Frequency of pulmonary embolism in ambulant patients with pelvic vein thrombosis: A prospective study H u g o Partsch, MAD, Karin Oburger, MD, Adolf Mostbeck, MD, Beatrix K6nig, MD, and H o r s t K6hn, MD, Vienna, Austria One hundred thirty-nine consecutive patients (average age 70.1 years) who were able to walk with a swollen leg were seen at the clinic where diagnosis of acute deep vein thrombosis (DVT) extending to the pelvis was confirmed by injecting microspheres labeled with technetium 99m into the dorsal foot vein (radionuclide venography). Thirty-nine (28%) of these patients had malignant disease. Perfusion lung scans performed immediately after radionuclide venography were supplemented by inhalation scans (99roTe-labeled diethylenetriamine pentaacetic acid aerosol) in case of perfusion defects. During scintigraphy patterns highly indicative of pulmonary embolism (PE) were found in 80 patients (58%), but only 11 (7.9%) had minor clinical symptoms. All patients were admitted to the ward, were given standard heparin subcutaneously (35,000 to 40,000 units/24 hr) and firm bandages, and were encouraged to walk. After 11 days pulmonary scintigraphy was repeated and revealed no change in 55 of 59 patients without PE and in 40 of 80 patients with PE. Thirty-three patients (23.7%) showed regression of perfusion defects. New PE developed in 11 patients (7.9%, four without and seven with previous PE). Autopsy revealed that one 80-year-old patient with prostatic carcinoma had died of massive PE. When comparing this frequency of newly developed PE during ambulation with the occurrence of PE after bed rest, according to the literature, it is no more dangerous for a mobile patient with proximal DVT to walk wearing a firm bandage than it is for the patient to be in bed. Therefore we recommend treating mobile patients with DVT by use of anticoagulation and firm compression bandages and without immobilization. (J VASC SURG 1992;16:715-22.)

In patients with acute, proximal deep vein thrombosis (DVT), fibrinolysis or thrombectomy may produce excellent results including complete restitution. However, there are several contraindications for fibrinolysis, and the side effects and possible complications of both regimens may be even more hazardous than DVT. These methods are restricted to yotmger patients who have fresh clots, with the rationale of keeping the valves intact to prevent postthrombotic syndrome. Older patients and patients with a history of DVT longer than I week can be better treated by conservative methods. The conventional therapeutic approach is to administer heparin for 5 to 10 days and then continue From the Department of Dermatology (Drs. Partsch and Oburger), Ludwig Boltzmann Institute for Nuclear Medicine (Dr. Mostbeck), and the Department of Nuclear Medicine, Wilhelminenspital (Drs. K6nig and K6hn), Vienna. Presented at the Fourth AnnualMeetingof the AmericanVenous Forum, Coronado, Calif., Feb. 26-28, 1992. Reprint requests: Hugo Partsch, MD, Department of Dermatology, Wilhelminenspital, Al171 Vienna, Austria. 24/1/41577

with oral anticoagulant therapy with a crossover period of about 5 days. Most centers administer heparin intravenously and keep the patient in bed, thereby hoping to avoid pulmonary embolism (PE), although this behavior is based more on tradition than on firm data. It has been our policy for years to encourage ambulant patients with DVT to walk while wearing a firm compression bandage and to administer heparin subcutaneously, not to immobilize the patient. The purpose of this prospective study was to detect the frequency of PE in mobile patients who have acute proximal DVT extending into the pelvis when they continue walking with a firm bandage and receive heparin therapy. PATIENTS A N D M E T H O D S Patients Patients enrolled in this study were referred to our department from April 1989 to May 1991 after DVT extending into the pelvis was confirmed by radionu715

716

~fournalof VASCULAR SURGERY

Partsch et al.

Table I. Clinical characteristics of 139 patients with DVT extending into the pelvis (percentages in parentheses) Age (yr) Range Mean < 60 >60 Sex Male Female Previous history of thrombosis Present history < 1 wk 1-4 wk Not clear Right Left Bilateral

18 to 94 70.1 26 (20) 113 (80) 53 (38) 86 (62) 61 (44) 78 (56) 50 (36) 11 (8) 47 (35) 90 (65) 2 (0.7)

clide venography. Patients in whom fibrinolytic treatment or thrombectomy was performed (seven patients) and patients for whom we had been consulted but who had been immobilized in other departments because of surgery, trauma, or internal diseases were not included in the study. The remaining 139 patients were able to walk and did not have complete bed rest for at least 36 hours before admission to the hospital (Table I). There were 53 men and 86 women with ages ranging between 18 and 94 years (70.1 on average). Only 26 patients (20%) were younger than 60 years, and six (4.2%) were younger than 30 years. Sixty-one patients (44%) reported having had previous DVT or superficial phlebitis. The history of leg symptoms was shorter than 1 week in 78 patients (56%), between 1 and 4 weeks in 50 (36%), and uncertain in 11 (8%). One 76-year-old woman came to the hospital because of symptoms of phlegmasia cerulea dolens. In six patients pulmonary symptoms such as dyspnea and pleuritic pain preceded leg swelling. The study covered the time of hospitalization (range 8 days to 2 months, 21 days average). Methods Radionuclide venography. After each patient was tested by means of venous occlusion plethysmography, Doppler scanning, and duplex scanning, radionuclide venography was done. Microspheres with a diameter of 10 to 20 t~m labeled with technetium 99m (37 MBq in about 0.3 ml) were injected into dorsal foot veins of both legs. A pressure cuff of 40 m m Hg over the distal lower leg directed the tracer into the deep veins. A conventional gamma camera placed over the proximal thighs and the lower abdomen registered the tracer flow through the

pelvic veins into the inferior vena cava? A typical example is shown in Fig. 1, A. Only patients in whom radionuclide venography revealed thrombotic involvement of the external and common pelvic vein on one or both side were included in this study. Lung imaging. The microspheres are trapped in the terminal pulmonary arterioles depending on regional blood flow, which results in a perfusion scintigram of the lung; 300 to 400 K images in at least four standard views (anterior, posterior, right posterior oblique, and left posterior oblique) were acquired. If the perfusion was normal, the investigation was stopped and no additional ventilation study was performed. In this case clinically important PE was excluded? 4 If perfusion defects were present, an additional inhalation scan with use of 99mTc-labeled diethylenetriamine pentaacetic acid (DPTA) aerosol was performed in the same projections. An especially developed delivery system, the APE-nebulizer (Aerosol Production Equipment; Mallinckrodt, St. Louis, Mo.), generated a dry submicronic DTPA aerosol with high alveolar targeting) Probability of PE was assessed by regional comparison of perfusion with ventilation images and a recent chest radiography. Perfusion defects mismatched to a normal inhalation scan and a normal radiograph were judged as diagnostic for PE because all patients had acute iliofemoral D V T . 2'6'7 Fig. 1, B, shows an example. In every patient a conventional perfusion lung scan was repeated after 9 to 14 days (11 days average) by injecting 70 MBq 99mTc-labeled microspheres into an arm vein. The results of this investigation were compared with the results of the baseline scan of the first examination. Supplementary examinations. Additional examinations included complete blood and urine analysis, chest radiography, electrocardiography, blood gas analysis in case of clinically symptomatic PE, and screening for malignancies (gynecologic and urologic investigation, abdominal ultrasonography, and enzyme tests). Therapeutic regimen. Standard heparin was administered intravenously as a bolus dose of 5000 units, followed by subcutaneous injections of approximately 30,000 to 40,000 units per 24 hours. The heparin dose was adjusted to maintain the activated partial thromboplastin time at 1.5 to 2 times the control value (normal value 27 to 35 seconds). Heparin injected subcutaneously every 8 to 12 hours was continued for 10 to 14 days. Oral anticoagulant therapy was introduced after the second lung scan, overlapping the last 4 days of heparin administration if there was no contraindication.

Volume 16 Number 5 November I992

Pulmonary embolism in ambulant patients with DVT

717

Fig. 1. A, Radionuclide phlebography reveals unimpeded flow of 99mTc-labeledmicrospheres in right iliac vein up to inferior caval vein bnt network of collateral vessels on left side. B, Reaching lungs, microspheres are trapped in pulmonary arterioles ofperfused regions. Perfusion defects can be seen in both lungs. C, Inhalation scintigraphy with 99mTc-labeleddiethylenetriamine pentaacetic acid reveals a homogenous distribution of activity. In presence of iliofemoral DVT, mismatch between normal ventilation and pathologic perfusion indicates PE.

Very firm compression bandages were applied to the affected leg with use of nonelastic zinc plaster and short-stretch material; these were changed every 2 to 3 days in the first week and every 7 days later on (Varicex F and Rosidal [Lohmann GMBH & Co. KG, Neuwied, Germany] or Comprilan on the lower leg, Panelast or Tricoplast [Beiersdorf AG, Hamburg, Germany] over the knee up to the groin). The resting pressure on the distal lower leg was between 40 and 50 mm Hg. The patients were encouraged to walk and to move their ankles as much as possible. Fixed bandages were applied for 4 weeks. Then the patients were advised to wear strong medical compression stockings (Sigvaris, class II or III; Ganzoni & CIE AG, St. Gallen, Switzerland). RESULTS Localization of thrombosis and underlying causes. Pelvic vein thrombosis occurred on the right

side in 47 patients (35%) and on the left side in 90 patients (65%), and bilateral acute DVT was observed in two. Two patients, both with gastrointestinal neoplasm, had caval filters because of previous thromboembolic events. The underlying causative factors that could be detected are listed in Table II and Table HI. No causes were identified in 77 patients (56%). Pulmonary embolism before therapy. At the time pelvic vein obstruction was diagnosed by radionuclide venography, a baseline scan combined with ventilation imaging disclosed a PE pattern in 80 patients (58%). Only i1 patients (7.9%) complained of symptoms such as dyspnea or chest pain. In 69 cases (49.6%), including one of the two patients with a vena cava filter, PE was clinically silent. Fifty-nine patients (42.4%) had no signs of embolism that could be seen by scintigraphy (Table IV). Analysis of the various subgroups reveals no statistically significant difference between the observed frequency and the expected frequency of PE depending on the side

718

Journal of VASCULAR SURGERY

Partsch et al.

Table II. Causative factors for pelvic vein thrombosis (n = 139; percentages in parentheses) Malignant disease Surgery/trauma Estrogen intake Coagulopathy Antithrombin III deficiency After delivery Pregnancy Intrafemoral drug injections Total

39 (28) 9 (5.5) 5 (3.6) 3 (3.2) 2 (1.4) 2 (1.4) I (0.7) 1 (0.7) 62 (44)

of DVT, previous history, and duration of current symptoms. However, men, patients over age 60 years, and patients with neoplasms showed a significantly higher frequency of PE than expected (chisquared test with Yates correction: p < 0.05, Table V, left column). Pulmonary embolism 11 days after heparin therapy, compression, and walking. After an average of 11 days, lung scans were unchanged in 95 patients (68.3%) and improved in 33 (23.7%). New PE occurred in 10 patients (7.1%) (Table IV). No change in the pulmonary scan was observed in 55 patients without primary PE and in 40 patients with primary PE. New emboli developed in six patients with PE at the first examination. Two of these patients had continuous symptoms but did not require immobilization. Four patients without signs of PE at the first investigation were diagnosed with acute PE at the repeat scan; only one had marked dyspnea. Autopsy revealed that an 80-year-old patient who had prostatic carcinoma died of massive pulmonary embolism 2 days after being admitted to the hospital after walking on a swollen leg for 4 weeks. The frequency of new PE did not differ significantly between the different subgroups (Table V, right column). Other complications. Minor bleeding occurred in five patients (3.7%). Predisposing conditions were hemorrhoids (two patients), urethral catheter (one patient), colon carcinoma (one patient), and trauma (one patient). One 56-year-old man died 12 days after admission; autopsy revealed myocardial infarction. DISCUSSION

Methodologic problems with diagnosing thromboembolic disease have to be considered with every investigative procedure. An advantage of radionuclide venography is that very thin foot and leg veins can be used for injection

Table III. Thirty-nine neoplasms detected in 139 patients with pelvic vein thrombosis (28%) Prostate Uterus Colon Breast Urinary bladder Stomach Rectum Ovaries Liver Lymphoma Melanoma Unknown origin

8 8 7 4 2 2 1 I I i 1 3

because of the small injection volume. Therefore adequate results can be obtained in every consecutive case, even in a large series. A disadvantage is the rather poor image resolution that provides only coarse information about the venous pathways in the proximal thigh and pelvis. Floating parts and the proximal edge of thrombi cannot be detected by this method. It is not possible to differentiate between an ascending and a descending thrombosis with use of radionucfide venography alone. One unique advantage of the use of labeled microspheres is the additional perfusion lung scan obtained with every examination. The predictive value of a negative scan (the probability that a patient with a normal scan does not have PE) is high? On the other hand, the specificity of a perfusion scan is low because perfusion defects are caused not only by vascular obstruction but also by reflex vasoconstriction (Euler-Liljestrand mechanism) in patients with obstructive airway disease. To increase specificity, perfusion scintigraphy has to be combined with ventilation imaging? Because DVT in the lower extremity is by far the most common source of PE, tests for the detection of DVT are of particular importance. The inclusion of these tests in the diagnostic scheme increases the probability of PE in patients suspected of having PE. 4'7'8 According to Bayesian analysis, the posttest likelihood ofa perfusion-ventilation mismatch for PE is significantly influenced by the probability, or the prevalence, of PE in a given population. 9 According to pathoanatomic and nuclear medicine studies, testing for DVT increases the probability of PE to 55% to 60%, and consequently the posttest likefihood of a pulmonary ventilation mismatch to present PE increases to greater than 90%. 6'10 A further improvement of the diagnostic accuracy of perfusion-ventilation imaging can be

Volume 16 Number 5 November 1992

719

Pulmonary embolism in ambulant patients with D V T

Table IV. Frequency of symptomatic and silent PE before and 11 days after therapy (anticoagulation, walking with compression bandages) (percentages in parentheses) Perf~sion scan after ll days Patients before therapy PE with symptoms PE with no symptoms No PE Total

11 69 59 139

(7.9) (49.6) (42.4) (100)

Improved

Unchanged

Deteriorated

4 29 __33 (23.7)

5 35 55 95 (68.3)

2 4 (1 + ) __4" 11 (7.9)

I +, Fatal PE in one patient 2 days after admission (nonsymptomatic PE at entry into the study). *One patient with symptoms (moderate dyspnea) and three without symptoms.

achieved by repeating lung scans before discharge from the hospital, as was done in this study. Treatment trials and clinic follow-up studies have shown that although the ventilation-perfusion scan is not always predictive of angiography results, it is a reliable predictor of patient outcome, n The occurrence of PE in patients with leg and pelvic vein thromboses is very frequent. In a previous study, combined perfusion-ventilation imaging was performed in a total of 169 patients referred to our depamnent because of suspected leg vein thrombosis? ° Sixty of 105 patients (57%) with proven DVT had PE, but only three of 64 (4.7%) without DVT had PE. In this study most of the PE (59%) were clinically silent. These figures are in accordance with recent data from the literature 1z~6 and correspond well with autopsy findings? 7as If scintigraphy is performed only in patients with symptoms, then reports on the frequency of PE will be unreliable because most embolic events will be missed. This is especially important if different therapeutic regimens are to be compared. Fatal PE was observed in one of 139 patients, a figure comparable with the results of other reports dealing with conventional therapy of DVT. 19-21After treating 155 patients who had DVT extending to the pelvis with very high doses of streptokinase, Martin et al.22 reported four deaths caused by PE; Holmstroem et al.23 reported four deaths after treating 150 patients with conventional doses of streptoldnase. The data that exist concerning the frequency of nonfatal PE after thrombectomy, fibrinolysis, and conservative therapy are based on systematic scintigraphic investigations. Compared with a baseline scan, the frequency of new perfusion defects during initial heparin therapy of DVT ranges between zero and 20%. Table VIA summarizes data from the literature that include the

Table V. Frequencies of primary PE and of new PEs in the subgroups after 11 days Primary PE (n=8O)

New PE (n = 11)

Age 60 yr (n = 113)

10 70*

2 9

Men (n = 53) Women (n = 86)

37* 43

3 8

Left (n = 91) Right (n = 48)

51 29

4 7

Previous history (n = 61) No previous history (n = 78)

37 43

4 7

< l w k ( n = 78) -> 1 wk (n = 61)

42 38

5 6

Neoplasm (n = 39) No neoplasm (n = 100)

28* 52

2 9

Chi-squared test with Yates correction: p < 0.05. *Higher frequency than expected.

frequency of initial perfusion defects. 12,1s,e4-e6Table VIB contains studies in which only new perfusion defects, detected by comparison with a baseline perfusion scan, are cited.21,2729 In a comparison of these data from patients who have DVT at all levels of the leg including isolated distal forms, our frequency of newly detected PE in pelvic vein thrombosis (7.9%) is almost in the same range. This means that our regimen of anticoagulant therapy, compression bandaging, and exercise in mobile patients is obviously no more dangerous than conventional bed rest. From a practical point of view two groups of patients should be differentiated: one group in which DVT develops during immobilization (e.g., after surgery) and a second group of patients who come to the clinic without previous bed rest. Unfortunately these two groups, whose prognoses probably differ, have not been separated in all the above mentioned studies. It has been our policy

720

Journal of VASCULAR SURGERY

Partsch et al.

Table VIA. Literature survey: Frequency of PE before and after therapy* Incidence of PE (%) before~aftertherapy Authors

DVT

Therapy

n

Initial PE

New PE

Plate et al.24 Plate et al. 24 Huisman et al.12 Albada et al. 2s Albada et al. 2s von Liebe et al. 26 Cuppini et al. is This study

Iliofemoral Iliofemoral V~'nole leg Whole leg Whole leg Iliofemoral Proximal DVT Iliofemoral

Heparin Surgery Intravenous heparin Intravenous heparin Intravenous Fragmin Surgery Intravenous heparin Subcutaneous heparin

29 20 95 98 96 29 100 139

45 45 51 47 35 24.1 59 58

+ 11 + 20 +8 + 13 + 11 + 17 +5 + 8

*Bed rest in five reports12,1s,24-26; compression therapy and ambulation in our study.

Table VIB. Literature survey: Frequency of new perfusion defects compared with a baseline scan after initial treatment Authors

Diagnosis

Therapy

n

New pe~fusion defects (%)

Gallus 27 Gallus 27 Prandoni 21 Prandoni 21 Monrea128 Monrea128 Girard 29 Girard 29

Venous thromboembolism Venous thromboembolism Proximal D V T Proximal D V T DVT PE Caval DVT Proximal D V T

Intravenous heparin, warfarin from day 7 Intravenous heparin, warfarin from day S Intravenous heparin Subcutaneous L M W H Intravenous heparin Intravenous heparin Intravenous heparin Intravenous heparin

127 139 81 79 83 38 18 38

8.5 3.6 19 5 6 18.4 0 0

LMWH, Low molecular weight heparin.

for years to treat patients in the latter category with ambulation on an outpatient basis when possible. We believe that immobilization, manipulations with a bedpan, sudden mobilization after several days of bed rest, and other well-known hemodynamic and respiratory risk factors together with long-term recumbency may even be more hazardous, especially in older patients. However, proving this thesis would have required forming a comparable control group treated with bed rest only. Because of bad experiences with the negatively selected group of immobile patients, it did not seem acceptable for us to randomize our mobile patients into one group receiving exercise therapy and one group receiving bed rest therapy. The same kind of argument is used by Schulman 3° who showed in a summarized comparison of different studies that patients who are mobilized after 0 to 2 days had a markedly higher rate of phlebographic improvement and a lower frequency of PE than had patients who were immobilized for 5 days. Hull et al. 31 encourage their patients to move around once the activated partial thromboplastin time indicates a therapeutic response. It must be stressed that the usual period of bed rest ranging between 2 and 20 days is based more on speculation than on experimental data..Am important prerequisite

for the mobilization of patients with D V T seems to be an adequate compression technique with fixed bandages applied with high pressurefi 2 Because the majority of our consecutive patients (56%) were admitted with a history of symptoms of no more than 1 week, leg pain, especially during walking, was a relatively consistent finding. A typical, criterion of a well-applied compression bandage is prompt relief of pain in most patients so they are able to walk again. Unfortunately the underestimated bandaging technique cannot be standardized; therefore comparison of results will be difficult. On the basis of phlebographic experience, it can be assumed that free-floating thrombi in the leg may be fixed by firm bandages. 32This of course will not be the case with thrombi extending to the pelvis where no external pressure can be exerted. As we know from a study by Norris et a1.,33 free-floating thrombi in the ihac vein have a higher risk of causing PE than occluding thrombi. Based on duplex examination Baldridge et al. 34 found 10% of acute D V T to be connected with free-floating thrombi; of these DVT, only 13% developed clinically significant PE confirmed by ventilation-perfusion scanning. We are not able to delineate free-floating thrombi with use of radionuclide venography. The high frequency of

Volume 16 Number 5 November 1992

emboli found in case of pelvic vein thrombosis and the clinical observation of pulmonary symptoms, which may precede leg swelling, point to the fact that DVT is just one aspect of thromboembolic disease. Our data suggest that in ambulant patients seen at the clinic with the main symptom of an acute leg problem, the life-threatening danger of fatal PE has been already surpassed at the time DVT is diagnosed. If adequate amounts of anticoagulant are administered, the risk remains low even if patients continue walking with bandages. The surprisingly high incidence (28%) of malignant diseases (Table II) exceeding recent data from the literature 35is one argument for admitting patients with proximal DVT to the hospital even if they are kept ambulant. The diagnosis of a neoplasm had been established before admission in 25 patients, but in 14 of 39 patients it was detected only because of tumor screening. Another argument in favor of hospitalization is the need to monitor therapy with unfractioned standard heparin by frequent laboratory testing. If these two problems can also be solved for outpatients, ambulant care of mobile patients with DVT would become possible. At least the hospital stay would be shortened by early introduction of oral anticoagulant therapy. The potential savings of in-hospital expenditures are evidenc 31, 36

REFERENCES

1. Partsch H, Lofferer O, Mostbeck A, Tham B. Die Isotopenphlebographie mit anschlie[3ender Lungenszintigraphie (markierte Mikrosph~en) bei akuter Beckenvenenthrombose. In: Ehringer H, ed. Akute tiefe Becken- und Beinvenethrombose. Bern: Hans Huber, 1976:81-90. 2. Alderson PO, Biello DR, Gottschalk A, et al. Tc-99m-DTPA aerosol and radioactive gases compared as adjuncts to perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1984;153:515-21. 3. Hull RD, Raskob GE, Coates G, Panju AA. Clinical validity of a normal perfusion lung scan in patients with suspected pulmonary embolism. Chest 1990;97:23-6. 4. Kelley MA, Carson JL, Palevsky HI, Schwartz JS. Diagnosing pulmonary embolism: new facts and strategies. Ann Intern Med 1991;114:300-6. 5. K6hn H, De Jong R, K6nig B, Mostbeck A. Aerosol delivery system with high lung deposition efficiency [Abstract]. )r Aerosol Med 1988;1:227. 6. K6hn H, K6nig B, Mostbeck A. Incidence and clinical feature of pulmonary embolism in patients with deep vein thrombosis. Eur J Nucl Med 1987;13:11-15. 7. Pavlevsky HI, Alavi A. A noninvasive strategy for the management of patients suspected of pulmonary embolism. Semin Nucl Med 1991;21:325-31. 8. Moser KM. Venous thromboembolism. Am Rev Respir Dis 1990;141:235-49. 9. Alderson PO. Reviewer's comments: Hull RID, Hirsch J,

Pulmonary embolism in ambulant patients with DVT

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

24. 25.

26.

27.

28.

721

Carter CJ, et ai. Diagnostic value ofventilation-perfusionlung scanning in patients with suspected pulmonary embolism. Invest Radiol 1987;22:87-9. Mostbeck A, Partsch H, K6hn H, K6nig B. Lungenembolie bei Bein-Beckenvenenthrombose: Ergebnisse einer prospekriven Studie. Wien Klin Wochenschr 1980;92:464-71. Juni J-E, Alavi A. Lung scanning in the diagnosis of pulmonary embolism: the emperor redressed. Semin Nucl Med 1991;21:281-96. Huisman MV, Biiller HR, ten Care JW, et al. Unexpected high prevalence of silent pulmonary embolism in patients with deep venous thrombosis. Chest 1989;95:498-502. Stiegler H, Weichenhain B, Chatzopulos D, et al. Untersuchungen zur H~iutigkeit und Symptomatologie der Lungenembolie in Abh~ngigkeit von der Lokalisation der tiefen Beinvenenthrombose. Vasa 1991;20:119-24. Moser LM, Le Moine JR. Is embolic risk conditioned by location of deep venous thrombosis? Ann Intern Med 1991;94:439-44. Cuppini S, Cattelan AM, Casara D, Prandoni P. Occult pulmonary embolism in patients with proximal deep venous thrombosis. Ann Ital Med Int 1991;6:1-5. Monreal M, Rey-Joly Barroso C, Ruiz-Manzano J, SalvadorTarrason R, Lafoz-Naval E, Viver-Manresa E. Asymptomafic pulmonary embolism in patients with deep vein thrombosis: is it useful to take a lung scan to rule out this condition? J Cardiovasc Surg 1989;30:104-7. Havig O. Deep vein thrombosis and pulmonary embolism: an autopsy study with multiple regression analysis of possible risk factors. Acta Chir Scand 1977;478(suppl): 1-120. McLachlin J, Paterson JC. Some basic observations on venous thrombosis and pulmonary embolism. Surg Gynecol Obstet 1951;93:1-18. Hull R, Delmore T, Carter C, et al. Adjusted subcutaneous heparin versus warfarin sodium in the long-term treatment of venous thrombosis. N Engl J Med 1982;306:18%94. Ehringer H, Minar E. Die Therapie der akuten BeckenBeinvenenthrombose. Internist 1987;28:317-51. Prandoni P, Lensing AWA~Biiller HR, et al. Comparison of subcutaneous low-molecular-weightheparin with intravenous standard heparin in proximal deep-vein thrombosis. Lancet 1992;339:441-5. Martin M, Eickerling B. T6dliche Lungenembolle unter ultrahoch-dosierter Streptokinase-Behandlung [Letter]. Dtsch ivied Wochenschr 1990;115:1812. Holmstroem M, Bratt G, Toernebohm E, Rhedin AS, Lockner D. Fatal pulmonary embolism caused by streptokinase treatment of deep venous thrombosis of the leg? [[ Intern Med 1990;228:647-9. Plate G, Ohlin P, Ekl6f B. Pulmonary embolism in acute iliofemoral venous thrombosis. Br J Surg 1985;72:912-5. Albada J, Nieuwenhuls HK, Sixma JJ. Treatment of acute venous thromboembolism with low molecular weight heparin (Fragmin): results of a double-blind randomized study. Circulation 1989;80:935-409. yon Liebe S, Kissler M, Zumtobel V. Zur Hiiufigkeit yon Lungenembolien vor und nach Thrombektomie defer BeinBeckenvenenthrombosen. Chirurg 1991;62:482-5. Gallus A, Jackaman J, Tillett J, Mills W, Wycherley A. Safety and efficacy of warfarin started early after submassive venous thrombosis or pulmonary embolism. Lancet 1986;2:1293-6. Monreal M, Ruiz J, Salvador R, Morera J, Arias A. Recurrent pulmonary embolism: a prospective study. Chest 1989;95:

976-9.

722

Journal of VASCULAR SURGERY

Partsch et al.

29. Girard P, Hauuy M-P, Musset D, Simonneau G, Pefitpretz P. Acute inferior vena cava thrombosis: early results of heparin therapy. Chest 1989;95:284-91. 30. Schulman S. Studies on the medical treatment of deep vein thrombosis. Acta Med Scand 1985;704(suppl):1-47. 31. Hull RD, Raskob GE, Rosenbloom D, et al. Heparin for 5 days as compared with 10 days in the initial treatment of proximal venous thrombosis. N Engl J Med 1990;322: 1260-4. 32. Partsch H. Compression therapy of the legs. J Dermatol Surg Oncol 1991;17:799-805. 33. Norris CS, Greenfield LJ, Herrmann JB. Free-floating il-

iofemoral thrombus: a risk of pulmonary embolism. Arch Surg 1985;120:806-8. 34. Baldridge ED, Martin MA, Welting RE. Clinical significance of free-floating venous thrornbi. J VAse SuR6 1990;11: 62-9. 35. Monreal M, Lafoz E, Casals A, et al. Occult cancer in patients with deep vein thrombosis. Cancer 1991;67:541-5. 36. Wheeler FIB. Ambulatory management of suspected deep vein thrombosis. Arch Intern Med 1989;149:501-2. Submitted April 14, 1992; accepted Aug. 6, 1992.

R E S I D E N T R E S E A R C H PRIZE INTERNATIONAL

SOCIETY FOR CARDIOVASCULAR NORTH AMERICAN CHAPTER

SURGERY

The International Society for Cardiovascular Surgery, North American Chapter, desires to stimulate laboratory research in the area of cardiovascular surgery. A resident research award has been established to achieve this goal. The award will consist of a $5000 stipend. In addition, the awardee will receive 1-year complimentary subscriptions to the JOURNALOF VASCULARSURGERY and the Journal of Cardiovascular Surgery. The Society will select a single awardee each year. The Program Committee of the Executive Council of the ISCVS-NA will be responsible for the selection process.

Policies 1. The research must be original and experimental. 2. The research must not be published or submitted for publication (American College of Surgeons Surgical Forum excepted). 3. The research must be performed by a resident in a surgical training program in North America. 4. A member of the ISCVS-NA must be a senior collaborator and assume responsibility for the research. 5. A manuscript must be submitted in English describing the work (six double-spaced copies with appropriate figures prepared in accordance with the Information for Authors of the JOUV,NAL OF VASCULAR SURGERY) and accompanied by a signed letter from the sponsoring member confirming the status of his/her role in the project as well as the submitter's status. The manuscript and an abstract must be submitted for consideration by the Program Committee of the ISCVS-NA for its annual scientific meeting. The prize-winning work will be presented at this meeting. Other submissions may be accepted for presentation even though they do not receive the prize. 6. The deadline for receipt of manuscripts is January 15, 1993. 7. The awardee is encouraged to submit his/her manuscript to the JOURNALOF VASCULAR SURGERY for consideration for publication. 8. Decisions regarding the award will be mailed to the recipient and sponsor by April 1, I993. Manuscripts should be sent and inquiries directed to: Secretary International Society for Cardiovascular Surgery North American Chapter 13 Elm St. Manchester, MA 01944