(Stroke. 1996;27:1516-1520.)
© 1996 American Heart Association, Inc.
Articles |
Northwestern University Medical School (R.L.H., E.J.R., P.R.Y., J.R.D., D.G.), the Rehabilitation Institute of Chicago (R.L.H., E.J.R., D.G.), and Northwestern Memorial Hospital (J.R.D.), Chicago, Ill.
| Abstract |
|---|
|
|
|---|
Method Plasma samples were drawn from 105 nonambulatory rehabilitation patients with recent ischemic or hemorrhagic stroke and assayed for D-dimer with an enzyme-linked immunosorbent method. Samples were drawn within 24 hours of venous duplex ultrasound (VDU) screening for DVT. Optimal discriminant analysis was used to determine whether plasma D-dimer level, age, sex, days after stroke onset, stroke etiology, National Institutes of Health Stroke Scale score, and ambulatory status could correctly classify patients' DVT status.
Results Fourteen of 105 patients had DVT identified by VDU scan. Of all attributes, only D-dimer level had significant ability to discriminate between patients with or without DVT (P<.0001). The optimal cut point for predicting DVT was D-dimer=1591 ng/mL, resulting in 79% sensitivity, 78% specificity, 35% positive predictive value, and 96% negative predictive value. Reducing the D-dimer cut point to 1092 ng/mL improved both sensitivity and negative predictive value to 100% but reduced specificity to 66% and positive predictive value to 31%.
Conclusions A D-dimer level
1092 ng/mL can exclude the presence of DVT in stroke rehabilitation patients. When a D-dimer level >1092 ng/mL occurs, further diagnostic testing is necessary to confirm DVT. Plasma D-dimer level is a simple and inexpensive screening test for DVT during stroke rehabilitation.
Key Words: diagnosis dimers rehabilitation venous thrombosis
| Introduction |
|---|
|
|
|---|
Clinical signs and symptoms are unreliable for detection of DVT.12 Although contrast venography is the gold-standard diagnostic test, it is invasive and expensive. VDU is a safe, noninvasive alternative that has proven reliable for detection of DVT in symptomatic pa-tients,13 14 15 16 17 but it is costly to perform and requires expertise for interpretation.
A novel diagnostic approach to DVT uses plasma levels of coagulation markers such as D-dimer, a cross-linked fibrin degradation product that is generated during thrombus formation. Several studies indicate that the D-dimer assay can be applied as a simple, inexpensive screening test for venous thromboembolism. Using meta-analysis, Bounameaux et al18 showed that plasma D-dimer level has an overall diagnostic sensitivity of 97% and a specificity of 47% in various patient populations suspected of having DVT who were referred for venography. Such studies have not been performed in the acute stroke population or in the rehabilitation setting.
The purpose of this study was to evaluate whether plasma D-dimer level could identify patients with DVT referred for VDU screening during acute inpatient stroke rehabilitation. We assumed a priori that the standard normal D-dimer value for healthy adults (<500 ng/mL) may not be the best cut point for discriminating DVT in this population. Therefore, we also sought to define the optimal plasma D-dimer level cut point for discriminating DVT status in patients with stroke.
| Subjects and Methods |
|---|
|
|
|---|
Procedures
Age, sex, date of stroke, and date of admission to rehabilitation were documented for each patient. Using clinical examination and review of available laboratory tests, we classified stroke etiology as ischemic or hemorrhagic for each subject. The method of DVT prophylaxis used before and during rehabilitation hospitalization and use of antiplatelet medications were noted.
A physical examination to assess neurological status on admission was performed by a physician investigator (R.L.H.) and scored according to the NIHSS.19 FIM score on admission,20 as assessed by a certified nurse evaluator using Uniform Data System criteria, was recorded retrospectively from each patient's chart.
All study patients underwent VDU scanning and blood sampling for D-dimer during inpatient rehabilitation. VDU scans were performed by an experienced vascular technologist using an Ultra Mark 4 Duplex instrument (Advanced Technology Laboratories) and interpreted by a vascular surgeon (J.R.D.). Studies were positive for DVT if thrombus was visualized on B-mode ultrasound or if noncompressibility was noted in proximal deep veins including the femoral, superficial femoral, and popliteal veins. Altered respiratory variation on venous Doppler distal to the occlusion was confirmatory.
Within 24 hours of the VDU scanning, a single (5-mL) plasma sample was drawn in a 3.2% citrated evacuated tube from each subject. Each sample was centrifuged immediately at 3000 rpm for 10 minutes and stored at -70°C for later assay. Plasma D-dimer assays were performed in a blinded manner with the Asserachrom D-Di enzyme immunoassay kit (Diagnostica Stago). All samples were run in duplicate at 1:200 and 1:400 dilutions because of the elevated D-dimer concentrations found in study patients. D-Dimer levels were determined by spectrophotometry at a wavelength of 492 nm (Reader 400, SLT Lab Instruments).
Analysis
Because the purpose of this study was to assess the diagnostic utility of plasma D-dimer level as a screening test for DVT, the data were analyzed from the perspective of the ODA para-digm.21 22 23 24 ODA is ideal for this purpose because it explicitly maximizes weighted or unweighted classification accuracy rather than variance or likelihood functions. Univariate analysis treated DVT status as the class variable. Attributes included plasma D-dimer level (nanograms per milliliter), subject age (years), sex, number of days since stroke, stroke etiology, admission NIHSS score, and FIM score for ambulatory status on admission. Weighting by prior odds was used because of differing numbers of patients in the two class categories, and permutation probabilities were estimated with data from 10 000 Monte Carlo experiments.21 22 Hypothesizing that plasma D-dimer levels would be higher in patients with DVT, we used a one-tailed analysis; other univariate analyses were two-tailed. We estimated cross-generalizability of each univariate model in discriminating DVT status for independent random samples using a "leave-one-out" (jackknife) validity analysis, in which each patient is classified using a model created on the basis of data from all patients except the one being classified. Multivariate analysis was unable to identify a statistically reliable multiattribute model for predicting DVT status.23 24 25 26 27
The preceding analysis produced an optimal D-dimer cut point to maximize model accuracy while minimizing both false-positives and false-negatives. Since classification accuracy was <100% and because misidentifying the presence of DVT would be contradictory to the goals of a screening test, a second optimal discriminant analysis was conducted that evaluated a model in which false-negative errors were minimized.
| Results |
|---|
|
|
|---|
|
The optimal D-dimer level cut point for discriminating the presence or absence of DVT was determined to be 1591 ng/mL by ODA. The model predicts that patients with D-dimer level
1591 ng/mL are free of DVT and patients with D-dimer level >1591 ng/mL have DVT. When applied to the total sample, this model correctly classified 81% of the patients (Table 2
). However, jackknife analysis indicated marginal instability, with cross-generalizability of overall accuracy estimated to be 78% (Table 3
). The jackknife values for sensitivity (79%), specificity (78%), positive predictive value (35%), and negative predictive value (96%) indicate that this model shows good utility for excluding the presence of DVT in stroke survivors within the rehabilitation setting. The overall effect strength is expressed on a scale of 0 to 1, in which 0 represents chance and 1 represents perfect classification performance. The plasma D-dimer cut point value of 1591 ng/mL for patients with stroke has a jackknife effect strength of 0.44, indicating that the model provides moderate ability to identify patients with or without DVT.
|
|
If plasma D-dimer is to be used as a screening test to primarily exclude the presence of DVT, thereby preventing the need for further diagnostic evaluation, the plasma D-dimer cut point should provide perfect sensitivity for DVT and eliminate risk for false-negatives. Accordingly, the data were reanalyzed to determine a new cut point for plasma D-dimer that maximizes test sensitivity and negative predictive value. The resulting ODA model predicts that patients with plasma D-dimer level
1092 ng/mL are free of DVT, yielding 100% sensitivity and negative predictive value in both total sample and jackknife analyses (P<.0001). However, this model shows reduced test specificity (66%) but a slightly improved overall effect strength (0.48) relative to the former model (Tables 4 and 5).![]()
|
|
| Discussion |
|---|
|
|
|---|
In the present study we tested whether several clinical factors could correctly classify patients with or without DVT as detected by VDU during acute inpatient stroke rehabilitation. Age and time since stroke were not predictive of DVT. Level of neurological impairment and ambulatory status also had no reliable predictive value, possibly as a result of sample bias since only patients with limited ambulatory skills were selected.
As hypothesized, we found that D-dimer level was significantly elevated in patients with either ischemic or hemorrhagic stroke who had DVT. A diagnostic cut point of 1591 ng/mL was determined with the aid of an ODA classification model that provided maximum test accuracy for discriminating the DVT status of patients. Using a second model, we showed that a cut point value of 1092 ng/mL reduced overall classification accuracy but correctly identified all patients with DVT. The clinical implications of this finding are extremely important. A D-dimer value
1092 ng/mL can exclude the presence of DVT with the same level of confidence as a normal VDU. However, further diagnostic testing with VDU or venography is necessary to confirm the presence of DVT in patients with D-dimer values >1092 ng/mL (Figure
). Future research with larger samples should test the reliability of these D-dimer cut points over multiple replications. Multiattribute models yielding greater overall accuracy may also be identified by the use of additional clinical and hematologic factors that are predictive of DVT.
|
Clinical use of the D-dimer immunoassay could result in considerable cost reduction in the care of patients with stroke. At our institution the charge for D-dimer immunoassay is $75, whereas bilateral lower limb VDU is $550, including physician interpretation. Fifty-seven percent of the present study sample could have had DVT ruled out with D-dimer assay alone. For our study sample, the screening model presented in the Figure
would reduce total charges by 44%. Avoiding transportation costs to outside vascular laboratories could also provide additional savings for freestanding rehabilitation facilities.
Based on this model, we suggest that plasma D-dimer level, measured by ELISA technology, is a simple and inexpensive initial screening test for DVT in the stroke rehabilitation setting. Our data indicate that a majority of physically impaired patients with stroke who are typically considered at risk can have DVT safely excluded by plasma D-dimer level and avoid the need for further diagnostic testing.
These findings are similar to results from other studies that have evaluated the use of plasma D-dimer level for detection of thromboembolism. For DVT, the range of diagnostic sensitivity with D-dimer assay is 92% to 100% with a specificity of 21% to 51% compared with VDU in various hospitalized and outpatient populations.31 32 33 34 35 When compared with venography, D-dimer level shows similar results: 95% to 100% sensitivity and 6% to 65% specificity.36 37 38 39 40 Sensitivity for pulmonary embolism confirmed by nuclear scanning or pulmonary angiography is 93% to 100%, with specificity of 19% to 100%.41 42 43 44 45 46 47 In general, plasma D-dimer level has proven to be a sensitive but nonspecific indicator of thromboembolic disease.
The poor test specificity found in this and other studies is an indication that plasma D-dimer level might be elevated in any condition in which intravascular thrombosis may occur, such as trauma, surgery, or severe illness. The present study involved patients with recent stroke, including some with intracerebral hemorrhage, craniotomy, recent gastric tube placement, and other medical comorbidities. A further explanation for the low specificity is that many patients examined in this study were asymptomatic for DVT. The sensitivity of VDU in asymptomatic postoperative neurosurgical patients was <70% in one study.48 This raises the possibility that some patients harbored venous thrombi that were undetected by VDU and yet were reflected by high levels of plasma D-dimer.
The use of the ELISA technique in the present study is important when the application of plasma D-dimer assays is considered. Most clinical laboratories use a latex agglutination technique that yields results in a matter of minutes, whereas the ELISA method requires approximately 4 hours. Unfortunately, the diagnostic sensitivity of the latex agglutination method remains controversial and has not been recommended over the ELISA method.* Therefore, the model presented in this study can only be applied when the ELISA technique is used. A newer whole blood D-dimer test that can be rapidly performed at bedside has recently been developed and shows some promise as a clinical tool, but further testing is needed.52
We conclude that D-dimer immunoassay is a simple method to exclude the presence of DVT in patients with stroke during inpatient rehabilitation and is a cost-effective initial screening tool for diagnostic use in this population. It is particularly valuable for freestanding rehabilitation facilities that have laboratory services but lack more elaborate diagnostic testing, such as duplex ultrasound or venography. Further clinical studies with diagnostic decision trees that incorporate the D-dimer assay as an initial screen are warranted.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
| Footnotes |
|---|
*References 31, 32, 36, 38, 39, 44, 46, 49-51.
Received April 9, 1996; revision received May 23, 1996; accepted May 23, 1996.
| References |
|---|
|
|
|---|
2. McCarthy ST, Turner JJ, Robertson D, Hawkey CJ. Low-dose heparin as a prophylaxis against deep-vein thrombosis after acute stroke. Lancet. 1977;11:800-801.
3.
McCarthy ST, Turner J. Low-dose subcutaneous heparin in the prevention of deep-vein thrombosis and pulmonary emboli following acute stroke. Age Ageing.. 1986;15:84-88.
4. Prins MH, Gelsema R, Sing AK, van Heerde LR, den Ottolander GJH. Prophylaxis of deep venous thrombosis with a low-molecular-weight heparin (Kabi 2165/Fragmin) in stroke patients. Haemostasis. 1989;19:245-250.[Medline] [Order article via Infotrieve]
5. Warlow C, Ogston D, Douglas AS. Deep venous thrombosis of the legs after strokes. BMJ.. 1976;1:1178-1183.
6. Bornstein NM, Norris JW. Deep vein thrombosis after ischemic stroke: rationale for a therapeutic trial. Arch Phys Med Rehabil. 1988;69:955-958.[Medline] [Order article via Infotrieve]
7. Gelmers HJ. Effects of low-dose subcutaneous heparin on the occurrence of deep vein thrombosis in patients with ischemic stroke. Acta Neurol Scand.. 1980;61:313-318.[Medline] [Order article via Infotrieve]
8. Izzo KL, Aquino E. Deep venous thrombosis in high-risk hemiplegic patients: detection by impedance plethysmography. Arch Phys Med Rehabil. 1986;67:799-802.[Medline] [Order article via Infotrieve]
9. Turpie AG, Levine MN, Hirsh J, Carter CJ, Jay RM, Powers PJ, Andrew M, Magnani HN, Hull RD, Gent M. Double-blind randomised trial of ORG 10172 low-molecular weight heparinoid in prevention of deep-vein thrombosis in thrombotic stroke. Lancet. 1987,1:523-526.
10. Sioson ER, Crowe WE, Dawson NV. Occult proximal deep vein thrombosis: its prevalence among patients admitted to a rehabilitation hospital. Arch Phys Med Rehabil. 1988;69:183-185.[Medline] [Order article via Infotrieve]
11. Oczkowski WJ, Ginsberg JS, Shin A, Panju A. Venous thromboembolism in patients undergoing rehabilitation for stroke. Arch Phys Med Rehabil. 1992;73:712-716.[Medline] [Order article via Infotrieve]
12. Haeger K. Problems of acute deep venous thrombosis, I: the interpretation of signs and symptoms. Angiology. 1969;20:219-223.
13.
Cronan JJ, Dorfman GS, Scola FH, Schepps B, Alexander J. Deep venous thrombosis: ultrasound assessment using vein compression. Radiology. 1987;162:191-194.
14.
Appelman PT, De Jong TE, Lampmann LE. Deep venous thrombosis of the leg: ultrasound findings. Radiology. 1987;163:743-746.
15.
Vogel P, Laing FC, Jeffrey RB, Wing VW. Deep venous thrombosis of the lower extremity: ultrasound evaluation. Radiology. 1987;163:747-751.
16. O'Leary DH, Kane RA, Chase BM. A prospective study of the efficacy of B-scan sonography in the detection of deep venous thrombosis in the lower extremities. J Clin Ultrasound.. 1988;16:1-8.[Medline] [Order article via Infotrieve]
17. Lensing AWA, Prandoni P, Brandjes D, Huisman PM, Vigo M, Tomasella G, Krekt J, Wouter ten Cate J, Huisman MV, Buller HR. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med. 1989;320:342-345.[Abstract]
18. Bounameaux H, de Moerloose P, Perrier A, Reber G. Plasma measurement of D-dimer as diagnostic aid in suspected thromboembolism: an overview. Thromb Haemost.. 1994;71:1-6.[Medline] [Order article via Infotrieve]
19. Lyden P, Brott T, Tilley B, Welch KMA, Mascha EJ, Levine S, Haley EC, Grotta J, Marler J, and the NINDS TPA Stroke Study Group. Improved reliability of the NIH Stroke Scale using video training. Stroke. 1994;25:2220-2226.[Abstract]
20. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil. 1993;74:531-536.[Medline] [Order article via Infotrieve]
21. Yarnold PR, Soltysik RC. Theoretical distributions of optima for univariate discrimination of random data. Decis Sci.. 1991;22:739-752.
22. Soltysik RC, Yarnold PR. ODA 1.0 Optimal Data Analysis for DOS. Chicago, Ill: Optimal Data Analysis, Inc; 1993.
23. Yarnold PR, Soltysik RC, Martin GJ. Heart rate variability and susceptibility for sudden cardiac death. Stat Med.. 1994;13:1015-1021.[Medline] [Order article via Infotrieve]
24. Yarnold PR. Discriminating geriatric and non-geriatric patients using functional status information: an example of classification tree analysis via UniODA. Educ Psych Measurement. 1996;56:656-667.[Abstract]
25. Silva APD, Stam A. Discriminant analysis. In: Grimm LG, Yarnold PR, eds. Reading and Understanding Multivariate Statistics. Washington, DC: APA Books; 1995:277-318.
26. Wright RE. Logistic regression. In: Grimm LG, Yarnold PR, eds. Reading and Understanding Multivariate Statistics. Washington, DC: APA Books; 1995:217-244.
27. Yarnold PR, Hart LA, Soltysik RC. Optimizing the classification performance of logistic regression and Fisher's discriminant analysis. Educ Psych Measurement. 1994;54:73-85.[Abstract]
28.
Bounds JV, Wiebers DO, Whisnant JP, Okazaki H. Mechanisms and timing of deaths from cerebral infarction. Stroke. 1981;12:474-477.
29. Viitanen M, Winblad B, Asplund K. Autopsy-verified causes of death after stroke. Acta Med Scand. 1987;222:401-408.[Medline] [Order article via Infotrieve]
30. White RH, McGahan JP, Daschbach MM, Hartling RP. Diagnosis of deep-vein thrombosis using duplex ultrasound. Ann Intern Med.. 1989;111:297-304.
31. Boneu B, Bes G, Pelzer H, Sie P, Boccalon H. D-dimers, thrombin antithrombin III complexes and prothrombin fragments 1+2: diagnostic value in clinically suspected deep vein thrombosis. Thromb Haemost.. 1991;65:28-32.[Medline] [Order article via Infotrieve]
32. Chang-Liem GS, Lustermans FATh, van Wersch JWJ. Comparison of the appropriateness of the latex and ELISA plasma D-dimer determination for the diagnosis of deep venous thrombosis. Haemostasis. 1991;21:106-110.[Medline] [Order article via Infotrieve]
33. Kroneman H, Van Bergan P, Knot E, Jonker J, de Maat M. Diagnostic value of D-dimer for deep venous thrombosis in outpatients. Haemostasis. 1991;21:286-292.[Medline] [Order article via Infotrieve]
34. Heijboer H, Ginsberg JS, Buller HR, Lensing AWA, Colly LP, Wouter ten Cate J. The use of the D-dimer test in combination with non-invasive testing versus serial non-invasive testing alone for the diagnosis of deep-vein thrombosis. Thromb Haemost.. 1992;67:510-513.[Medline] [Order article via Infotrieve]
35. Carter CJ, Doyle DL, Dawson N, Fowler S, Devine DV. Investigations into the clinical utility of latex D-dimer in the diagnosis of deep vein thrombosis. Thromb Haemost.. 1993;69:8-11.[Medline] [Order article via Infotrieve]
36. Heaton DC, Billings JD, Hickton CM. Assessment of D dimer assays for the diagnosis of deep vein thrombosis. J Lab Clin Med.. 1987;110:588-591.[Medline] [Order article via Infotrieve]
37. Rowbotham BJ, Carroll P, Whitaker AN, Bunce IH, Cobcroft RG, Elms MJ, Masci PP, Bundesen PG, Rylatt DB, Webber AJ. Measurement of crosslinked fibrin derivatives: use in the diagnosis of venous thrombosis. Thromb Haemost.. 1987;57:59-61.[Medline] [Order article via Infotrieve]
38. Ott P, Astrup L, Jensen RH, Nyeland B, Pedersen B. Assessment of D-dimer in plasma: diagnostic value in suspected venous thrombosis of the leg. Acta Med Scand. 1988;224:263-267.[Medline] [Order article via Infotrieve]
39. Bounameaux H, Schneider PA, Reber G, de Moerloose P, Krahenbuhl B. Measurement of plasma D-dimer for diagnosis of deep venous thrombosis. Am J Clin Pathol.. 1989;91:82-85.[Medline] [Order article via Infotrieve]
40. Mossaz A, Grille S, Vitoux JF, Abdoucheli-Baudot N, Aiach M, Fiessinger JN. Valeur des D-dimeres dans le diagnostic en urgence des thromboses veineuses. La Presse Medicale. 1990;19:1055.
41.
Rowbotham BJ, Egerton-Vernon J, Whitaker AN, Elms MJ, Bunce IH. Plasma cross linked fibrin degradation products in pulmonary embolism. Thorax. 1990;45:684-687.
42. Bounameaux H, Cirafici P, de Moerloose P, Schneider PA, Slosman D, Reber G, Unger PF. Measurement of D-dimer in plasma as diagnostic aid in suspected pulmonary embolism. Lancet. 1991;337:196-200.[Medline] [Order article via Infotrieve]
43. Demers C, Ginsberg JS, Johnston M, Brill-Edwards P, Panju A. D-dimer and thrombin-antithrombin III complexes in patients with clinically suspected pulmonary embolism. Thromb Haemost. 1992;67:408-412.
44.
Ginsberg JS, Brill-Edwards PA, Demers C, Donovan D, Panju A. D-Dimer in patients with clinically suspected pulmonary embolism. Chest. 1993;104:1679-1684.
45.
Goldhaber SZ, Simons GR, Elliott CG, Haire WD, Toltzis R, Blacklow SC, Doolittle MH, Weinberg DS. Quantitative plasma D-dimer levels among patients undergoing pulmonary angiography for suspected pulmonary embolism. JAMA. 1993;270:2819-2822.
46. van Beek EJR, van den Ende B, Berckmans RJ, van der Heide YT, Brandjes DPM, Sturk A, ten Cate JW. A comparative analysis of D-dimer assays in patients with clinically suspected pulmonary embolism. Thromb Haemost.. 1993;70:408-413.[Medline] [Order article via Infotrieve]
47.
Quinn RJ, Nour R, Butler SP, Glenn DW, Travers PL, Wellings G, Kwan YL. Pulmonary embolism in patients with intermediate probability lung scans: diagnosis with venous US and D-dimer measurement. Radiology. 1994;190:509-511.
48. Jongbloets LM, Lensing AW, Kooperman MM, Buller HR, ten Cate JW. Limitations of compression ultrasound for the detection of symptomless postoperative deep vein thrombosis. Lancet. 1994;343:1142-1144.[Medline] [Order article via Infotrieve]
49. de Moerloose P, Minazio Ph, Reber G, Perrier A, Bounameaux H. D-Dimer determination to exclude pulmonary embolism: a two-step approach using latex assay as a screening tool. Thromb Haemost.. 1994;72:89-91.[Medline] [Order article via Infotrieve]
50.
Heit JA, Nichols WL. Plasma D-dimer levels and diagnosis of pulmonary embolism. JAMA. 1994;271:1404. Letter.
51.
van Beek EJR, Buller HR, Bounameaux H. Latex D-dimer for diagnosing pulmonary embolism. J Nucl Med.. 1994;35:189.
52. Ginsberg JS, Wells PS, Brill-Edwards P, Donovan D, Panju A, van Beek EJR, Patel A. 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.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
![]() |
S. Goodacre, F.C. Sampson, A.J. Sutton, S. Mason, and F. Morris Variation in the diagnostic performance of D-dimer for suspected deep vein thrombosis QJM, July 1, 2005; 98(7): 513 - 527. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Yablon, W. A. Rock Jr., T. G. Nick, M. Sherer, C. M. McGrath, and K. H. Goodson Deep vein thrombosis: Prevalence and risk factors in rehabilitation admissions with brain injury Neurology, August 10, 2004; 63(3): 485 - 491. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Kelly, A. Rudd, R. R. Lewis, and B. J. Hunt Plasma D-Dimers in the Diagnosis of Venous Thromboembolism Arch Intern Med, April 8, 2002; 162(7): 747 - 756. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Kelly, A. Rudd, R.R. Lewis, and B.J. Hunt Screening for subclinical deep-vein thrombosis QJM, October 1, 2001; 94(10): 511 - 519. [Full Text] [PDF] |
||||
![]() |
J. Kelly, A. Rudd, R. Lewis, and B. J. Hunt Venous Thromboembolism After Acute Stroke Stroke, January 1, 2001; 32(1): 262 - 267. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1996 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |