| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Phoenix, AZ
Dr. Richman is Vice-Chairman, Division of Research, Department of Emergency Medicine, Mayo Clinic Hospital.
Correspondence to: Peter B. Richman, MD, MBA, Department of Emergency Medicine, Mayo Clinic Hospital, Phoenix, AZ 85054; e-mail: PRichmanMD{at}aol.com
In the 16 years since the first Prospective Investigation of Pulmonary Embolism Diagnosis report,1 clinicians have gained numerous diagnostic tools beyond ventilation-perfusion imaging and conventional pulmonary angiography to rule out pulmonary embolism. Many still consider angiography, an invasive modality, the reference standard to diagnose the condition. However, a recent metaanalysis2 of 15 studies (3,500 patients) revealed that negative CT angiographic findings of the chest alone might be sufficient to safely rule out pulmonary embolism. Quiroz et al2 reported that CT angiography had a negative predictive value for the disease of 99.1% (95% confidence interval [CI], 98.7 to 99.5%) and a negative predictive value for fatal pulmonary embolism of 99.4% (95% CI, 98.7 to 99.9%).
In view of the capability of CT to safely exclude pulmonary embolism, it is not surprising that it is the imaging modality of choice to evaluate emergency department patients in many settings.3 CT holds the additional advantage of allowing an emergency physician to concomitantly identify other acute thoracic diseases.4 Furthermore, an investigation5 revealed that following imaging of the chest and without additional contrast infusion, indirect CT venography of the lower extremities may be performed to increase the sensitivity of the CT protocol to rule out thromboembolism.
Despite the advantages of CT imaging, many questions remain as to its ultimate place in an algorithm to exclude pulmonary embolism. The procedure is not without risk, as there are potential complications including allergy and nephrotoxicity. A recent investigation6 also reveals that CT has some limitations with respect to interobserver disagreement between radiologists as to the presence of pulmonary embolism and risk of false-positive diagnosis. Perhaps of equal importance, CT imaging is expensive and time consuming. In an era of emergency department overcrowding and explosive health-care inflation, it is imperative that investigators seek more efficient and cost-effective methods to rule out diseases.
In this issue of CHEST (see page 1417), Kline et al7 add to a growing body of evidence that supports the notion that clinical evaluation and blood testing may be sufficient to exclude pulmonary embolism in selected populations. The authors7 describe the emergency department use of Simplify (Agen Biomedical; Brisbane, Australia), a rapid, bedside, qualitative d-dimer test, in combination with either clinical judgment or one of two previously validated decision rules to rule out pulmonary embolism for low-risk patients.89 The study is important for at least two reasons.
First, it is a step toward a protocol to safely and efficiently rule out pulmonary embolism at the bedside. Of the current commercially available d-dimer tests, only one assay (VIDAS; Biomerieux S.A.; Marcy L’Etoile, France) has an indication for use to exclude pulmonary embolism. In contrast to the bedside Simplify assay used by Kline et al,7 which may be performed in 10 min, the VIDAS assay takes 35 min to perform on a laboratory machine. When one accounts for factors such as nursing time, the turnaround time for a result from the VIDAS assay is probably closer to 2 h in even the most efficient emergency department. While further investigation is warranted to validate the safe use of the Simplify assay in a rapid rule out pulmonary embolism algorithm, the time advantages of the protocol reported by Kline et al7 are obvious.
Second, the novel methodology of the study by Kline et al7 appears to be another advance. A previous investigation10 of this tool was limited by its small population size, low disease prevalence, and absence of a defined method to identify patients with low pretest probability who were appropriate for laboratory screening. Kline et al7 advance our understanding of how to utilize the Simplify assay within the context of two well-designed and previously validated decision rules to define low-risk populations.89 When the Simplify assay result was negative in the current study, the authors7 found a very low prevalence of pulmonary embolism in low-risk patients identified by both the Charlotte (1.1%; 95% CI, 0.6 to 1.7%) and Canadian (1.2%; 95% CI, 0.7 to 2.0%) rules, respectively.
Despite the promise of these findings, there are several points of caution worth noting with respect to the Simplify d-dimer test. Its reported sensitivity here (80.6%; 95% CI, 71.8 to 87.5%) and previously does not appear adequate for use without adjunctive clinical decision-making tools.710 In addition, the authors7 did not assess for interobserver agreement for positive and negative results. Certainly, before widespread use, investigators should confirm that different observers would draw the same conclusion when reading the Simplify test cartridge.
In addition to concerns regarding the testing device, further study is also warranted to determine the validity of unstructured assessment of pretest probability for pulmonary embolism. The authors11 previously found that when physicians at their center used only clinical "gestalt" (ie, unstructured assessment) to estimate patients at < 15% risk (low risk), the prevalence of disease was only 2.6%. However, interobserver agreement was only moderate among study physicians for a designation of patients at low risk.
In the current study, Kline et al7 describe that patients had a very low prevalence of disease (0.7%; 95% CI, 0.3 to 1.4%) when physicians defined a patient as low risk via a similar unstructured estimate; and, then, the subsequent Simplify test result was negative. However, one should question whether this "unstructured" estimate is truly unstructured. The Charlotte rule was previously developed by the study authors in the same setting as the current investigation.9 Further, the investigators have also analyzed the Canadian score with many of the same faculty participating in patient enrollment.11 As a consequence, it is likely that the physicians providing the unstructured estimates in this institution were familiar with both validated decision rules prior to study initiation. Therefore, the observed physician performance with respect to unstructured estimate may, alternatively, represent some combination use either consciously or unconsciously of the two rules. Anecdotally, we know that many emergency physicians do not routinely utilize any validated algorithm in their decision making related to the evaluation of possible pulmonary embolism because they are unfamiliar with such tools. It remains uncertain whether physicians unfamiliar with the Charlotte and Canadian scoring systems would perform as well the physicians who enrolled patients in the current study. Until independent studies across various settings confirm the reliability of unstructured estimates of risk, patients will be better served if their physicians rely on the aforementioned Charlotte and Canadian scoring systems to determine pretest probability.
Footnotes
Dr. Richman does not have any financial interests to disclose.
References
This article has been cited by other articles:
|
|
 |
|
  Negative D-Dimer Testing to Exclude PE Journal Watch (General), July 25, 2006; 2006(725): 2 - 2. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
