Article Text

Download PDFPDF

Repeated echocardiography after the diagnosis of endocarditis: too much of a good thing?
  1. C H Cabell,
  2. V G Fowler Jr
  1. Department of Medicine, Duke University School of Medicine, and the Duke Clinical Research Institute, Durham, North Carolina, USA
  1. Correspondence to:
    Dr Christopher H Cabell
    Box 3850, Duke University Medical Center, Durham, NC 27710, USA;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Although echocardiography is seen as the technology of choice for the diagnosis of infective endocarditis, it is being increasingly overused in clinical scenarios with a low pre-test probability of disease

Since the advent of two dimensional transthoracic echocardiography (TTE) in the 1970s and high frequency transoesophageal echocardiography (TOE) imaging in the 1980s, echocardiography has become a standard diagnostic tool in patients with suspected infective endocarditis (IE). It is now well established that echocardiography is a technology of choice for the diagnosis of IE,1 and that echocardiography can detect cardiac involvement in a significant proportion of patients with clinically occult IE.2,3 Because IE is a lethal infection that can be difficult to diagnose clinically, clinicians who care for patients at risk for IE often have a low threshold for employing echocardiography. This clinical practice has several implications. While echocardiography can often provide a rapid diagnosis, its optimal use is predicated on the appropriate pre-test probability of disease.4,5 By contrast, echocardiography is increasingly overused in clinical scenarios with a low (< 2–3%) pre-test probability of disease, where its diagnostic utility diminishes.6,7


It is in this light that Vieira and colleagues,8 in a study reported in this issue of Heart, have asked an important question: “What is the diagnostic yield of repeated echocardiograms in patients with suspected IE?” Over a three year period, they evaluated 262 patients with suspected IE referred for echocardiography. They found that repeat echocardiography was frequent: TTEs were repeated at least once in 192 (72.2%) patients while TOEs were repeated in 49 (18.4%) of patients. The average number of TTE examinations was 2.4, but six patients had at least six TTEs. In a similar fashion, the mean number of TOE examinations was 1.7, although four patients had at least four TOEs and one patient had five. The authors found that while repeated echocardiograms were occasionally helpful, no additional diagnostic information was provided after the second or third echocardiogram (TTE or TOE).

Although this information is helpful in understanding the limitations of excessive imaging, it is unclear how many of the 262 patients had both TTE and TOE examinations and how these findings affected subsequent evaluation. In addition, the clinical setting of a large university cardiology hospital with a high percentage of patients with prosthetic valves and intracardiac devices may have skewed the findings toward an increase in the use of echocardiography. Finally, the aggressive use of imaging may reflect a high clinical suspicion for underlying IE, as most of these patients were referred to the university based cardiology hospital in which the study took place with positive blood cultures and other suggestive findings.


These limitations notwithstanding, it is evident from this and previous studies that imaging technologies such as echocardiography can be overused in certain clinical scenarios such as the evaluation of suspected endocarditis. For instance, Kurupuu and colleagues have shown that 53% of echocardiograms could be avoided without loss of diagnostic accuracy by using a simple algorithm in patients with a low pre-test probability of disease.6 In addition, Greaves and colleagues have shown that the collective absence of five simple clinical criteria indicated a zero probability of a TTE showing evidence of endocarditis.7 Collectively, these studies have shown that in patients with very low pre-test probability of disease that echocardiography may be avoided without the loss of diagnostic accuracy.

Although current guidelines highlight the use of TTE and/or TOE imaging depending on the clinical scenario,9 studies have shown that a strategy of TOE imaging is the most cost effective in the majority of clinical situations. For example, Heidenreich and colleagues have show that in suspected endocarditis a diagnostic strategy that focuses on TOE as the initial imaging modality is more cost effective than a staged procedure with TTE and is a dominant strategy over empiric antibiotic treatment alone.5

For a variety of reasons it is important to limit the indiscriminate use of diagnostic technologies. From a safety standpoint, low risk procedures such as TOE may have an unacceptable risk:benefit ratio as the diagnostic yield approaches zero. In addition, in the era of rapidly rising medical costs, excessive procedures lead to substantial financial waste. For example, in the USA it has been estimated that the appropriate use of TOE to establish the length of treatment in catheter associated Staphylococcus aureus bacteraemia could save over $120 million annually.10


It is clear from the work by Vieira and colleagues,8 as well as previous studies, that the real issue in evaluating patients with suspected endocarditis is the pre-test probability of disease. In any given patient at risk for IE, such as patients with prosthetic valves, this pre-test probability may be different for different microorganisms. Studies are needed that focus attention on understanding which clinical characteristics are the most important in predicting risk for the subsequent development of IE. From these studies, scoring systems could be developed that help clinicians at the bedside improve their precision in developing the appropriate pre-test probability for disease. A scoring system such as this could be tested prospectively, validated, and then disseminated into clinical practice by existing guideline committees. Until such data become available, it will be very difficult to limit the indiscriminate use of expensive diagnostic modalities by our technology driven society.


This work was supported in part by National Institutes of Health grants HL70861 (CHC) and R01AI59111 (VGF)

Although echocardiography is seen as the technology of choice for the diagnosis of infective endocarditis, it is being increasingly overused in clinical scenarios with a low pre-test probability of disease