| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Diagnostic Imaging, Kaiser Foundation Hospital, 3288 Moanalua Rd., Honolulu, HI 96814.
Received September 5, 2003;
accepted after revision November 25, 2003.
Address correspondence to H.-C. Yoon
(hyo-chun.yoon{at}kp.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. From July 2001 to December 2002, 755 patients underwent pulmonary CT angiography for the evaluation of acute pulmonary embolism. A rapid, fully automated quantitative D-dimer assay was obtained in more than half the patients. The electronic medical records of the patients were subsequently reviewed to analyze the negative predictive value of the D-dimer assay in the diagnostic workup of acute pulmonary embolism and to determine the outcome of the patients who had negative findings on both D-dimer assay and pulmonary CT angiography at 3-month follow-up.
RESULTS. Of the 755 patients who underwent pulmonary CT angiography, 666 (88.2%) had negative findings, 73 (9.7%) had positive findings, and 16 (2.1%) were indeterminate. A total of 426 patients underwent both pulmonary CT angiography and D-dimer level evaluation, and 84 of these had negative findings (< 0.4 µg/mL) on D-dimer assay. Eighty-two of the 84 patients with negative findings on D-dimer assay had negative findings on pulmonary CT angiography; two were indeterminate and both subsequently had low-probability ventilation–perfusion studies. Among patients with positive D-dimer assays, no one with a level between 0.4 and 1.0 µg/mL had pulmonary CT angiography with findings positive for pulmonary embolism.
CONCLUSION. A quantitative D-dimer assay was effective in excluding the need for pulmonary CT angiography and had high negative predictive value when the D-dimer level was less than 1.0 µg/mL.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
90%) of which are negative for pulmonary embolism. Because CT
examinations are costly and involve radiation and contrast exposure, an
inexpensive rapid adjunctive test with high negative predictive value would be
useful to diminish the number of negative pulmonary CT angiograms. Several previous studies have evaluated the utility of D-dimer assays in the exclusion diagnosis of venous thromboembolism [9–11]. However, to our knowledge, none have yet investigated the role of a rapid, highly sensitive, quantitative assay in determining the need for pulmonary CT angiography in the diagnostic workup for pulmonary embolism. The purpose of this retrospective study was to evaluate a quantitative D-dimer assay (STA-Liatest D-DI, Diagnostica Stago) with the hypothesis that a D-dimer level less than a certain value in patients suspected of acute pulmonary embolism could obviate pulmonary CT angiography.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All CT angiography was performed on an MDCT scanner (LightSpeed Plus, General Electric Medical Systems) with 0.8-sec helical rotation speed, 1.25-mm collimation, and a pitch of 3:1. Patients were injected with 100 mL of iopamidol (Isovue 370, Bracco Diagnostics) diluted with normal saline to 120 mL total volume at a rate of 3 mL/sec. Imaging began 20 sec after initiation of contrast infusion.
The pulmonary CT angiogram findings were interpreted as either positive, negative, or indeterminate (Fig. 1A, 1B, 1C). The findings were interpreted as positive if thrombus, appearing as a filling defect outlined by intravascular contrast material, was present in one or more pulmonary arteries. The examinations were interpreted as having negative findings if the pulmonary arteries showed normal enhancement without the presence of filling defects. Findings were indeterminate if they could not be characterized as either positive or negative because of technical factors such as poor contrast bolus; excessive respiratory or cardiac motion; and beam hardening artifacts caused by pacemaker leads, obesity, or inability to abduct and exclude the arms from the imaging volume.
|
|
|
After 18 months, we reviewed the electronic medical records of all patients undergoing pulmonary CT angiography at our institution. The results of pulmonary CT angiography and any other diagnostic imaging tests (ventilation–perfusion scans, lower extremity Doppler studies, pulmonary angiograms) ordered at the time of pulmonary embolism evaluation were noted. In addition, the results of all positive and indeterminate pulmonary CT angiograms were reviewed by one of the authors. If a discrepancy was present between the officially reported results and the findings of the reviewing author for the presence or absence of pulmonary embolism, the pulmonary CT angiogram was further reviewed independently by two other radiologists who were unaware of the D-dimer results and of the previous interpretations. Although it was not feasible to review the large number of negative pulmonary CT angiograms, all of the pulmonary CT angiograms in patients who had negative D-dimer assays were reviewed by one of the authors.
All quantitative serum D-dimer measurements were performed using a latex agglutination turbidimetric immunoassay method with a fully automated coagulation analyzer (STA-Compact, Diagnostica Stago). Serum D-dimer levels less than 0.4 µg/mL are considered normal as reported by the manufacturer (package insert, STA-Liatest D-Di) and confirmed at the Kaiser Permanente Southern California Regional Coagulation Laboratory.
The source of the patient referral was also evaluated. Three separate categories were delineated: patients seen in the emergency medical department and acute care centers of the health maintenance organization, inpatients or patients referred by hospital-based physicians and surgeons, and patients from intensive care units.
Finally, the electronic medical record was reviewed for all patients who had negative or indeterminate pulmonary CT angiograms to determine if any patients developed acute pulmonary embolism within 3 months of their initial CT study.
All statistical analyses were performed using SPSS version 10 (Statistical Package for the Social Sciences). Continuous variables were compared using the Student's t test. Nominal variables were evaluated using the chi-square test. Logistic regression was performed using the outcome of pulmonary CT angiography as the dependent variable with age, sex, referral source, and serum D-dimer value as the covariates.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1 presents the results of the pulmonary CT angiograms based on D-dimer values and includes the patients who did not undergo D-dimer evaluation. Of the total 755 CT pulmonary angiograms, 88% were negative for pulmonary embolism and 10% were positive. Only 2% were indeterminate. Among the 16 patients with indeterminate studies, 10 underwent additional imaging studies: six had low-probability ventilation–perfusion studies, one had an intermediate-probability ventilation–perfusion scan and a negative bilateral lower extremity venous Doppler sonogram, one had a negative pulmonary angiogram, and two had negative bilateral lower extremity venous Doppler sonograms. These 16 patients were excluded from subsequent analysis because the primary purpose of this study was to evaluate the utility of the D-dimer assay rather than to examine the accuracy of pulmonary CT angiography.
|
The remaining 739 patients were stratified into one of four groups on the
basis of their D-dimer values. The data and distribution of
positive and negative pulmonary CT angiograms are presented in
Table 2. Review of the data
showed that among patients with a negative D-dimer level, none had
a pulmonary CT angiogram positive for pulmonary embolism. Furthermore, among
the 159 patients with a positive serum D-dimer level (
0.4
µg/mL), no one with a level less than 1.0 µg/mL had a pulmonary CT
angiogram positive for pulmonary embolism. Overall, significant difference
occurred in the likelihood of a positive pulmonary CT angiogram based on this
stratification scheme (p < 0.001). Pairwise comparisons showed no
difference between group 1 (D-dimer < 0.4 µg/mL) and group 2
(D-dimer 0.4 but < 1.0 µg/mL) because no subject in
either group proved to have a positive pulmonary CT angiogram. A significant
difference occurred in the likelihood of a positive pulmonary CT angiogram
between subjects in group 2 (D-dimer 0.4 but < 1.0
µg/mL) and group 3 (D-dimer 1.0 µg/mL) (p <
0.001). No significant difference occurred in the likelihood of a positive
pulmonary CT angiogram between subjects in group 3 (D-dimer
1.0 µg/mL) and group 4 (no serum D-dimer level drawn)
(p = 0.24).
|
A total of 419 patients were evaluated with both a D-dimer level and a pulmonary CT angiogram. Using pulmonary CT angiography as the reference standard and a positive D-dimer threshold greater than or equal to 0.4 µg/mL, sensitivity was 100%, negative predictive value was 100%, specificity was 21%, and positive predictive value was 9%. If the positive D-dimer threshold is increased to greater than or equal to 1.0 µg/ml, sensitivity and negative predictive value remain 100%, but specificity rises to 62% and positive predictive value increases to 17%.
A significant difference occurred in the number of patients who underwent serum D-dimer assay on the basis of their referral source (p < 0.001). Patients seen in acute outpatient care settings who were much more likely to have a D-dimer level checked. However, no significant difference occurred in the distribution of patients with positive pulmonary CT angiograms (p = 0.19). Furthermore, among those patients who did not undergo a serum D-dimer assay, no significant difference occurred in the distribution of patients with positive pulmonary CT angiography based on their referral source (p = 0.55).
Logistic regression was performed to determine which covariates were significant using the outcome of the pulmonary CT angiogram as the dependent variable. Four variables were analyzed: age, sex, referral source (inpatient, intensive care unit, or emergency medicine department), and serum D-dimer value. Only the serum D-dimer value proved to be a significant covariate.
Finally, 3-month follow-up of all patients with D-dimer values less than 1.0 µg/ml was performed. None of these 247 patients proved to have subsequent acute pulmonary embolism, including the six patients with indeterminate pulmonary CT angiography results.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The results of this study have already had a significant impact on the algorithm for pulmonary embolism detection in our hospital. Patients whose medical history suggests that their D-dimer level is likely to be elevated (e.g., postoperative patients, patients on anticoagulation, patients with recent trauma) proceed directly to pulmonary CT angiography unless a contraindication is present. All others undergo serum D-dimer evaluation. Patients with a D-dimer value of less than 1.0 µg/mL do not undergo pulmonary CT angiography unless a high clinical index of suspicion is present for pulmonary embolism. We believe that until the high negative predictive value of a negative D-dimer assay shown in our study can be confirmed by prospective trials, imaging should be performed in patients with a high pretest probability for pulmonary embolism regardless of the D-dimer results.
This retrospective study has several limitations, and the data should be interpreted with some caution. Primary care and emergency room physicians at our hospital did not use a structured clinical assessment model for pulmonary embolism. Although all patients referred for pulmonary CT angiography in our study were suspected of having pulmonary embolism, the clinical pretest probability was not quantified, and the stratification of patients into low, moderate, and high probability is not reported. Despite this limitation, outcome analysis of all 247 patients with D-dimer levels less than 1.0 µg/mL at 3 months after the diagnostic workup for pulmonary embolism disclosed no patient with clinical evidence of venous thromboembolism or mortality attributable to pulmonary embolism.
Most subjects who did undergo serum D-dimer assay were outpatients, and those who did not were inpatients or intensive care patients. Referring physicians of the latter group often decided against ordering a D-dimer assay because they thought their patients would invariably have an elevated D-dimer level because of iatrogenic causes such as recent surgery, indwelling vascular catheters, and daily blood drawings for laboratory studies. This line of reasoning may be valid because no difference occurred in the distribution of positive pulmonary CT angiograms among patients with positive D-dimer levels and those in whom a D-dimer level was not evaluated (groups 3 and 4, respectively; Table 2). However, this circumstance resulted in a heavier weighting of the outpatient group to assess the value of the D-dimer assay for pulmonary embolism in this study, and the results of this study may best be applied to this group of patients.
This study focused on patients suspected of having acute pulmonary embolism. Patients with chronic pulmonary embolism may have lower D-dimer values caused by clot stabilization; therefore, the D-dimer cutoff value found in our study may not be applicable to patients evaluated for chronic pulmonary embolism.
Another caveat concerns the use of pulmonary CT angiography as the gold standard for the determination of pulmonary embolism in this study. Although catheter pulmonary angiography has been considered the gold standard for the detection of pulmonary embolism, catheter pulmonary angiography and pulmonary CT angiography have similar sensitivity and specificity when compared to necropsy in animal models of pulmonary embolism [14, 15]. Both imaging techniques also suffer from similar limitations in the evaluation of subsegmental pulmonary embolism [16, 17]. Outcome analyses have shown that it is safe to withhold anticoagulation in patients evaluated for pulmonary embolism who have negative CT angiography examinations and no other evidence of venous thromboembolism [18]. Hence, although it is possible that a few patients with both negative D-dimer assays and negative pulmonary CT angiograms could have had pulmonary embolism, the observation that none developed clinically apparent pulmonary embolism during the 3-month follow-up period would suggest that there were few, if any, false-negative studies.
A final important point that warrants emphasis is that no standardization exists among the several different commercially available D-dimer assays [19]. The different assays vary in sensitivity and specificity, largely because of variations in the monoclonal antibodies and the methods of detection and instrument calibration. Cutoff values used for one assay cannot simply be applied to a different assay. Comparison among assays is further hampered by the use of different units by different manufacturers: micrograms per milliliter, nanograms per milliliter, milligrams per milliliter, fibrinogen equivalent units, and D-dimer units.
Three different types of D-dimer assay are available: the enzyme-linked immunosorbent assay, the whole-blood agglutination assay, and the latex agglutination assay. Although widely accepted as the gold standard, the manual enzyme-linked immunosorbent assay is relatively expensive, time-consuming, usually performed in batches, and not widely available, making it an unrealistic candidate for widespread clinical use. The whole-blood agglutination assay is a simple qualitative study that requires only a drop of blood and can be performed at the bedside. However, optimal test reliability requires testing by a trained technician with strict adherence to recommended technique, and the negative predictive value of this test is highest and comparable to the quantitative assays when combined with a clearly delineated clinical assessment model [11].
Newer automated enzyme-linked immunosorbent and quantitative latex agglutination assays compare favorably with the manual enzyme-linked immunosorbent assay. In addition to being highly sensitive, these automated assays have a fast turnaround time of 1 hr or less, are available 24 hr a day, show good duplication of results around the normal and abnormal cutoff values, and are relatively immune to biologic interference from variables such as rheumatoid factor, lipemia, and hyperbilirubinemia. These assays have been shown to be sufficiently sensitive for use in the diagnostic evaluation of venous thromboembolism [12, 13].
The assay used in our study is based on the agglutination of a suspension of antibody-coated microlatex particles in the presence of plasma containing D dimer. The change in absorbency caused by agglutination of the microlatex particles, measured at 540 nm, is proportional to the concentration of D dimer. Assay time on STA analyzers is less than 10 min, with all procedures performed automatically including sample dilution and quality control. Unlike the previous latex agglutination assays, this system allows D dimer to be measured on a fully automated coagulation analyzer in true random access fashion.
In conclusion, for the fully automated quantitative D-dimer assay used in this study, a value less than 1.0 µg/mL effectively excluded the diagnosis of acute pulmonary embolism in patients seen in an acute care setting. The use of the D-dimer assay as the first step in the diagnosis of acute pulmonary embolism could have eliminated the need for up to 60% of all pulmonary CT angiograms.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  V. King, A. A. Vaze, C. S. Moskowitz, L. J. Smith, and M. S. Ginsberg D-Dimer Assay to Exclude Pulmonary Embolism in High-Risk Oncologic Population: Correlation with CT Pulmonary Angiography in an Urgent Care Setting Radiology, June 1, 2008; 247(3): 854 - 861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. D. Raptopoulos, P. B. Boiselle, N. Michailidis, J. Handwerker, A. Sabir, J. A. Edlow, I. Pedrosa, and J. B. Kruskal MDCT Angiography of Acute Chest Pain: Evaluation of ECG-Gated and Nongated Techniques Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S346 - S356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. T. Johnson and E. K. Fishman IV Contrast Selection for MDCT: Current Thoughts and Practice Am. J. Roentgenol., February 1, 2006; 186(2): 406 - 415. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
