Computer-Aided Auscultation of the Heart: The Clinical Opportunity.
Clinicians have long depended on the stethoscope as the clinical tool that helps them to characterize heart sounds and identify abnormalities suggestive of underlying cardiac disease. However, as technology advances, the confidence of most physicians in their ability to accurately detect abnormal cardiac findings through physical examination is diminishing. Advanced technologies have become so much the standard in the practice of cardiology that physicians have abandoned basic clinical assessment techniques and are unable to appreciate the cardiac findings that may be obtained with the stethoscope. Before the technological revolution, the well-trained physician was capable of diagnosing cardiac disease consistently and accurately using basic clinical skills with the assistance of the stethoscope. Despite the current emphasis on technology in medicine, it still holds true that most patients seen in clinical practice today for evaluation of cardiovascular disease can be accurately diagnosed in the physician's office through careful cardiac clinical examination.1
Precise auscultation of the heart with a stethoscope throughout the cardiac cycle provides the clinician with information on the timing, duration, pitch, location, and intensity of normal heart sounds (S1 and S2) and can reveal the presence or absence of other sounds, such as S3, S4, and murmurs. However, auscultation performed with a stethoscope is inherently difficult in that it is based on the examiner's ability to hear and interpret a wide variety of sounds, most of which fall within the lower ranges of sounds detectable by the human auditory system. Furthermore, the practice is highly subjective, with descriptions of auscultatory sounds varying greatly among clinicians. In the current era of advanced technology, cost containment, and managed care, there is a great need to renew the clinical skill of cardiac auscultation and bring it back to the forefront of cardiac care and diagnosis. Today, the clinical skills of the well-trained physician who can make an accurate and cost-effective cardiovascular diagnosis in the office or at the bedside are becoming increasingly important.1 A cost-effective device that would improve the clinical skills of cardiac auscultation and provide physicians with additional information as to whether they should refer patients for more expensive and sometimes invasive procedures is warranted.
A computer-aided auscultation device would be expected to improve the diagnostic accuracy of this fundamental practice by reinforcing good auscultatory practices and providing objective, repeatable findings that are readily interpretable. The purpose of this monograph is to describe the potential role of computer-aided auscultation devices in clinical practice.
The Impact of Computer-assisted Auscultation on Physician Referrals of Asymptomatic Patients with Heart Murmurs
Studies suggest that as many as 50% to 70% of asymptomatic children referred for specialist evaluation or echocardiography because of a murmur have no heart disease, although it is widely held that the distinction of innocent from pathologic murmurs should and could be reliably made at the point of care using auscultation and other basic clinical skills.1-6 Several investigators have reported that patients are over-referred for specialist evaluation or echocardiography.2,7,8 Over-referral and over use of echocardiography are likely due at least in part to declining expertise in cardiac auscultation among referring physicians, resulting in diminished certainty about the presence or absence of disease.
Among various strategies that are being pursued to compensate for declining auscultatory skills, computer-assisted auscultation (CAA) of the heart is an emerging and promising technology. Automatically derived characteristics of heart sounds have been shown to provide a basis for accurate detection of pathological murmurs,9 and identification of systolic murmurs that are louder in standing than supine position.10 What has not yet been established is the potential improvement in clinical decision-making by physicians that can be achieved using computer-based analysis of heart sounds obtained at the point of care.
The objective of this study was to assess the impact of CAA on primary care physician's accuracy of murmur detection and characterization as well as their decisions to refer asymptomatic patients with murmurs, as compared with the American College of Cardiology/American Heart Association (ACC/AHA) referral guidelines (Table 1).11 Physicians were evaluated using prerecorded heart sounds from actual patients without and with the use of CAA provided by Cardioscan (Zargis Medical Corp., Princeton, N.J., USA), a new diagnostic decision support system.