DOI: 10.1148/radiol.2423051441
(Radiology 2007;242:889-897.)
© RSNA, 2007
Pulmonary Embolism: Prognostic CT Findings1
Philip A. Araoz, MD,
Michael B. Gotway, MD,
Jeffrey R. Harrington, MA,
W. Scott Harmsen, MS and
Jayawant N. Mandrekar, PhD
1 From the Departments of Radiology (P.A.A.) and Health Sciences Research, Division of Biostatistics (J.R.H., W.S.H., J.N.M.), Mayo Clinic, 200 First St SW, Rochester, MN 55905; and Scottsdale Medical Imaging, Scottsdale, Ariz (M.B.G.). Received August 26, 2005; revision requested November 2; revision received February 23, 2006; accepted April 3; final version accepted, June 8.
Address correspondence to P.A.A. (e-mail: paraoz{at}mayo.edu).
 |
ABSTRACT
|
|---|
Purpose: To retrospectively determine whether three computed tomographic (CT) findings—ventricular septal bowing (VSB), ratio between the diameters of right ventricle (RV) and left ventricle (LV), and embolic burden—are associated with short-term death, defined as in-hospital death or death within 30 days of CT, whichever was longer, due to acute pulmonary embolism (PE).
Materials and Methods: Institutional Review Board approval was obtained, and patient information was reviewed in compliance with HIPAA regulations. A total of 1193 patients with CT scans positive for PE from January 1, 1997, to December 31, 2002, who had given authorization for retrospective research were included. Scans were independently reviewed by two observers. CT findings were compared with risk of death by using univariate analysis (
2 statistic) and multivariate logistic regression. Interobserver variability (
statistic or intraclass correlation coefficient), sensitivity, and specificity of CT findings for predicting death were calculated. A third observer reviewed discrepant cases post hoc.
Results: Fifty-four percent of patients were women and 46% were men (mean age, 63 years ± 16). For observer 1, VSB was associated with death in univariate (odds ratio [OR], 1.98; P = .04) and multivariate modeling (OR, 1.97; P = .05). Interobserver variability was only fair ( = 0.54) for VSB, and observer 2 found no association with death (OR, 1.52; P = .22). For both observers, VSB had low sensitivity (21% and 18%) and high specificity (88% and 87%) for predicting death. Neither RV/LV diameter ratio nor embolic burden was associated with increased risk of death. For observer 3, VSB was associated with death in univariate (OR, 2.10; P = .05) and multivariate analyses (OR, 2.18; P = .05).
Conclusion: CT-depicted VSB is predictive of death due to PE, but with low sensitivity and high interobserver variability. RV/LV diameter ratio and embolic burden are not associated with short-term death due to PE.
© RSNA, 2007
|
INTRODUCTION
|
|---|
Acute pulmonary embolism (PE) is a disease with a variable prognosis. In recent years, many attempts have been made to stratify patients according to their risk of death due to PE. From the results of large retrospective registries (1,2) and randomized trials of thrombolytics (3), clinical presentation is considered the most powerful predictor of death due to PE. However, there have been attempts to further risk-stratify patients with imaging. In several small prospective studies, right-sided heart strain, as demonstrated with echocardiography, has been shown to be predictive of death in acute PE (4).
Given the promising results found with echocardiography, several investigators have attempted to determine prognosis from findings at computed tomography (CT). Authors of several retrospective studies have suggested that right-sided heart strain and embolic burden at CT are prognostic findings (5,6).
However, the studies to date have been small and have had limited multivariate analysis. Also, none of the existing CT studies have attempted to control for clinical presentation variables, which are the most powerful predictors of death. Thus, the purpose of our study was to retrospectively determine whether three CT findings—ventricular septal bowing, ratio between the diameter of right ventricle (RV) and that of the left ventricle (LV), and embolic burden—are associated with short-term death, which was defined as in-hospital death or death within 30 days of CT, whichever was longer, due to acute PE.
|
MATERIALS AND METHODS
|
|---|
Patient Selection
Institutional Review Board approval was obtained, and patient information was reviewed in compliance with Health Insurance Portability and Accountability Act regulations. Patients with CT scans coded positive for PE from January 1, 1997, to December 31, 2002, who had given authorization for retrospective research were identified. For patients with more than one scan, only their first CT scan was included so that a patient could not be included more than once. A total of 2249 patients were thus identified. The report of the CT scan was then reviewed by a nurse study coordinator from the radiology department's Study Coordinator Unit, which comprises approximately 12 nurses with expertise in coordinating research and chart abstraction. If the CT report indicated no PE, the scan was considered to have been miscoded as positive and the patient was excluded.
The remaining CT scans were reviewed by two board-certified radiologists (P.A.A., M.B.G., with 2 and 5 years experience, respectively, as attending-level thoracic radiologist) with fellowship training in thoracic radiology. Both observers reviewed the scans for PE, defined as the presence of a discrete low-attenuation filling defect in the pulmonary arterial tree. Both observers reviewed the scans for signs of chronic PE, defined as laminated thrombus adherent to the pulmonary arterial wall, recanalized thrombus, webs within the lumen of pulmonary arteries, and small tapered pulmonary arteries (7). Arbitrarily, the results of observer 1 were used to further exclude patients; if observer 1 detected no PE or signs of chronic PE, the patient was excluded. A total of 1056 of the original 2249 patients were excluded, leaving 1193 patients in the study (Table 1).
CT Acquisition
Of 1193 CT scans included, 1105 (93%) were for the clinical indication of PE and were obtained according to the PE protocol of our institution. The PE protocol requires contrast material injections at rates of at least 4 mL/sec, at least 100 mL of iodinated contrast material, a section thickness of 3 mm or less depending on the type of scanner (Table 2), and the use of either bolus timing or bolus tracking software for optimal opacification of pulmonary arteries. The remaining 88 CT scans were for indications other than PE and were obtained by using routine chest CT protocols, which at our institution involve contrast material injections (as deemed necessary) of 3 mL/sec, no bolus timing or tracking, and section thickness of 5 mm or less depending on the type of scanner.
CT Interpretation
CT scans were reviewed independently by the two observers who were blinded to the clinical history. CT scans were reviewed electronically by using commercially available software (E-film Medical, Toronto, Ontario, Canada). Each radiologist evaluated for the presence or absence of ventricular septal bowing, measured the RV/LV diameter ratio, and scored the embolic burden.
Ventricular septal bowing.—Ventricular septal bowing was subjectively judged as being present or absent. It was considered present if any image showed ventricular septal bowing Leftward toward the left ventricle (Fig 1).
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: Ventricular septal bowing in 73-year-old man. Transverse contrast material–enhanced chest CT scan shows that ventricular septum bows leftward (arrow) into the left ventricular lumen. Small pulmonary emboli are visible in left lower lobe basal segmental pulmonary arteries.
|
|
RV/LV diameter ratio.—The RV diameter was measured on the transverse section that showed the tricuspid valve at its widest. The RV diameter was measured from the inner wall to the inner wall at the widest point in the chamber, which was typically in the basal third of the RV. The LV diameter was measured on the transverse image that showed the mitral valve at its widest and, like the diameter of RV, was measured from the inner wall to the inner wall at the widest portion of the LV. The RV/LV diameter ratio was calculated (Fig 2).
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a: RV/LV diameter ratio. (a) Transverse contrast-enhanced chest CT scan in 78-year-old woman at level where the tricuspid valve is widest. RV diameter is measured (line) at this level from inner wall to inner wall. (b) LV diameter is measured (line) at the level where the mitral valve is widest. Small pulmonary emboli are visible in basal segmental pulmonary arteries bilaterally.
|
|
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b: RV/LV diameter ratio. (a) Transverse contrast-enhanced chest CT scan in 78-year-old woman at level where the tricuspid valve is widest. RV diameter is measured (line) at this level from inner wall to inner wall. (b) LV diameter is measured (line) at the level where the mitral valve is widest. Small pulmonary emboli are visible in basal segmental pulmonary arteries bilaterally.
|
|
Embolic burden.—Each observer scored the main, lobar, right interlobar, and segmental pulmonary arteries for the presence of emboli and also graded whether emboli were occlusive. Emboli were defined as low-attenuation filling defects within the pulmonary arterial tree. A previously reported embolic burden scoring system was applied (8,9). The scoring system was based on the number of segmental pulmonary arteries involved, with weight added to an occlusive thrombus (Fig 3). Thrombus in a segmental artery received a score of 1. Thrombus in more proximal arteries received a score equal to the number of segmental arteries supplied. If a thrombus in a vessel was occlusive, the score in that vessel was multiplied by two. For example, the right upper lobe pulmonary artery supplies three lung segments. Thus, an isolated thrombus in the right upper lobe pulmonary artery received a score of 3 and an occlusive thrombus received a score of 6. The maximal score was 36 for an occlusive thrombus in the main pulmonary artery.
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3: Embolic burden scoring system. Schematic of the pulmonary arterial tree with scores for nonocclusive emboli according to vessel. Emboli in a segmental pulmonary artery are given a score of 1. Emboli in more proximal pulmonary arteries are given a score based on the total number of segmental pulmonary arteries supplied.
|
|
Clinical Information
Outcome variable.—Nurses from the Radiology Study Coordinator Unit reviewed patient medical records for short-term death due to PE. The association of death with PE was determined from chart review by observers blinded to the CT interpretation. Patients with severe or terminal illness were considered to have died due to PE if it was the clinical impression that the patient would not have succumbed to the underlying illness at that time if not for the PE.
Patient comorbidities.—In addition to sex and age, charts were reviewed for 13 comorbid conditions: congestive heart failure, ischemic heart disease, pulmonary disease, pulmonary hypertension, asthma, malignancy, current anticoagulant use, organ transplantation, history of thromboembolic disease, current thromboembolic disease, diabetes, prior inferior vena cava filter, and immunocompromise.
Clinical presentation.—Patients' charts were also reviewed for six clinical variables available at the time the decision to order a CT scan was made: room air oxygen saturation, heart rate, blood pressure, whether the patient received supplemental oxygen without intubation, whether the patient was intubated prior to the clinical presentation that prompted CT scan, and whether the patient was intubated after the clinical event but prior to the CT scan. Also, patient treatment for the PE was recorded.
Post Hoc Analysis
Results of statistical analysis showed that ventricular septal bowing was significantly associated with death for observer 1 but not for observer 2. Among the 1193 patients in the study, observers 1 and 2 disagreed about ventricular septal bowing in 122 (10%) patients. To clarify whether ventricular septal bowing is associated with death, a third observer reviewed the findings for the 122 patients. Observer 3 was blinded to the clinical outcomes and to the results of observers 1 and 2. Observer 3 was a radiologist, fellowship trained in noninvasive vascular imaging with 1-year experience as an attending-level radiologist. For observer 3, the association of ventricular septal bowing and short-term death was determined by using the results of observers 1 and 2 in those cases when the observers agreed on the presence or absence of ventricular septal bowing and by using the observations of observer 3 when observers 1 and 2 disagreed.
Statistical Analysis
The end point for the analysis was short-term death due to PE, as defined earlier. CT findings, comorbid conditions, and clinical variables were compared with short-term death due to PE by using the 2 statistic. Age, heart rate, room air oxygen saturation, and blood pressure were converted into dichotomous variables (age > 65 years, heart rate > 100 beats per minute, room air oxygen saturation < 90%, systolic blood pressure < 100 mm Hg). Multivariate logistic regression was used to determine the incremental prognostic value of CT findings over clinical information. Interobserver variability for ventricular septal bowing was calculated by using the statistic. For continuous variables (RV/LV diameter ratio and embolic burden), intraclass correlation coefficients were calculated to determine interobserver variability. For predictors that showed an association with short-term death due to PE, analysis of sensitivity and specificity was performed. For this analysis, short-term death due to PE was considered the reference standard measure.
|
RESULTS
|
|---|
Patient Demographics
Fifty-four percent of patients were women and 46% were men. The mean age was 63 years ± 16. A large percentage (41%) of patients had a history of malignancy, which was the most common comorbid condition (Tables 3, 4).
Deaths
One hundred thirteen (9%) patients died of all causes in a hospital or within 30 days of the CT scan. Of these, 58 patients (5% overall) were thought to have died of PE.
CT Parameters and Interobserver Variability
Interobserver variability for ventricular septal bowing was fair ( = 0.54) as was interobserver variability for RV/LV diameter ratio (intraclass correlation coefficient = 0.51). Interobserver variability for embolic burden was very good (intraclass correlation coefficient = 0.85.) (Table 5).
Univariate Analysis: Comorbid Conditions, Clinical Presentation, and Treatment
In univariate analysis, age and seven of the 13 comorbid conditions were associated with increased risk of short-term death due to PE (Table 6). Of these, pulmonary disease was associated with the highest risk (odds ratio [OR], 2.56; P < .001). Five of the six clinical presentation variables were significantly associated with short-term death due to PE. Only heart rate greater than 100 beats per minute was not associated with increased risk of death (P = .10). Patients who received no treatment were by far more likely to die than were patients who received any treatment (OR, 5.43; P < .001). This was not surprising, as all patients in this study had known PE and treatment was typically withheld in terminally ill patients. Only 16 patients received thrombolytics, and none had died. Therefore, univariate analysis of this subgroup was not performed. Among the clinical presentation variables, intubation prior to the clinical event was associated with the highest risk (OR, 10.24; P < .001).
Univariate Analysis: CT Findings
Association of CT findings with short-term death due to PE was calculated independently for observers 1 and 2 (Table 7). Ventricular septal bowing was associated with death due to PE for observer 1 (OR, 1.98; P = .04) but not for observer 2 (OR, 1.52; P = .22). Unexpectedly, a higher embolic burden was associated with a very small but statistically significant decreased risk of death for observer 2 (OR, 0.95; P = .02). However, for observer 1, embolic burden was not associated with short-term death due to PE (P = .31).
Multivariate Analysis
Because of the low number of events (58 short-term deaths from PE), the strongest clinical presentation variables were combined into one event to allow for placement in the multivariate model. A new variable was created, called hemodynamic instability. Patients were considered to have hemodynamic instability if they had any one of the following: room air saturation less than 90%, systolic blood pressure less than 100 mm Hg, intubation prior to the clinical event that prompted the CT scan, or intubation after the clinical event but prior to the CT scan. For observer 1, a multivariate model was created by using age, hemodynamic instability, malignancy, pulmonary disease, and ventricular septal bowing. A similar model was created for observer 2. For observer 1, ventricular septal bowing remained significantly associated with short-term death due to PE, though the association was borderline (P = .05) (Table 8). For observer 2, ventricular septal bowing remained unassociated with short-term death due to PE (Table 9).
Sensitivity and Specificity of Ventricular Septal Bowing for Predicting Death Due to PE
In this analysis, patients with ventricular septal bowing who died were considered to have true-positive results and those who did not die were considered to have false-positive results. For both observers, ventricular septal bowing had low sensitivity but high specificity for predicting death (Table 10). The main difference between observers 1 and 2 was the number of false-positive results (patients with ventricular septal bowing who did not die). Observer 1 and observer 2 had a similar number of true-positive results (12 and 11 patients, respectively). However, observer 1 had fewer false-positive observations than did observer 2 (132 vs 151).
Post Hoc Analysis
Because of the discrepant results between observers 1 and 2, the cases for which these observers disagreed were selected and analyzed by a third observer. The univariate and multivariate analyses of ventricular septal bowing were performed by using the results of observers 1 and 2 in those cases when observer 1 and 2 agreed on the presence or absence of ventricular septal bowing and by using the observations of observer 3 when observers 1 and 2 disagreed. For observer 3, ventricular septal bowing was associated with short-term death in both univariate analysis (OR, 2.10; lower 95% confidence limit, 1.00; upper 95% confidence limit, 4.41; P = .05) and in multivariate analysis (OR, 2.18; P = .05, Table 11). Sensitivity and specificity analyses demonstrated that observer 3 also had a low sensitivity (16%) and high specificity (92%) for predicting death (positive predictive value, 9%; negative predictive value, 96%). Observer 3 had nine true-positive and 91 false-positive results.
View this table:
[in this window]
[in a new window]
|
Table 11. Post Hoc Multivariate Analysis of Ventricular Septal Bowing for Predicting Short-term Death Due to PE for Observer 3
|
|
|
DISCUSSION
|
|---|
The findings of this study are that ventricular septal bowing, as detected at CT, had a low sensitivity for predicting death due to PE and a high degree of interobserver variability ( = 0.54) and that neither the RV/LV diameter ratio nor embolic burden was associated with death due to PE.
Ventricular Septal Bowing
We found ventricular septal bowing at CT to be an inconsistent predictor of death due to PE. For observer 1, ventricular septal bowing was associated with short-term death even when the data were corrected for very powerful predictors such as clinical presentation. However, because of a small number of additional false-positive findings (19 additional patients with observed ventricular septal bowing who did not die), ventricular septal bowing was not associated with death for observer 2. Because of this discrepancy, a third observer analyzed the 122 cases in which observers 1 and 2 had disagreed. Using results of observer 3 as a tiebreaker, we found that the relationship between death and ventricular septal bowing persisted, which led us to believe that ventricular septal bowing reflects an underlying pathologic process that predisposes toward short-term death. However, detection of ventricular septal bowing at CT may be of limited practical value given the low sensitivity of ventricular septal bowing for predicting death (16%–21% for our observers), its high interobserver variability ( = 0.54), and the strength of existing clinical predictors.
Like our findings, the literature shows mixed results for the prognostic power of ventricular septal bowing. On the one hand, five of six prospective studies (1,3,10–12) have shown echocardiographically detected right-sided heart strain to be significantly associated with death due to PE. On the other hand, neither of the two studies (5,13) that explicitly examined ventricular septal bowing at CT found it to be predictive of death due to PE. However, both studies had far fewer patients than did our current study, and in both articles there was a suggestion of an association between PE and death that did not reach significance. Ventricular septal bowing detected with CT may be a predictor, but it may not be a very powerful one.
RV/LV Diameter Ratio
We found no association between the RV/LV diameter ratio and death. As opposed to ventricular septal bowing, CT findings reported in two articles (5,6) have shown the RV/LV diameter ratio to be predictive of death due to PE. Shoepf et al (6) retrospectively studied 431 patients who underwent four-detector CT and found the RV/LV diameter ratio to be predictive of a 30-day all-cause mortality with both univariate and multivariate analyses. Van der Meer et al (5) reviewed 120 patients and found the RV/LV diameter ratio to be predictive of death due to PE with univariate analysis but did not attempt multivariate analysis.
The reason for the discrepancy between these studies and our study appears to be that the two other CT studies had a consensus measurement of the RV/LV diameter and a more uniform CT acquisition. Both Schoepf et al and van der Meer et al interpreted images in consensus as opposed to individually, as was done in our study; this is important given the high interobserver variability we noted (intraclass correlation coefficient = 0.51). Also, both Schoepf et al and Van der Meer et al used a single type of a CT scanner, while in our study five types of CT scanners were used, and some patients were not scanned using a dedicated PE protocol. Schoepf et al reconstructed all the images into a four-chamber view, which is more suited to the measurement of RV/LV diameter than the transverse images used in our study. Patients in the van der Meer et al study were selected from a prospective cohort as opposed to a retrospective review of all cases, as was done in our study.
Nevertheless, given our findings that ventricular septal bowing is predictive by at least some observers and the preponderance of echocardiographic data, it seems possible that the CT-defined RV/LV diameter ratio may be a predictor of death due to PE, even though it was not found to be so in our study. Future studies should take care to account for the variability in the measurement of the RV/LV diameter ratio.
Embolic Burden
We found embolic burden not to be associated with an increased risk of death due to PE. Unlike ventricular septal bowing, our findings for embolic burden were unequivocal. For observer 2, a higher embolic burden was associated with a small but significant decreased risk of death (OR, 0.99; P = .02). Also, unlike ventricular septal bowing and the RV/LV diameter ratio, our measurement of embolic burden was very consistent between observers (intraclass correlation coefficient = 0.85). This study therefore strongly supports the idea that in patients with PE diagnosed at CT, the embolic burden is not associated with an increased risk of death.
It may seem paradoxical that an increased embolic burden could be associated with a decreased risk of death. However, this is likely due to the selection bias inherent in our study. Findings of autopsy studies performed primarily in the 1960s have shown that 36%–75% of patients with PE die less than 1 hour after the onset of symptoms (14–19). Those who survive long enough to undergo CT are a distinct subgroup. It may be that patients with large pulmonary emboli who survive to undergo CT are healthier than average patients.
Our finding that embolic burden is not associated with an increased risk of death is consistent with the bulk of the literature. From the early 1970s until very recently, no study had ever shown embolic burden, defined either with angiography or CT, to be predictive of mortality (13,20). Recently, however, two small CT studies did report a relationship between embolic burden and mortality. In a study of van de Meer et al (5), the embolic burden was significantly associated with death due to PE over 3 months. In a study of 59 patients by Wu et al (21), the embolic burden was also associated with death due to PE.
It appears that the difference between the previous studies and our study may be the patient populations. Even though van der Meer et al excluded hemodynamically unstable patients, their PE-related mortality was higher than ours (6% vs 5%) and their mean embolic burden was nearly twice that in our study. Likewise, Wu et al had a higher PE-related mortality (8%), though they did not report their mean embolic burden.
Limitations
Our study had limitations. First, the outcome of interest was PE-related death, which requires a subjective determination of how much a given PE contributed to death. This is difficult to do in a retrospective review. Moreover, in this study, the individual nurse study coordinators abstracting the charts and not an expert panel, as has been done in other studies, decided the association of PE with death. Second, the initial identification of cases for inclusion was performed by using codes that were imperfect. These codes were recorded by clinical radiologists at the time the CT was performed. Of the 2249 CT scans coded as positive, 517 (23%) had been read by the interpreting radiologist as negative for PE, meaning that some CT scans had been initially miscoded as positive. It is unknown how many scans with pulmonary emboli might have been coded as negative and were therefore not included in the study. Third, one of the exclusion criteria, chronic PE, required a subjective determination. Also, this exclusion means that the results of this study are not applicable to cases with chronic PE. Fourth, the majority (66%) of CT scans in our study were obtained with an electron-beam CT scanner. These scanners have slightly higher temporal resolution and lower spatial resolution than do the current state-of-the-art multidetector scanners. It is unclear what effect these differences might have on the applicability of our results to the current scanners. Fifth, our tiebreaking post hoc analysis for observer 3 was performed.
In conclusion, our results show that ventricular septal bowing was predictive of death for one observer but not another owing to the fact that observer 2 diagnosed ventricular septal bowing in more patients who did not die than did observer 1. Ventricular septal bowing in patients with PE may reflect an underlying hemodynamic derangement that places patients at increased risk of death, but considering the low mortality related to PE and the interobserver variability of this observation, it may be of limited clinical value. We found no association between the RV/LV diameter and death due to PE, which is in contrast to existing echocardiographic findings and CT literature. The ability of RV/LV measurements to predict death may depend on strict and consistent measurements, as well as the use of reconstructed four-chamber views. We found no association between the embolic burden and an increased risk of death for either observer, which is consistent with the bulk of the literature about the outcome related to PE.
|
ADVANCES IN KNOWLEDGE
|
|---|
- Ventricular septal bowing as detected at CT is predictive of death due to pulmonary embolism (PE) but with a low sensitivity and high interobserver variability.
- The ratio between the diameter of right ventricle and left ventricle was not associated with short-term death due to acute PE.
- Embolic burden was not associated with short-term death due to acute PE.
|
FOOTNOTES
|
|---|
Abbreviations: LV = left ventricle • OR = odds ratio • PE = pulmonary embolism • RV = right ventricle
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, P.A.A.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, P.A.A., M.B.G.; clinical studies, P.A.A., M.B.G.; statistical analysis, J.R.H., W.S.H., J.N.M.; and manuscript editing, all authors
|
References
|
|---|
- Goldhaber SZ, Visani L, De Rosa M, et al. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999;353:1386–1389.[CrossRef][Medline]
- Kasper W, Konstantinides S, Geibel A, et al. Management strategies and determinants of outcome in acute major pulmonary embolism: results of a multicenter registry. J Am Coll Cardiol 1997;30:1165–1171.[Abstract]
- Jerjes-Sanchez C, Ramirez-Rivera A, de Lourdes Garcia M, et al. Streptokinase and heparin versus heparin alone in massive pulmonary embolism: a randomized controlled trial. J Thromb Thrombolysis 1995;2:227–229.[Medline]
- ten Wolde M, Sohne M, Quak E, Mac Gillavry MR, Buller HR. Prognostic value of echocardiographically assessed right ventricular dysfunction in patients with pulmonary embolism. Arch Intern Med 2004;164:1685–1689.[Abstract/Free Full Text]
- van der Meer RW, Pattynama PM, van Strijen MJ, et al. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3-month follow-up in patients with acute pulmonary embolism. Radiology 2005;235:798–803.[Abstract/Free Full Text]
- Schoepf UJ, Kucher N, Kipfmueller F, et al. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004;110:3276–3280.
- Han D, Lee KS, Franquet T, et al. Thrombotic and nonthrombotic pulmonary arterial embolism: spectrum of imaging findings. RadioGraphics 2003;23:1521–1539.[Abstract/Free Full Text]
- Qanadli SD, El Hajjam M, Vieillard-Baron A, et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR Am J Roentgenol 2001;176:1415–1420.[Abstract/Free Full Text]
- Bankier AA, Janata K, Fleischmann D, et al. Severity assessment of acute pulmonary embolism with spiral CT: evaluation of two modified angiographic scores and comparison with clinical data. J Thorac Imaging 1997;12:150–158.[Medline]
- Ribeiro A, Lindmarker P, Juhlin-Dannfelt A, Johnsson H, Jorfeldt L. Echocardiography Doppler in pulmonary embolism: right ventricular dysfunction as a predictor of mortality rate. Am Heart J 1997;134:479–487.[CrossRef][Medline]
- Kasper W, Konstantinides S, Geibel A, et al. Prognostic significance of right ventricular afterload stress detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 1997;77:346–349.[Abstract/Free Full Text]
- Grifoni S, Olivotto I, Cecchini P, et al. Short-term clinical outcome of patients with acute pulmonary embolism, normal blood pressure, and echocardiographic right ventricular dysfunction. Circulation 2000;101:2817–2822.
- Araoz PA, Gotway MB, Trowbridge RL, et al. Helical CT pulmonary angiography predictors of in-hospital morbidity and mortality in patients with acute pulmonary embolism. J Thorac Imaging 2003;18:207–216.[CrossRef][Medline]
- Turnier E, Hill JD, Kerth WJ, Gerbode F. Massive pulmonary embolism. Am J Surg 1973;125:611–622.[CrossRef][Medline]
- Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961;102:19–28.[CrossRef][Medline]
- Morrison MC. Is pulmonary embolectomy obsolete? Br J Dis Chest 1963;57:187–191.[CrossRef][Medline]
- Gifford RW Jr, Groves LK. Limitations in the feasibility of pulmonary embolectomy: a clinicopathologic study of 101 cases of massive pulmonary embolism. Circulation 1969;39:523–530.
- Donaldson GA, Williams C, Schnnel JG, Shaw RS. A reappraisal of the application of the trendelenburg operation to massive fatal pulmonary embolism. N Engl J Med 1963;268:171–174.[Medline]
- Coon WW, Coller FA. Clinicopathologic correlation in thromboembolism. Surg Gynecol Obstet 1959;109:259–269.[Medline]
- Urokinase pulmonary embolism trial: phase 1 results—a cooperative study. JAMA 1970;214:2163–2172.[CrossRef][Medline]
- Wu AS, Pezzullo JA, Cronan JJ, Hou DD, Mayo-Smith WW. CT pulmonary angiography: quantification of pulmonary embolus as a predictor of patient outcome—initial experience. Radiology 2004;230:831–835.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
I. Ocak and C. Fuhrman
CT Angiography Findings of the Left Atrium and Right Ventricle in Patients with Massive Pulmonary Embolism
Am. J. Roentgenol.,
October 1, 2008;
191(4):
1072 - 1076.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
