Annals of Emergency Medicine
Volume 52, Issue 1 , Pages 76-79, July 2008

Clinical Assessment of the Patient With a Suspected Pulmonary Embolism

Department of Emergency Medicine, SUNY–Downstate Medical Center, Brooklyn, NY.

published online 28 August 2007.

Article Outline

 

[Ann Emerg Med. 2008;52:76-79.]

Back to Article Outline

Rational Clinical Examination Review Source 

This is a rational clinical examination abstract, a regular feature of the Annals' Evidence-Based Emergency Medicine (EBEM) series. Each features an abstract of a rational clinical examination review from the Journal of the American Medical Association and a commentary by an emergency physician knowledgeable in the subject area.

The source for this rational clinical examination review abstract is: Chunilala SD, Eikelbloom JW, Attia J, et al. The rational clinical examination. Does this patient have a pulmonary embolism? JAMA. 2003;290:2849-2858.1 The Annals' EBEM editors assisted in the preparation of the abstract of this rational clinical examination review, as well as selection of the Evidence-Based Medicine Teaching Points.

Back to Article Outline

Objective 

To summarize the value of the clinical examination, clinical prediction rules, and D-dimer assays in detecting pulmonary embolism in adult patients.

Back to Article Outline

Data Sources 

The authors report a MEDLINE search for English-language articles from 1966 to March 2003 and hand searching of reference lists of identified articles. The search strategy is described in detail.

Back to Article Outline

Study Selection 

Studies were included that described the operating characteristics of the clinical examination for the diagnosis of pulmonary embolism. The studies selected for inclusion had to satisfy 3 conditions: estimated pretest probability of pulmonary embolism, blinding, and validated methodology for confirming or excluding pulmonary embolism. In addition, each study had to have a minimum number (N>50) of confirmed cases of pulmonary embolism. No selection criteria are reported for the 3 studies cited describing D-dimer operating characteristics. Two of these are medical record review studies on selected populations, whereas a third is a large-scale, multicentered, prospective evaluation.

Back to Article Outline

Data Extraction and Analysis 

Potentially eligible articles were reviewed by 3 of the authors, with a senior reviewer resolving disagreements. The operating characteristics of the clinical prediction rules of pulmonary embolism were summarized with likelihood ratios (LRs) and 95% confidence intervals (CIs), and κ and weighted κ values are reported as measures of precision for clinical prediction rules. Pooled LRs were summarized with random-effects measures.

Back to Article Outline

Main Results 

Performance of Clinical Gestalt 

From 1,709 articles identified using the search strategy, 7 studies were found that assessed physicians' clinical gestalt in judging the pretest probability of pulmonary embolism. Of these studies, the physician's clinical gestalt of low, moderate, and high pretest probabilities showed that the actual rates of pulmonary embolism ranged from 8% to 19%, 26% to 47%, and 46% to 91%, respectively. Three of the 7 studies (Prospective Investigation of Pulmonary Embolism Diagnosis,2 Sanson et al,3 and Musset et al)4 used inpatients and outpatients, with the clinical gestalt determined by either the attending physician caring for the patient or the physician in charge of the patient. Perrier et al5, 6, 7 had internal medicine residents working in emergency departments to predict pretest probabilities according to their clinical gestalt. Last, pulmonary specialists in the study completed by Miniati et al8 determined the clinical gestalt.

Performance of Clinical Prediction Rules 

In derivation studies, the clinical prediction rules divided patients suspected of having pulmonary embolism into low (3% to 10%), moderate (26% to 46%), and high (63% to 98%) pretest probabilities. Pooled LRs for each category of predicted risk were low (0.12; 95% CI 0.05 to 0.31), moderate (1.1; 95% CI 0.76 to 1.6), and high (23; 95% CI 7.6 to 69). Table 1 compares the performance of each clinical prediction rule in assigning pretest probabilities in derivation cohorts.

Table 1. Pretest probabilities of pulmonary embolism of clinical decision rules.
SourcePatients, No.Pretest Probability CategoryPretest Probability, %LR (95% CI)
Wells et al, 1998 (extended)1,239Low30.17(0.12–0.25)
Moderate281.8(1.25–2.1)
High7817(11–27)
Miniati et al, 1999 (PISA-PED)750Unlikely60.05(0.03–0.10)
Possible460.99(0.75–1.3)
Verylikely9747(23–98)
High638.6(5.7–13)
Wicki et al, 2001 (Geneva Rule)986Low100.31(0.24–0.40)
Moderate381.7(1.5–1.9)
High8111(6.1–21)
Kline et al, 2002934Nonhigh13.30.64(0.56–0.73)
High42.13.0(2.4–3.8)
Miniati et al, 2003 (PISA-PED II)1,100Low40.07(0.04–0.11)
Moderate260.72(0.6–0.87)
High9866(31–137)

The extended Wells rule9 was tested for validation by 2 studies. Although the prospective validation study by Kruip et al10 confirmed the discriminatory power of the rule to determine pretest probability (low=4%; moderate=28%; high=63%), the validation study by Sanson et al,3 using a retrospective study design, found lower discriminatory power (28%, 39%, 46%).

Three studies attempted to validate the simplified Wells rule. Sanson et al,3 again using a retrospective methodology, found a lower (28%, 30%, 38%) discriminatory power than the original rule, whereas prospective validation studies were performed by Wells et al11 (1.3%, 16.2%, 40.6%) and Chagnon et al12 (12%, 40%, 91%). Chagnon et al12 also prospectively tested the Wicki (Geneva)13 clinical prediction rule and found a similar range of pretest probabilities of pulmonary embolism (13%, 38%, 67%).

Table 2 compares the results of prospective validation studies, reported for the extended Wells rule,9 simplified Wells rule,14 and Wicki (Geneva).13 The Prospective Investigtive Study of Acute Pulmonary Embolism Diagnosis and Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis II rules developed by Miniati et al15, 16 and the clinical prediction rule by Kline et al17 did not have prospective validation studies evaluated in this review.

Table 2. Comparisons of clinical decision rules with their validation studies.
Clinical Prediction RuleValidation StudyPosttest Probability, %LR (95% CI)
LowModerateHighLowModerateHigh
Simplified WellsWellsetal1.316.240.60.13(0.06–0.26)1.9(1.6–2.3)5.9(3.7–9.3)
Chagnonetal1240910.39(0.26–0.58)2.0(1.5–2.6)29(3.8–223)
Extended WellsKruipetal428630.15(0.07–0.33)1.5(1.01–2.2)5.85(3.51–9.74)
Wicki(Geneva)Chagnonetal1338670.44(0.30–0.65)1.8(1.4–2.3)5.8(1.8–19.0)

Precision of Clinical Prediction Rules 

Interobserver variability was not measured in any of the studies of clinical gestalt. High agreement for the clinical prediction rules was identified for Wells et al9 (κ=0.86) and Kline et al17 (κ=0.81).

D-dimer Assays 

Results are reported for 3 individual studies of differing subtypes of D-dimer assays, with positive and negative LRs ranging from 1.7 to 2.7 and 0.09 to 0.22, respectively.

Back to Article Outline

Conclusions 

Although clinical assessment alone is insufficient to rule in or rule out pulmonary embolism, both the clinical gestalt of experienced clinicians and clinical prediction rules can assign a pretest probability with reasonable accuracy, though given their consistent performance, inexperienced clinicians are encouraged to use clinical prediction rules. Negative D-dimer assays increasingly appear to be adequate for ruling out pulmonary embolism in low-pretest-probability patients.

Rational Clinical Examination Author Contact 


Jeffrey S. Ginsberg, MD, FRCPC

Department of Medicine

McMaster University

Hamilton, Ontario, Canada

E-mail ginsbrgj@mcmaster.ca

Back to Article Outline

Commentary: Clinical Implication 

In 2007, the United States can expect almost 100,000 new cases of pulmonary embolism.18 In addition, the combination of high mortality rate (5%19 to 25%20) and a widely varied presentation often leads to the consideration of pulmonary embolism in the setting of chief complaints from syncope to unexplained fever. The emergency physician's estimation of an individual patient's risk of pulmonary embolism allows for the calculation of posttest probability justifying treatment, further testing, or discharge. Clinical gestalt is central to the practice of emergency medicine, in which high patient turnover leaves little time for extensive testing and consultation. Unfortunately, clinical judgment is extremely variable and depends on training level, local practices, and individual physicians' personal experience. Clinical prediction rules use the experience of many physicians treating diverse patient groups to derive estimates of pretest probability, with the potential for greater precision than any one physician's clinical gestalt.

This rational clinical examination installment addresses the value of clinical gestalt and clinical prediction rules in estimating the pretest probability of pulmonary embolism and attempts to address the value of D-dimer for detection of pulmonary embolism. Clinical gestalt of experienced internal medicine and chest physicians had similar results to clinical decision rules; for example, from the Prospective Investigation of Pulmonary Embolism Diagnosis study, the calculated LRs for high (5.3; 95% CI 4.0 to 8.0) intermediate (1.1; 95% CI 1.0 to 1.2), and low (0.3; 95% CI 0.2 to 0.4) risk.2 Of the 7 studies of clinical gestalt for pulmonary embolism, none tested implementation of the decision rules using emergency physicians. In addition, for the 5 clinical decision rules covered, only the extended Wells rule9 and simplified Wells rule14 had more than 1 validation study, and these demonstrated widely varying results, making conclusions difficult and reliability and accuracy difficult to predict (Table 2). In particular, the Wells group (using their own rule) was able to successfully demonstrate both accuracy and high sensitivity with an internally validated decision rule although there remains a need for external validation.

The utility of D-dimer testing to diagnose pulmonary emboli was also briefly discussed. The author chose 3 studies evaluating D-dimer assays to review, each an example of a separate assay type (latex agglutination, enzyme-linked immunosorbent assay, and whole-blood agglutination tests). Because of the limited nature of these data (2 medical record reviews and a single prospective investigation), firm conclusions cannot be drawn from this review. For a more detailed review of D-dimer testing for pulmonary emboli, high-quality systematic reviews are available.21, 22

Given the high stakes and variability in presentation of the disease, emergency physicians await prospective external validation of clinical gestalt and decision rules in multiple ED settings. Until then, emergency physicians will have to continue to use a combination of clinical judgment and the selective and cautious application of decision rules for their patients in whom pulmonary embolism is suspected.

Back to Article Outline

Take-Home Message 

When suspecting a patient with pulmonary embolism, clinical gestalt should be buttressed with any of these clinical decision rules to improve the accuracy and precision of estimating pretest probability of disease. Combining this with an algorithm approach using D-dimer testing in low-risk patients in the ED has been shown to result in excellent outcomes.

EBEM Commentator Contact 


Richard Sinert, DO

Matthew Foley, MD

Department of Emergency Medicine

SUNY-Downstate Medical Center

Brooklyn, NY

Back to Article Outline

EBEM Teaching Point 

The hierarchy and utility of clinical decision rules. The Evidence-Based Medicine Working Group in 2000 set forth a users guide to assessing the utility of clinical decision rules.23 The authors defined a hierarchy of evidence for clinical decision rules. Level 4, the lowest assessment, is for rules that have yet to be externally validated and thus should not be applied clinically. Level 3 includes rules that have been prospectively validated in only a narrow range of patients. These rules may be used clinically if the patient population is similar to the validation group. Level 2 denotes clinical rules prospectively validated in a variety of clinical settings. This level of confidence allows level 2 rules to be used in a broad spectrum of patients. Finally, level 1 refers to rules that have been prospectively validated as in the level 2 group, with the addition of an impact analysis that demonstrates both a change in physician behavior and a clear clinical benefit.

Back to Article Outline

References 

  1. Chunilal SD, Eikelboom JW, Attia J, et al. Does this patient have pulmonary embolism?. JAMA. 2003;290:2849–2858
  2. The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). The PIOPED Investigators. JAMA. 1990;263:2753–2759
  3. Sanson BJ, Lijmer JG, MacGillavry MR, et al. Comparison of a clinical probability estimate and two clinical models in patients with suspected pulmonary embolism (ANTELOPE-Study Group). Thromb Haemost. 2000;83:199–203
  4. Musset D, Parent F, Meyer G, et al. Diagnostic strategy for patients with suspected pulmonary embolism: a prospective multicentre outcome study. Lancet. 2002;360:1914–1920
  5. Perrier A, Bounameaux H, Morabia A, et al. Diagnosis of pulmonary embolism by a decision analysis-based strategy including clinical probability, D-dimer levels, and ultrasonography: a management study. Arch Intern Med. 1996;156:531–536
  6. Perrier A, Desmarais S, Goehring C, et al. D-dimer testing for suspected pulmonary embolism in outpatients. Am J Respir Crit Care Med. 1997;156(2 pt 1):492–496
  7. Perrier A, Desmarais S, Miron MJ, et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet. 1999;353:190–195
  8. Miniati M, Prediletto R, Formichi B, et al. Accuracy of clinical assessment in the diagnosis of pulmonary embolism. Am J Respir Crit Care Med. 1999;159:864–871
  9. Wells PS, Ginsberg JS, Anderson DR, et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med. 1998;129:997–1005
  10. Kruip MJ, Slob MJ, Schijen JH, et al. Use of a clinical decision rule in combination with D-dimer concentration in diagnostic workup of patients with suspected pulmonary embolism: a prospective management study. Arch Intern Med. 2002;162:1631–1635
  11. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med. 2001;135:98–107
  12. Chagnon I, Bounameaux H, Aujesky D, et al. Comparison of two clinical prediction rules and implicit assessment among patients with suspected pulmonary embolism. Am J Med. 2002;113:269–275
  13. Wicki J, Perneger TV, Junod AF, et al. Assessing clinical probability of pulmonary embolism in the emergency ward: a simple score. Arch Intern Med. 2001;161:92–97
  14. Wells PS, Anderson DR, Rodger M, et al. Derivation of a simple clinical model to categorize patients' probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost. 2000;83:416–420
  15. Miniati M, Prediletto R, Formichi B, et al. Accuracy of clinical assessment in the diagnosis of pulmonary embolism. Am J Respir Crit Care Med. 1999;159:864–871
  16. Miniati M, Monti S, Bottai M. A structured clinical model for predicting the probability of pulmonary embolism. Am J Med. 2003;114:173–179
  17. Kline JA, Nelson RD, Jackson RE, et al. Criteria for the safe use of D-dimer testing in emergency department patients with suspected pulmonary embolism: a multicenter US study. Ann Emerg Med. 2002;39:144–152
  18. White RH. The epidemiology of venous thromboembolism. Circulation. 2003;107(23 suppl 1):I4–I8
  19. Calder KK, Herbert M, Henderson SO. The mortality of untreated pulmonary embolism in emergency department patients. Ann Emerg Med. 2005;45:302–310
  20. Cushman M, Tsai AW, White RH, et al. Deep vein thrombosis and pulmonary embolism in two cohorts: the longitudinal investigation of thromboembolism etiology. Am J Med. 2004;117:19–25
  21. Brown MD, Rowe BH, Reeves MJ, et al. The accuracy of the enzyme-linked immunosorbent assay D-dimer test in the diagnosis of pulmonary embolism: a meta-analysis. Ann Emerg Med. 2002;40:133–144
  22. Stein PD, Hull RD, Patel KC, et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med. 2004;140:589–602
  23. McGinn TG, Guyatt GH, Wyer PC, et al. Users' guides to the medical literature: XXII: how to use articles about clinical decision rules (Evidence-Based Medicine Working Group). JAMA. 2000;284:79–84

PII: S0196-0644(07)01299-1

doi:10.1016/j.annemergmed.2007.07.010

Annals of Emergency Medicine
Volume 52, Issue 1 , Pages 76-79, July 2008

Article Tools