1288.