Elsevier

Thrombosis Research

Volume 125, Issue 4, April 2010, Pages e123-e127
Thrombosis Research

Regular Article
Safety of excluding acute pulmonary embolism based on an unlikely clinical probability by the Wells rule and normal D-dimer concentration: A meta-analysis

https://doi.org/10.1016/j.thromres.2009.11.009Get rights and content

Abstract

Introduction

The Wells clinical decision rule (CDR) and D-dimer tests can be used to exclude pulmonary embolism (PE). We performed a meta-analysis to determine the negative predictive value (NPV) of an “unlikely” CDR (≤ 4 points) combined with a normal D-dimer test and the safety of withholding anti-coagulants based on these criteria.

Methods

Prospective studies that withheld anti-coagulant treatment from patients with clinically suspected PE and an “unlikely” CDR in combination with a normal D-dimer concentration without performing further tests were searched for in Medline, Cochrane and Embase. Primary endpoints were the recurrence rate of venous thromboembolism (VTE) and PE-related mortality during 3-months follow-up.

Results

Four studies including 1660 consecutive patients were identified. The pooled incidence of VTE after initial exclusion of acute PE based on an “unlikely” CDR and normal D-dimer was 0.34% (95%CI 0.036-0.96%), resulting in a NPV of 99.7% (95%CI: 99.0-99.9%, random effects-model). The risk for PE related mortality was very low: 1/1660 patients had fatal PE (0.06%, 95%CI 0.0017-0.46%).

Conclusion

Acute PE can be safely excluded in patients with clinically suspected acute PE who have an “unlikely” probability and a negative D-dimer test and anticoagulant treatment can be withheld. There is no need for additional radiological tests in these patients to rule out PE.

Introduction

The possibility of excluding the diagnosis of acute pulmonary embolism (PE) without the need for radiological imaging is a great step forward in the complex management of patients suspected of this disease. In this way diagnostic time, costs and potential complications from performing computed tomographic pulmonary angiography (CTPA) or ventilation-perfusion (V-Q) lung scintigraphy, including radiation exposure and allergic reaction to contrast dye, are kept limited.

The Wells rule (Table 1) is widely used as a clinical decision rule for the assessment of clinical probability for PE [1], [2], [3], [4], [5]. Originally using this rule, patients could be divided in to three categories of increasing risk for having PE. These categories were low (< 2.0 points, 2.0% PE), moderate (2.0-6.0 points, 18.8% PE) and high (> 6.0 points, 50% PE) clinical probability [5]. It has previously been shown that anticoagulant therapy could be safely withheld from patients classified as low risk with a normal D-dimer concentration (28.4% of the population) [5]. In a post hoc analysis, it was suggested that using dichotomization of the Wells rule – dividing patients in to PE “unlikely” (≤ 4 points) and PE “likely” (> 4 points) – PE might be safely excluded based on a low as well as an “unlikely” clinical probability and a normal D-dimer test, with very low three months venous thromboembolism (VTE) recurrence [5]. Importantly, in that study 50% of patients had a Wells score of 4 points or less (5.1% PE) compared to only 31% of patients who had a score of 2 points or less (2.0% PE). Thus, the 2-level approach increases the number of patients in whom radiological tests can be avoided although it puts more weight on the sensitivity of the D-dimer test. An additional advantage of the dichotomization of the Wells rule is that it involves a simpler triage in patients.

In spite of the important advantages of a diagnostic strategy including the dichotomized Wells rule and D-dimer testing to exclude acute PE, widely implementation of such a diagnostic algorithm is lacking [6]. In one recent study, only 58% of the patients with a positive D-dimer underwent, as should be, CT-scanning, and in 7% of the patients with negative D-dimer results, superfluous CT scans were performed [7]. One explanation for this is that the safety of this algorithm is understudied and untreated PE is a major concern for every physician since that has been shown to have a high mortality rate ranging from 9.2-51% [8], [9].

For this reason, we have performed a systematic review and meta-analysis of studies in patients that excluded acute PE on the basis of an “unlikely” clinical probability (Wells rule ≤ 4 points) and a normal high sensitive D-dimer test to evaluate the safety of such a diagnostic strategy.

Section snippets

Data sources

A literature search was performed to locate all prospective studies using a diagnostic strategy including a dichotomized clinical decision rule and a D-dimer test to rule out PE. Predefined search terms were used as Mesh terms as well as free text words (Appendix A). All full articles published after the introduction of the Wells score (2000) till the December 1st 2008 (date of our final search) were eligible for inclusion in the analysis. The articles were limited to the English, German,

Study selection

The literature search revealed 157 articles. Of these, 139 were excluded after review of the title and abstract. After full review another 14 articles were excluded because of use of other clinical decision rules, retrospective assessment of the Wells rule or performance of additional diagnostic tests to rule out PE. Eventually, 4 studies were approved for inclusion in this meta-analysis (Fig. 1) [1], [2], [3], [4].

Quality and characteristics of included studies

All four identified studies were prospective studies with a follow-up period of

Discussion

The main finding of our study is that the pooled incidence of morbidity due to PE after an “unlikely” clinical decision rule (Wells Rule ≤ 4 points) in combination with a normal D-dimer test is 0.34% (95%CI 0.036- 0.96%), resulting in a NPV of 99.7% (95%CI 99.0-99.9%). The upper limit of the 95% confidence interval of the three months VTE-rate after negative pulmonary angiography is 2.7% [11]. Using this 2.7% as the threshold for the safe exclusion of PE, our analysis shows that ruling out PE on

Conflict of Interest

None of the authors has a conflict of interest with a commercial entity.

Acknowledgements

All authors had full access to and take responsibility for the integrity of the data and the accuracy of the data analysis.

Authorship

SM Pasha: Study concept and design, acquisition of data, analysis and interpretation of data, drafting the manuscript

FA Klok: Study concept and design, acquisition of data, analysis and interpretation of data, drafting the manuscript

JD Snoep: analysis and interpretation of data, drafting the manuscript, critical revision of the manuscript for important

References (12)

There are more references available in the full text version of this article.

Cited by (82)

View all citing articles on Scopus
View full text