Annals of Emergency Medicine
Volume 35, Issue 2 , Pages 181-187, February 2000

Clots in the lung☆

Department of Emergency Medicine Albert Einstein College of Medicine Bronx, NY

Article Outline

Abstract 

[Gallagher EJ. Clots in the lung. Ann Emerg Med. February 2000;35:181-187.]

 

See related articles, p. 121 and p. 168.

Back to Article Outline

Clinical probability estimates 

Although pulmonary embolism (PE) is not a clinical diagnosis,1, 2 there is good evidence that clinicians can accurately classify patients presenting to the ED with suspected PE into low-, intermediate-, and high-probability categories. This stratification appears to be driven by some intuitive amalgam of risk factors, symptoms, signs, findings on chest film, and likelihood of alternative diagnoses.3, 4, 5

In one large, well-conducted trial of more than 900 consecutive ED patients, this strategy produced an intercategory gradient for prediction of PE that was both clinically and statistically significant3: (1) 9% (95% confidence interval [CI] 6% to 13%) of the low clinical probability group (defined as an estimated pretest probability of 0% to 20%) eventually had a final diagnosis of PE; (2) 38% (95% CI 30% to 47%) of the intermediate clinical probability group (defined as an estimated pretest probability of 21% to 79%) had PE; and (3) 64% (95% CI 49% to 77%) of the high clinical probability group (defined as an estimated pretest probability of 80% to 100%) had PE. The overall prevalence of PE was 24% (95% CI 20% to 27%).3 These data indicate that clinical estimates of PE among emergency department patients, although imprecise, tend to be directionally and categorically accurate. This observation lays the foundation for development of an evidence-based algorithm outlining a serial diagnostic testing strategy for PE, as shown in the Figure.

  • View full-size image.

  • Figure. Suggested serial diagnostic testing strategy for patients presenting to the ED with varying clinical pretest probabilities of pulmonary embolism (PE). a, Conservative assessment of evidence supports use of VIDAS DD D -dimer assay in this setting. b, Negative (–) VIDAS DD D -dimer assay in low clinical pretest probability patients essentially excludes PE (<1% posttest probability). c, Evidence supporting use of negative VIDAS DD D -dimer result in intermediate clinical pretest probability patients for exclusion of PE appears valid; however, because the precision of the LR – cannot be estimated, this test cannot be recommended at this time as unequivocally safe among intermediate clinical probability patients. d, Negative VIDAS DD D -dimer result does not exclude PE among high clinical pretest probability patients. e, A/C denotes anticoagulation for treatment of venous thromboembolic disease. f, Patients with underlying cardiopulmonary disease, especially chronic lung disease, may bypass radionuclide lung scanning and proceed directly to sCTA. g, See text for indications for ventilation scanning after a perfusion study. h, Negative lung scan finding indicates a reading of normal or near-normal. i, (±) finding on lung scan indicates a clinically indeterminate reading (ie, neither normal/near-normal nor high-probability [very low, low, intermediate, and so forth]). j, Positive (+) lung scan indicates a high probability reading. k, sCTA denotes spiral (helical) CT angiography with contrast, not requiring catheterization. l, A plausible alternative diagnosis seen on sCTA is currently the only evidence-based means of excluding PE with this test. m, Angio denotes conventional pulmonary angiography, requiring catheterization.

This stepwise approach begins with a clinical estimate of the patient’s pretest probability as low, intermediate, or high. Then, depending on the results of each test obtained in sequence, patients have a succession of tests: a D -dimer assay, lower extremity compression ultrasonography, radionuclide lung scanning, and spiral computed tomographic (CT) angiography. For those individuals still remaining at the bottom of the algorithm, conventional pulmonary angiography is indicated as the diagnostic criterion standard.

Back to Article Outline

D -Dimer assays 

These assays, which measure fibrin degradation products, comprise a large and heterogenous group of blood tests of highly varying quality.6, 7, 8, 9, 10, 11, 12 For clinicians to know whether these assays are sufficiently accurate to be used in the care of their patients, they must first know which D -dimer test is performed by their hospital laboratory. Currently, there is at least one assay on the market that appears to meet the 2 conditions necessary for ED application: provision of results in a time frame consistent with the cadence of emergency practice, and demonstration of sufficiently powerful negative likelihood ratios (LR–) to be clinically useful as a screening test. This is the VIDAS DD assay (Biomerieux, France), a quantitative, fully automated rapid enzyme-linked immunosorbent assay (ELISA).3, 10, 13 Combining raw data from the meticulous systematic review by Kline et al14 in this issue of Annals, with information from other sources facilitated cumulative meta-analytic examination of several D -dimer assays.15, 16, 17 This in turn provided the foundation for an evidence-based algorithm supporting use of the VIDAS DD assay as an accurate means of excluding PE in selected ED patients (Figure).

In addition to the VIDAS DD assay, there is a second, well-studied D -dimer assay, the SimpliRED (Agen Diagnostics Ltd, Brisbane, Australia) worth noting. This is a qualitative RBC agglutination assay that can be performed at the bedside within minutes.11, 18, 19, 20, 21, 22 Kline et al14 recommend use of the latter to exclude PE in low clinical probability patients.14 However, the well-conducted observational cohort study of Farrell et al23 also published in this issue of Annals , appears to provide sound data that not only contradict Kline et al’s recommendation, but warn against general use of the SimpliRED assay to exclude the diagnosis of PE in all ED patients. These divergent conclusions can be reconciled by examination of the low pretest probability subset of patients with suspected PE in Farrell et al’s study. Among these patients, the SimpliRED assay showed a sensitivity of 100% (95% CI 48% to 100%), and an LR– of 0 (95% CI 0.0 to 0.5),23 which is entirely consistent with Kline et al’s recommendation. However, because Farrell et al’s subgroup contains a very small sample (N=29), the estimates are imprecise and must be combined15, 16, 17 with other data to obtain a better sense of this test’s performance within this low-risk stratum.11, 18, 19, 20, 21, 22, 23 When examined in this fashion, the summary LR– emerging from a meta-analysis of the SimpliRED assay appears to be less than 0.1. Nevertheless, possible technical problems with this D -dimer assay,24 combined with lack of rigorous quality control and interobserver discordance inherent in most point-of-care testing, argue for caution in use of the SimpliRED assay to exclude such a high-stakes diagnosis as PE. In contrast, the combination of a low clinical pretest probability and a negative VIDAS DD assay result (<500 μg/L) essentially excludes PE by driving the posttest probability below 1% (Figure).

In contrast to the low-probability patients, use of any negative D -dimer result to exclude PE in patients judged to have high clinical pretest probability is not supported by evidence at this time.

The application of negative D -dimer assay results to patients categorized as intermediate clinical pretest probability is more complex than either the low- or high-probability strata. Although the SimpliRED assay is inaccurate in patients judged to have an intermediate clinical probability of PE,23 a negative VIDAS DD assay result makes PE highly unlikely in this subgroup. Unfortunately, there is insufficient information available in published sources to calculate the precision of the estimated posttest probability of PE among intermediate clinical pretest probability patients with a negative VIDAS DD. Therefore, the evidence base is not yet sufficiently solid to recommend use of any D -dimer assay to exclude PE among intermediate or high clinical pretest probability patients (Figure).

None of the D -dimer assays currently on the market have a sufficiently high positive likelihood ratio (LR+) to make these tests of much value when results are positive. This is particularly true among the elderly, many of whom have comorbid conditions that produce false-positive results.25

Back to Article Outline

Compression ultrasonography 

Depending on institutional custom, resource availability, and presence of concomitant pulmonary disease, one may choose to perform either radioisotope lung scanning or sonographic evaluation of the lower extremities as the next step in the workup of suspected PE. In the algorithm, leg sonography is shown next because it is noninvasive, less expensive, and faster.4, 26

Serial B-mode ultrasonography of the lower extremities, assessing probe compressibility of the external iliac, common femoral, and popliteal veins, is highly accurate and superior to impedance plethysmography in the identification of proximal thrombi in outpatients with symptomatic deep vein thrombosis (DVT).27 Limitations of ultrasonography include its insensitivity to distal (calf) DVT, the need for repeated testing among patients with initially negative results, and, of greatest concern, the identification of DVT in less than half of patients with subsequently proven PE.28, 29 Addition of color Doppler to provide Duplex scanning capability30 fails to generate an LR– that is sufficiently powerful to exclude PE among patients lacking symptoms of DVT.28, 29, 30 Therefore, the utility of ultrasonography in this setting is tied to its LR+, which is such that positive ultrasound findings would drive all remaining patients out of the algorithm and provide sufficient indication for anticoagulation. A single negative ultrasound finding, in contrast, is of little help, leaving postultrasound probabilities not substantially changed from pretest status (Figure).

Back to Article Outline

Radionuclide lung scanning 

Most patients remaining in the algorithm up to this point should have perfusion scintigraphy performed. The clinical value of lung scans is compromised by technical limitations of the procedure coupled with liberal use of a richly ambiguous terminology in their interpretation. Because of this, classification of lung scans into 3 categories (normal/near-normal, high probability, and nondiagnostic [all others]), each stratified by LRs, seems to be the most clinically sensible approach.31 A normal/near-normal lung scan is used for its sensitivity of 98% (95% CI 95% to 99%) and LR– of 0.10 (95% CI 0.04 to 0.24).32 Because of the imprecision of the LR–, as reflected by the width of its CI, a conservative evidence-based strategy argues against use of a normal scan to exclude PE among patients with a high clinical pretest probability, negative ultrasound and D -dimer results notwithstanding. In all other patients, a normal scan should essentially remove PE from further consideration (Figure).

In contrast, a high-probability lung scan, which has a specificity of 98% (95% CI 96% to 99%) and LR+ of 17 (95% CI 10 to 29),32 is sufficient to confirm the diagnosis of PE in virtually all patients.

Alternative scan interpretations (very low, low, intermediate, indeterminate probability, and so on), which constitute the majority of readings, have LRs that range between 0.40 and 1.1, thus providing no clinically useful opportunities for revision of disease probability.31, 32

Although findings on most perfusion scans among patients with suspected PE will be abnormal,33, 34 a ventilation scan is indicated only if one or more of the perfusion defects is segmental or larger. This is because subsegmental defects alone, whether matched or mismatched on ventilation studies, preclude obtaining either a normal/near-normal or a high-probability reading. Because no other scan result is clinically meaningful, it is pointless to pursue a ventilation study on individuals with small perfusion defects.33, 35

Perfusion lung scanning is not indicated if findings on chest film suggest that a normal or high-probability scan is unlikely. Usually this is due to cardiopulmonary disease, typically chronic obstructive pulmonary disease (COPD). The evidence base for omission of the lung scan among these patients derives from the Prospective Investigation Of Pulmonary Embolism Diagnosis (PIOPED) data,32 showing that approximately 90% of patients with COPD and approximately 80% of those with any cardiopulmonary disease had nondiagnostic scans.36

In settings where a lung scan is either nondiagnostic or is likely to be so, a pulmonary spiral CT angiogram with contrast is the procedure of choice.

Back to Article Outline

Spiral CT Angiography 

Spiral (helical) CT angiography (sCTA) is, along with the D -dimer assay, the other major advance in the diagnosis of PE in recent years. sCTA is analyzed extensively in the systematic review by Kline et al.14 It is indicated in patients with (1) lung scans that are neither normal/near-normal nor high-probability (ie, nondiagnostic); (2) normal/near-normal lung scans in the setting of a high clinical pretest probability; and (3) abnormalities seen on chest film that make a nondiagnostic lung scan likely.32, 36

Using the pooled sensitivity and specificity from Kline et al’s14 raw data to calculate adjusted specificity and sensitivity, respectively,15, 16 suggests that sCTA provides an LR+ of at least 8 and an LR– of about 0.2. Another meta-analysis of sCTA, using the same approach to data aggregation,15, 16 estimated a similar LR–, but found a much larger LR+ ranging from 13 to 35, depending on classification of indeterminate readings.37

Several additional pieces of clinically important information are pertinent to the interpretation of sCTA LRs. The first is that a substantial proportion of the insensitivity of sCTA is attributable to inability to identify peripheral emboli (ie, subsegmental clot, lodged distal to segmental branches of the pulmonary arterial tree).38 However, reanalysis of PIOPED data32 provides some encouraging information, suggesting that the LR– of 0.2 for SCTA (caused mainly by a relatively low sensitivity) may be less of a practical clinical problem than previously thought. This is because isolated subsegmental PEs appear to be uncommon, representing only about 6% of all PEs (95% CI 4% to 9%).39

A second mitigating feature is that difficulties in detection of isolated peripheral PE are not confined to sCTA alone, but are also shared by the criterion standard of conventional invasive pulmonary angiography. Indeed, another recent reanalysis of PIODPED32 found that when isolated subsegmental defects were the sole basis for the final radiographic diagnosis of PE, experienced angiographers disagreed in about 1 of every 3 patients.40

Finally, although there is substantial test referral bias evident in the use of sCTA, as reflected by the roughly 40% prevalence of PE seen in this subgroup of patients,14, 37 this represents the setting in which this test is most likely to be used in clinical practice.

One of the many advantages of sCTA compared with lung scanning is a marked reduction in indeterminate readings. In PIOPED, which enrolled more inpatients than outpatients, 72% of readings were neither normal/near-normal nor high-probability, and thus were of no clinical value.31, 32 This proportion of nondiagnostic lung scans decreased only slightly to 66% in a recent study that enrolled only ED patients.3 In contrast, a meta-analysis of sCTA that included 277 patients drawn from 8 publications reported inconclusive results on only 4% of sCTAs.37 As might be expected from this, a collateral advantage of sCTA is the improved interobserver concordance in readings (κ=0.85) compared with scintigraphy (κ=0.61).41

Data do not presently support use of sCTA to exclude PE because the LR– of a normal result appears weak and shows substantial interstudy variation.14, 37 However, newer methods of image acquisition42 and use of electron beam technology43 promise to improve the LR– of this test in the future. Pending improvement in the LR– of sCTA, the visualization of an alternative diagnosis is currently the only safe and reliable means of excluding PE with this technology. Estimates of the proportion of patients with suspected PE in whom sCTA excludes PE through identification of an alternative diagnosis are highly variable, ranging from a low of 10%44 to a high of 70%.45 Much of this may be attributable to variation in test referral bias.

Back to Article Outline

Clinical bottom line 

A highly conservative examination of current data (summarized in the Figure) suggests that 2 relatively recent diagnostic advances ought to inform any evidence-based approach to the diagnosis of PE in the ED. The first of these, the D -dimer VIDAS DD assay, resides near the top of the algorithm, and is useful as a screening test because of its LR–. The specific role of this assay is to exclude the diagnosis of PE among patients with low clinical pretest probabilities. The application of this assay to intermediate- and high-probability PE patients or the use of any other D -dimer assay to exclude PE among patients classified at any level of clinical pretest probability is not firmly supported by evidence at this time. The second major advance in PE diagnosis is sCTA, which resides near the bottom of the algorithm, just proximal to the criterion standard of invasive pulmonary angiography. In contrast to the D -dimer assay, sCTA is useful for its LR+ and for its ability to identify a convincing diagnostic alternative to PE. Just as an elevated or positive D -dimer level should not be used to confirm the diagnosis of a PE, normal or negative findings on sCTA should not be used to exclude this diagnosis.

The Figure summarizes a sequential evidence-based diagnostic strategy applicable to patients presenting to the ED with suspected PE. As the patients move down through each of 5 diagnostic strata, the number requiring conventional invasive angiography steadily decreases. Each stratum tends to contribute either to driving patients below the testing threshold (good LR–s), thus removing the diagnosis of PE from further consideration, or driving them above the treatment threshold (good LR+s), thus pushing the posttest probability high enough to warrant anticoagulation.46

The first stratum, the assessment of clinical pretest probability cannot, by itself, move patients out of the algorithm in either direction. However, it sets the stage for the application of D -dimer testing.

In the second stratum, the D -dimer assay, specifically the VIDAS DD, when applied to those patients with low clinical pretest probability, appears to be sufficient to exclude the diagnosis of PE with a negative result.

In the third stratum, any positive leg compression ultrasound finding among the remaining patients is sufficient to warrant anticoagulation for venous thromboembolic disease.

In the fourth stratum, normal/near-normal lung scan findings will exclude the diagnosis of PE among patients with low or intermediate pretest clinical probability. However, patients with high pretest probability with normal/near-normal lung scan findings should proceed to sCTA. Because the LR+ of a high-probability scan is even more powerful than the LR– of a normal one, anyone with a high-probability scan becomes a candidate for anticoagulation, independent of pretest probability, D -dimer results, or ultrasound findings.

In the fifth and final stratum of the Figure, a positive sCTA warrants anticoagulation. A negative sCTA result (ie, one that does not show a PE) can only reliably exclude this diagnosis if a plausible alternative explanation for the patient’s signs and symptoms is identified.

For those not driven out of the algorithm by any of the above, conventional invasive angiography will usually be necessary to obtain a definitive diagnosis.

In the future, it is likely that combinations of existing noninvasive diagnostic tests will be developed that capitalize on the multiplicative power of “chained” LRs.47 For example, a D -dimer assay result with an LR– of 0.1 and a low-probability lung scan with an LR– of 0.4 might be combined to produce an aggregate LR of 0.1 × 0.4 = 0.04. This would be sufficient to revise a patient with a pretest clinical probability as large as 20% (the equivalent of a pretest odds of 0.25) downward to a posttest odds of 0.25 × 0.04 = 0.01, which is equivalent to a posttest probability of slightly less than 1%.47 A 1% probability of disease is sufficiently low that most clinicians would consider this patient to have dropped below the testing threshold for PE and would pursue alternative diagnoses.46 This conclusion would not have been possible if the very same test results were used independently of each other rather than in combination. Because the chaining of LRs, although mathematically sound for short chains, is theoretical, it can only be used to generate plausible hypotheses. These must then be tested in clinical trials before any tandem diagnostic strategy can be safely introduced into clinical practice.

Back to Article Outline

References 

  1. Hull RD, Raskob GE, Ginsberg JS, et al.  A noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med. 1994;154:289–297
  2. Kelley MA, Carson JL, Palevsky HI, et al.  Diagnosing pulmonary embolism: New facts and strategies. Ann Intern Med. 1991;114:300–306
  3. Perrier A, Desmarais S, Miron J, et al.  Noninvasive diagnosis of venous thromboembolism in outpatients. Lancet. 1999;353:190–195
  4. 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
  5. Morabia A, Steinig-Stamm M, Unger PF, et al.  Applicability of decision analysis to everyday clinical work: a controlled feasibility trial. J Gen Intern Med. 1994;9:496–502
  6. Goldhaber SZ, Vaughn DE, Tumeh SS, et al.  Utility of cross-linked fibrin degradation products in the diagnosis of pulmonary embolism. Am Heart J. 1988;116:505–508
  7. Elms MJ, Bunce IH, Bundesen PG, et al.  Rapid detection of cross-linked fibrin degradation products in plasma using monoclonal antibody-coated latex particles. J Clin Pathol. 1986;85:360–364
  8. Freyburger G, Trillaud H, Labrouche S, et al.  D -dimer strategy in thrombosis exclusion: a gold standard study in 100 patients suspected of deep venous thrombosis or pulmonary embolism: 8 DD methods compared. Thromb Haemost. 1998;79:32–37
  9. Goldhaber SZ, Simons GR, Elliott G, et al.  Quantitative plasma D -dimer levels among patients undergoing pulmonary angiography for suspected pulmonary embolism. JAMA. 1993;270:2819–2822
  10. Janssen MCH, Heebels AE, de Metz M, et al.  Reliability of five rapid D -dimer assays compared to ELISA in the exclusion of deep venous thrombosis. Thromb Haemost. 1997;77:262–266
  11. Veitl M, Hamwi A, Kurtaran A, et al.  Comparison of four rapid D -dimer tests for diagnosis of pulmonary embolism. Thromb Res. 1996;82:399–407
  12. Wells PS, Anderson DR, Bormanis J, et al.  SimpliRed D -dimer can reduce the diagnostic tests in suspected deep vein thrombosis. Lancet. 1998;351:1405–1406
  13. de Moerloose P, Desmarais S, Bounameaux H, et al.  Contribution of a new, rapid, individual and quantitative automated D -dimer ELISA to exclude pulmonary embolism. Thromb Haemost. 1996;75:11–13
  14. Kline JA, Johns KL, Colucciello SA, et al.  New diagnostic tests for pulmonary embolism. Ann Emerg Med. 2000;35:168–180
  15. Irwig LI, Tossteson ANA, Gatsonis C, et al.  Guidelines for meta-analyses evaluating diagnostic tests. Ann Intern Med. 1994;120:667–676
  16. Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations. Stat Med. 1993;12:1293–1316
  17. Lau J, Antman EM, Jimenez-Silva J, et al.  Cumulative meta-analysis of therapeutic trials for myocardial infarction. N Engl J Med. 1992;327:248–254
  18. Reber G, de Moerloose P, Coquoz C, et al.  Comparison of two rapid D -dimer assays for the exclusion of venous thromboembolism. Blood Coagul Fibrinolysis. 1998;9:387–388
  19. Turkstra F, van Beek EJR, ten Cate JW, et al.  Reliable, rapid blood test for the exclusion of venous thromboembolism in symptomatic outpatients. Thromb Haemost. 1996;76:9–11
  20. Ginsberg JS, Wells PS, Brill-Edwards P, et al.  Application of a novel and rapid whole blood assay for D -dimer in patients with clinically suspected pulmonary embolism. Thromb Haemost. 1995;73:35–38
  21. Egermayer P, Town GI, Turner JG, et al.  Usefulness of D -dimer, blood gas and respiratory rate measurements for excluding pulmonary embolism. Thorax. 1998;53:830–834
  22. Ginsberg JS, Wells PS, Kearon C, et al.  Sensitivity and specificity of a rapid whole-blood assay for D -dimer in the diagnosis of pulmonary embolism. Ann Intern Med. 1998;129:1006–1011
  23. Farrell S, Hayes T, Shaw M. A negative SimpliRED D -dimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med. 2000;35:121–125
  24. Mauron T, Baumgartner I, Z’Brun A, et al.  SimpliRED D -dimer assay: comparability of capillary and citrated venous whole blood, between assay variability, and performance of the test for exclusion of deep vein thrombosis in symptomatic outpatients. Thromb Haemost. 1998;79:1217–1219
  25. Tardy B, Tardy-Poncet B, Viallon A, et al.  Evaluation of D -dimer ELISA test in elderly patients with suspected pulmonary embolism. Thromb Haemost. 1998;79:38–41
  26. Perrier A, Buswell L, Bounameaux H, et al.  Cost-effectiveness of noninvasive diagnostic aids in suspected pulmonary embolism. Arch Intern Med. 1997;157:2309–2316
  27. Heijboer H, Buller HR, Lensing AW, et al.  A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med. 1993;329:1365–1369
  28. Turkstra F, Kuijer PM, van Beek EJ, et al.  Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med. 1997;126:775–781
  29. 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
  30. Davidson BL, Elliott CG, Lensing AW. Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. Ann Intern Med. 1992;117:735–738
  31. Jaeschke R, Guyatt GH, Sackett DL, for the Evidence-Based Working Group . Users’ guide to the medical literature III. How to use an article about a diagnostic test. B. What are the results and how will they help me in caring for my patients?. JAMA. 1994;271:703–707
  32. The PIOPED Investigators . Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA. 1990;263:2753–2759
  33. Oudkerk M, van Beek EJR, van Putten WLJ, et al.  Cost-effectiveness analysis of various strategies in the diagnostic management of pulmonary embolism. Arch Intern Med. 1993;153:947–954
  34. Van Erkel AR, van Rossum AB, Bloem JL, et al.  Spiral CT angiography for suspected pulmonary embolism: a cost-effectiveness analysis. Radiology. 1996;201:29–36
  35. Kuiker PMM, Turkstra F, van Beek EJR, et al.  A survey of the diagnostic and therapeutic management of patients with suspected pulmonary embolism in The Netherlands. Neth J Med. 1997;50:261–266
  36. Lesser BA, Leeper KV, Stein PD, et al.  The diagnosis of acute pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest. 1992;102:17–22
  37. Vo TT, Jackson RE. Helical CT as a diagnostic tool for pulmonary embolism: a meta-analysis. [abstract] Acad Emerg Med. 1999;6:540
  38. Goodman LR, Curtain JJ, Mewissen MW, et al.  Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. AJR Am J Roentgenol. 1995;164:1369–1374
  39. Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest. 1997;111:1246–1248
  40. Stein PD, Henry JW, Gottschalk A. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology. 1999;210:689–691
  41. Grenier PA, Beigelman C. Spiral computed tomographic scanning and magnetic resonance angiography for the diagnosis of pulmonary embolism. Thorax. 1998;53(Suppl 2):S25–S31
  42. Remy-Jardin M, Remy J, Artaud D, et al.  Peripheral pulmonary arteries: optimization of the spiral CT acquisition protocol. Radiology. 1997;204:157–163
  43. Teigen CL, Maus TP, Sheedy PF, 2nd , et al.  Pulmonary embolism: diagnosis with contrast-enhanced electron-beam CT and comparison with pulmonary angiography. Radiology. 1995;194:313–319
  44. Ferretti GR, Bosson JL, Buffaz PD, et al.  Acute pulmonary embolism: role of helical CT in 164 patients with intermediate probability at ventilation-perfusion scintigraphy and normal results at duplex US of the legs. Radiology. 1997;205:453–458
  45. van Rossum AB, Treurniet FEE, Kieft GJ, et al.  Role of spiral volumetric computed tomographic scanning in the assessment of patients with clinical suspicion of pulmonary embolism and an abnormal ventilation-perfusion lung scan. Thorax. 1996;51:23–28
  46. Pauker SG, Kassirer JP. The threshold approach to clinical decision-making. N Engl J Med. 1980;302:1109–1117
  47. Gallagher EJ. Clinical utility of likelihood ratios. Ann Emerg Med. 1998;31:391–397

☆ Reprints not available from the author.

PII: S0196-0644(00)70138-7

Annals of Emergency Medicine
Volume 35, Issue 2 , Pages 181-187, February 2000

Article Tools

* Image Usage & Resolution