Upper Extremity Deep Vein Thrombosis: A Community-Based Perspective

Article Outline

• Abstract
• Methods
  • Data Collection
  • Analysis
• Results
  • Incidence and Attack Rates of Venous Thromboembolism
  • Characteristics of Patient with Upper Extremity Deep Vein Thrombosis versus Lower Extremity Deep Vein Thrombosis
  • Characteristics of Patients with Upper Extremity Deep Vein Thrombosis and Recent History of Central Venous Catheter
  • History of Prophylaxis
  • Treatment Practices
  • Outcomes
• Discussion
  • Clinical Profile of Patients with Upper Extremity Deep Vein Thrombosis
  • Prophylaxis
  • Treatment and Outcomes
  • Study Limitations
• Conclusions
• Acknowledgments
• Appendix
• References
• Copyright

Abstract 

Purpose

The purpose of this study was to examine the magnitude, risk factors, management strategies, and outcomes in a population-based investigation of patients with upper, as compared with lower, extremity deep vein thrombosis diagnosed in 1999.

Methods

The medical records of all residents from Worcester, Massachusetts (2000 census=478,000) diagnosed with ICD-9 codes consistent with possible deep vein thrombosis at all Worcester hospitals during 1999 were reviewed and validated.

Results

The age-adjusted attack rate (per 100,000 population) of upper extremity deep vein thrombosis was 16 (95% confidence interval [CI], 13-20) compared with 91 (95% CI, 83-100) for lower extremity deep vein thrombosis. Patients with upper extremity deep vein thrombosis were significantly more likely to have undergone recent central line placement, a cardiac procedure, or an intensive care unit admission than patients with lower extremity deep vein thrombosis. Although short-term and 1-year recurrence rates of venous thromboembolism and all-cause mortality were not significantly different between patients with upper, versus lower, extremity deep vein thrombosis, patients with upper extremity deep vein thrombosis were less likely to have pulmonary embolism at presentation or in follow-up.

Conclusions

Patients with upper extremity deep vein thrombosis represent a clinically important patient population in the community setting. Risk factors, occurrence of pulmonary embolism, and timing and location of venous thromboembolism recurrence differ between patients with upper as compared with lower extremity deep vein thrombosis. These data suggest that strategies for prophylaxis and treatment of upper extremity deep vein thrombosis need further study and refinement.

Keywords: Deep vein thrombosis, Incidence, Population-based, Upper extremity

It is generally assumed that the occurrence of upper extremity deep vein thrombosis has increased in the last several decades. However, the actual incidence of upper extremity deep vein thrombosis in the community setting, the profile of patients at increased risk for the development of upper extremity deep vein thrombosis, and the clinical sequelae of this condition, particularly with regards to development of recurrent venous thromboembolic events, remains unclear.

The objectives of the Worcester Venous Thromboembolism study are to provide more contemporary population-based data about the clinical epidemiology of venous thromboembolism, as well as its management and associated outcomes. The purpose of the present investigation was to describe and compare incidence rates, patient profiles, management strategies, and subsequent outcomes in residents of the Worcester, Massachusetts metropolitan area diagnosed with upper versus lower extremity deep vein thrombosis in 1999.

Back to Article Outline

Methods 

Computerized printouts of all Worcester residents with health care encounters in which any of 34 ICD-9 diagnosis codes possibly consistent with venous thromboembolism (Appendix A, available online) had been listed in 1999 were obtained from each of the 12 hospitals serving the Worcester area. These data queries were not limited to discharge diagnoses but also encompassed all outpatient activities. In order to identify Worcester residents with potential venous thromboembolism who sought care outside of the Worcester area, we queried the Massachusetts Health Data Consortium, which collects information on all Massachusetts residents seeking health care at hospitals throughout Massachusetts, as well in adjacent states.

The medical records of all persons meeting the geographic inclusion criteria were subsequently reviewed and validated by trained abstractors using prespecified criteria based on a modification of a classification schema proposed by Silverstein et al (Appendix B, available online).¹ Patients were considered to have an upper extremity deep vein thrombosis if an internal jugular, innominate, subclavian, or axillary vein thrombosis was confirmed by ultrasonography or venography. Patients were considered to have a lower extremity deep vein thrombosis if thrombosis of the iliac, femoral, popliteal, or calf veins was confirmed by ultrasonography or venography.

Data Collection 

Information was collected from medical records about patient demographic and clinical characteristics, diagnostic test results, and hospital management practices. Surgery included major operations where general or epidural anesthesia lasted 30 minutes or longer. Medical history variables defined as “recent” were those occurring or active in the 3 months before deep vein thrombosis.

Simultaneous conduct of medical record review at all area hospitals and review of state and national mortality records enabled collection of information about recurrent venous thromboembolism, major bleeding, or mortality at the same time as information about the index event. Some form of additional follow-up was obtained in >99% of all patients.

Potential cases of recurrence of venous thromboembolism were classified using criteria similar to that employed for incident cases—however, a definite recurrence of deep vein thrombosis required the new occurrence of thrombosis in a previously uninvolved venous or pulmonary segment. Major bleeding was defined as any episode of bleeding requiring transfusion, resulting in cessation of anticoagulants, or resulting in hospitalization (or prolongation of hospitalization), stroke, myocardial infarction, or death.

Analysis 

Rates of initial and total deep vein thrombosis, stratified according to location (upper versus lower), were calculated based on US census estimates of the Worcester population in 2000 (n=477,800). Differences in the distribution of characteristics between patients with upper versus lower extremity deep vein thrombosis were examined using chi-squared tests of statistical significance for categorical variables and t tests for continuous variables.

Multivariate regression analysis was carried out to identify medical history variables independently associated with upper as compared with lower extremity deep vein thrombosis. Candidate variables for inclusion in the regression model included patient age, sex, race, current hospitalization, and the medical history variables listed in Table 1. Candidate variables possibly associated with the outcomes of interest (P <.25 after univariate analysis) were included in the multivariate models. Variables with P >.05 were eliminated in a stepwise fashion so that only variables with a statistically significant association with the outcome of interest were included in the final regression models.

Table 1. Demographic and Clinical Characteristics of Patients According to Location of Deep Vein Thrombosis^⁎

Variable	Upper Extremity Deep Vein Thrombosis (n=69)	Lower Extremity Deep Vein Thrombosis (n=414)	P Value
Demographic factors
Age (mean, years)	59	66	<.001
Age (years, %)
<55	42.0	25.6
55-64	8.7	11.1	.04
65-74	18.8	22
≥75	30.4	41.3
Female (%)	47.8	54.6	.29
Race (%)
Caucasian	86.7	91.7
Black	1.5	3.2	.002
Asian	1.5	0.0
Hispanic	8.8	2.0
Unknown	1.4	3.1
Body mass index (%)†			.02
<25	45	34.7
25-30	42.5	30.3
>30	12.5	35.0
Risk factors (%)
>48 h bed rest in last month	39.1	46.9	.48
Recent prior hospitalization‡	56.5	36.5	<.001
Recent surgery‡	48.5	27.9	<.001
Recent malignancy‡	43.5	32.8	.07
Recent severe infection‡	49.3	32.4	.01
Admission of non-venous thromboembolism-related diagnosis (immediately prior venous thromboembolism)	43.5	23	<.001
Recent central venous catheter‡	62.3	11.8	<.001
Prior venous thromboembolism	8.7	19.8	.03
Recent intensive care unit discharge‡	24.6	15.2	.05
Recent intubation‡	18.8	15.0	.41
Recent hormonal therapy‡	7.3	7.7	.89
Recent fracture‡	15.9	10.1	.15
Recent chemotherapy‡	20.3	6.3	<.001
Recent heart failure‡	8.7	5.3	.26
Recent cardiac procedures‡	11.6	3.3	.001

Back to Article Outline

Results 

The study sample consisted of 483 Worcester men and women with validated acute deep vein thrombosis. Of these, 69 patients were diagnosed with upper extremity deep vein thrombosis (14%), while the majority (86%) was diagnosed with lower extremity deep vein thrombosis. The mean age of the study sample was 65 years, 54% were women, and 91% were Caucasian.

Incidence and Attack Rates of Venous Thromboembolism 

The age-adjusted incidence and total rates of upper extremity deep vein thrombosis were 15 (95% confidence interval [CI], 12-19) and 16 (95% CI, 13-20) respectively, compared with 74 (95% CI, 67-82) and 91 (95% CI, 83-100) per 100,000 population for lower extremity deep vein thrombosis. Incidence and total rates of upper extremity deep vein thrombosis did not differ significantly by sex (males: 16 and 17, respectively, per 100,000; females 14 and 15, respectively, per 100,000).

Characteristics of Patient with Upper Extremity Deep Vein Thrombosis versus Lower Extremity Deep Vein Thrombosis 

Patients with upper extremity deep vein thrombosis were younger, more likely to be nonwhite, and had a lower body mass index than patients with lower extremity deep vein thrombosis (Table 1). Patients with upper extremity deep vein thrombosis were more likely to experience deep vein thrombosis during a hospital admission for a non-venous-thromboembolism-related diagnosis, to have had a recent central venous catheter, infection, active malignancy, or intensive care unit discharge, but less likely to have a prior history of venous thromboembolism.

After multivariate analysis, patients with upper extremity deep vein thrombosis were more likely to have had a recent central venous catheter (odds ratio [OR] 21.7, 95% CI, 9.3-50.0), recent cardiac procedure (OR 4.2, 95% CI, 1.2-14.2), or recent intensive care unit discharge (OR 3.8, 95% CI, 1.4-10) than patients with a lower extremity deep vein thrombosis.

Characteristics of Patients with Upper Extremity Deep Vein Thrombosis and Recent History of Central Venous Catheter 

Of the 69 patients with upper extremity deep vein thrombosis, 43 (62%) had a history of recent central venous catheter placement. In 38 of these 43 patients, central venous catheter position was in the same upper extremity as the subsequent deep vein thrombosis. Type/location of central line placement in these patients included PICC (peripherally inserted central catheters) (29%), internal jugular (29%), subclavian (12%), and other (eg, Hickman, Tessio, Groshong) (30%).

Patients with upper extremity deep vein thrombosis and a history of recent central venous catheter placement were more likely to have a history of recent hospitalization, surgery, severe infection, intensive care unit discharge, intubation, or fracture than patients without central venous catheter placement (Table 2).

Table 2. Demographic and Clinical Characteristics of Patients with Upper Extremity Deep Vein Thrombosis According to Recent Central Venous Catheter Status

Variable	Central Venous Catheter (+) (n=43)	Central Venous Catheter (−) (n=26)	P Value
Demographic factors
Age (years, %)			.41
<55	37.2	50
55-64	11.6	3.9
65-74	16.3	23.1
≥75	34.9	23.1
Female (%)	46.5	50	.78
Race (%)			.17
Caucasian	78.6	100
Black	2.4	0.0
Asian	2.4	0.0
Hispanic	14.3	0.0
Unknown	2.3	0.0
Body mass index (%)⁎			.48
<25	48	40.0
25-30	36	53.3
>30	16	6.67
Risk factors (%)
>48 h bed rest in last month	46.5	26.9	.23
Recent prior hospitalization†	67.4	38.5	.001
Recent surgery†	60.5	28.0	.01
Recent malignancy†	39.5	50	.72
Recent severe infection†	62.8	26.9	.004
Admission of non-VTE-related diagnosis (immediately prior VTE)	53.5	26.9	.08
Prior VTE	7.0	11.5	.51
Recent intensive care unit discharge†	34.9	0.0	.01
Recent intubation†	30.2	15.0	.002
Recent hormonal therapy†	7.0	7.7	.91
Recent fracture†	25.6	0.0	.005
Recent chemotherapy†	25.6	11.5	.16
Recent heart failure†	9.3	7.7	.81
Recent cardiac procedures†	11.6	11.5	.99

History of Prophylaxis 

We examined the prior utilization of deep vein thrombosis prophylaxis in three subsets of patients who developed deep vein thrombosis during or after hospitalization or surgery: patients who developed deep vein thrombosis during hospitalization for another illness (n=125); patients who developed deep vein thrombosis as outpatients but had been hospitalized in the preceding 3 months (n=190); and patients who developed deep vein thrombosis within the 3 months following surgery (n=147; 95 of these patients were also included in group 2). Prior utilization of anticoagulation prophylaxis during these high-risk periods in patients with upper extremity deep vein thrombosis was 60%, 49%, and 56%, respectively. Corresponding rates of anticoagulant prophylaxis in patients with lower extremity deep vein thrombosis were 43%, 36%, and 36%, respectively.

Treatment Practices 

Patients with upper extremity deep vein thrombosis were less likely to be acutely treated with unfractionated heparin, but equally likely to be treated with low-molecular-weight heparin, as compared with patients with lower extremity deep vein thrombosis (Table 3). Fewer patients with upper extremity deep vein thrombosis were prescribed warfarin at hospital discharge than those with lower extremity deep vein thrombosis.

Table 3. Treatment Strategies in Patients with Venous Thromboembolism According to Location of Deep Vein Thrombosis

Variable	Upper Extremity Deep Vein Thrombosis	Lower Extremity Deep Vein Thrombosis	P Value
Hospital therapy (%)
IV heparin	47.8	61.8	.03
SQ Enoxaparin	69.6	70.5	.87
Other parenteral anticoagulant	4.4	0.5	.03
Warfarin	56.5	73.7	.004
Aspirin	15.2	7.8	.07

Outcomes 

None of the 30-day, 6-month, or 1-year outcomes (major bleeding, venous thromboembolism recurrence, mortality) were significantly different in patients with upper versus lower extremity deep vein thrombosis (Table 4). Of the 10 recurrent deep vein thromboses occurring in patients with upper extremity deep vein thrombosis, 7 occurred in the ipsilateral upper extremity. Of 40 recurrent deep vein thromboses occurring in patients with lower extremity deep vein thrombosis, 26 were in the ipsilateral leg, 10 were in the contralateral leg, and 4 were in the upper extremity.

Table 4. Outcomes of Patients with Venous Thromboembolism According to Location of Deep Vein Thrombosis

Variable	Upper Extremity Deep Vein Thrombosis	Lower Extremity Deep Vein Thrombosis	P Value
30 day outcomes (%)
Major bleeding	11.6	7.5	.25
Recurrent VTE	8.7	4.6	.15
Recurrent pulmonary embolism (with or without deep vein hrombosis)	1.5	1.2	.87
Mortality	4.4	5.8	.63
6 month outcomes (%)
Major bleeding	13.0	9.9	.43
Recurrent VTE	15.9	9.4	.10
Recurrent pulmonary embolism (with or without deep vein thrombosis)	1.5	2.2	.70
Mortality	14.5	12.1	.57
1 year outcomes (%)
Major bleeding	13.3	10.9	.60
Recurrent VTE	14.5	11.4	.47
Recurrent pulmonary embolism (with or without deep vein thrombosis)	1.5	2.9	.46
Mortality	20.3	14.7	.25

VTE=venous thromboembolism.

None of the patients presenting with upper extremity deep vein thrombosis were also diagnosed with a concomitant pulmonary embolism, whereas pulmonary embolism was clinically recognized in 15% of patients presenting with lower extremity deep vein thrombosis. Only 1 patient with upper extremity deep vein thrombosis was diagnosed with pulmonary embolism at 30 days—no other clinically recognized pulmonary embolism occurred in this group over a 1-year follow-up period. Five patients with lower extremity deep vein thrombosis suffered a pulmonary embolism within the first month; an additional 7 patients developed clinically recognized pulmonary embolism by 1 year.

All-cause mortality at 1, 6, and 12 months did not differ significantly between patients with upper versus lower extremity deep vein thrombosis.

Back to Article Outline

Discussion 

This population-based study provides the only available data on the actual incidence of upper extremity deep vein thrombosis in a well-defined community. Approximately 1 in 7 deep vein thromboses in this community of approximately 500,000 people occurred in the upper extremity. Extrapolated to the US population, our data suggest approximately 50,000 cases of upper extremity deep vein thrombosis occur annually.

Clinical Profile of Patients with Upper Extremity Deep Vein Thrombosis 

We were able to identify important differences in the clinical profile of patients experiencing upper versus lower extremity deep vein thrombosis. Patients with upper extremity deep vein thrombosis were twice as likely to suffer their event during hospitalization and were more likely to have additional hospital-related risk factors. Most notably, approximately 60% of all patients with upper extremity deep vein thrombosis had undergone a central line placement in the preceding 3 months, compared with 10% of those with lower extremity deep vein thrombosis. After multivariate analysis, recent central line placement was the clinical variable most strongly associated with upper extremity (as opposed to lower) deep vein thrombosis.

Clinical characteristics of patients with upper extremity deep vein thrombosis compared with those with lower extremity deep vein thrombosis also were examined in a U.S. multi-center registry of nearly 5500 patients, 592 of whom had upper extremity deep vein thrombosis.² Data from this study suggested that upper extremity deep vein thrombosis occurred in approximately 70% of patients during hospitalization. As in our study, a recent central venous catheter was the strongest independent predictor of upper (versus lower) extremity deep vein thrombosis. Similarly, in a nested population-based case-control study of 625 patients with deep vein thrombosis from 1976 to 1990, recent central line or cardiac pacemaker placement was independently associated with an increased risk of deep vein thrombosis—approximately 9% of all deep vein thromboses within this community were attributable to these procedures.³

These findings also have important implications for the utilization of deep vein thrombosis prophylaxis. Because the most important risk factors of upper extremity deep vein thrombosis prophylaxis are identifiable and are hospital-related, optimal targeting of prophylaxis to patients at risk should be achievable.

Prophylaxis 

Anticoagulant prophylaxis use during high-risk periods in patients subsequently developing upper extremity deep vein thrombosis (DVT) was suboptimal. The DVT-Free Study reported an even lower utilization rate of anticoagulant prophylaxis in the 30 days before upper extremity deep vein thrombosis (∼33%).²

Our data suggest that placement of a central venous catheter, particularly in acutely ill hospitalized patients, represents a readily identifiable risk factor for upper extremity deep vein thrombosis. Unfortunately, there is very limited literature describing the efficacy of prophylaxis in such patients. Studies assessing prophylaxis with mini-dose warfarin or low-molecular-weight heparins in patients with in-dwelling central lines have been limited to ambulatory cancer patients and have provided mixed results.⁴, ⁵, ⁶, ⁷ No clear benefits associated with prophylaxis in these patients have been reproducibly demonstrated, and in one study, patients receiving mini-dose warfarin suffered an increase in bleeding.⁶ In our study, most of the patients experiencing upper extremity deep vein thrombosis were nonambulatory or hospitalized with an acute illness, approximately 25% had a recent intensive care unit stay, and most had additional risk factors, suggesting that they were at greater risk for both thrombosis and for bleeding than previously studied patients. Further studies are needed to evaluate the efficacy and safety of prophylaxis in nonambulatory patients at increased risk for upper extremity deep vein thrombosis, particularly those undergoing placement of a central venous catheters or requiring an intensive care unit admission.

Treatment and Outcomes 

The majority of patients with deep vein thrombosis in our community study were treated acutely with a heparin product. However, prescription of warfarin at the time of hospital encounter was significantly lower in patients with upper extremity than in those with lower extremity deep vein thrombosis. Aspirin therapy was prescribed at discharge in twice as many patients with upper, versus lower, extremity deep vein thrombosis, suggesting that some clinicians may still consider aspirin to be an acceptable alternative to warfarin therapy. Although recent guidelines from the American College of Chest Physicians suggest that upper extremity deep vein thrombosis should be treated similarly to lower extremity deep vein thrombosis,⁸ our data suggest a lingering perception among clinicians that upper extremity deep vein thrombosis does not necessarily require aggressive treatment.

Our data also provide some insights about outcomes associated with this condition in the community setting. At 30 days and 6 months, recurrent venous thromboembolism was approximately 1.7 times more frequent in patients with upper extremity deep vein thrombosis than in those with lower extremity deep vein thrombosis. Interestingly, the temporal profile of recurrent events differed according to initial thrombosis location. Patients with upper extremity deep vein thrombosis experienced all of their recurrent events within the first 6 months of diagnosis, whereas approximately one quarter of recurrent events in patients with lower extremity deep vein thrombosis occurred between 6 months and 1 year after diagnosis.

None of the patients with upper extremity deep vein thrombosis were diagnosed with pulmonary embolism at presentation, whereas pulmonary embolism was clinically recognized in approximately 15% of patients with lower extremity deep vein thrombosis. During follow-up, 1 patient with upper extremity deep vein thrombosis (1.5%) suffered a pulmonary embolism by 30 days and there were no pulmonary embolism events noted thereafter. In contrast, the cumulative rate of pulmonary embolism increased from 1.2% at 30 days to 2.2% at 6 months, and to 2.9% at 1 year in patients with lower extremity deep vein thrombosis.

These observations suggest that the population at risk, the natural history, and associated outcomes of upper extremity deep vein thrombosis differ from that of lower extremity deep vein thrombosis. Further studies are needed if we are to effectively target prophylaxis, improve diagnosis, and optimize our treatment (anticoagulant type, duration, intensity) of this condition.

Study Limitations 

It is important to recognize that our study is limited by its retrospective observational design. Information about medical history variables and clinical characteristics is limited to that available from the medical record. In addition, although this is one of the larger studies of patients with upper extremity deep vein thrombosis, its sample size is still relatively small. As such, these data should not be used to support the contention that upper extremity deep vein thrombosis does not warrant treatment to prevent life-threatening pulmonary embolism. Indeed, in a study of 27 patients with venographically confirmed upper extremity deep vein thrombosis, 8 (36%) had objective evidence for concomitant pulmonary embolism by ventilation-perfusion scan or pulmonary angiography.⁹ Clearly, the incidence of pulmonary embolism associated with catheter and non-catheter-related upper extremity deep vein thrombosis requires further study.

Back to Article Outline

Conclusions 

In summary, approximately 14% of cases of deep vein thrombosis in this community-based study occurred in the upper extremity. The clinical profile of patients with upper extremity deep vein thrombosis varied considerably from patients with deep vein thrombosis of the lower extremities. These findings have important implications for the targeting of deep vein thrombosis prophylaxis in patients at risk for upper extremity deep vein thrombosis. Patients with upper extremity deep vein thrombosis were unlikely to experience pulmonary embolism (at presentation or during follow-up), and recurrent deep vein thrombosis tended to occur early and in the previously affected limb. These findings suggest that further study is warranted to define appropriate treatment (eg, agent, intensity, duration) in patients with upper extremity deep vein thrombosis.

Back to Article Outline

Acknowledgments 

Dr. Frederick Spencer had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Back to Article Outline

Appendix 

Appendix A.

Appendix B.

Back to Article Outline

References 

1. Silverstein MD, Heit JA, Mohr DN, et al. Trends in the incidence of deep vein thrombosis and pulmonary embolism (A 25-year population-based study). Arch Intern Med. 1998;158:585–593
   • View In Article
   • MEDLINE
   • CrossRef
2. Joffe HV, Kucher N, Tapson VF, Goldhaber SZ Deep Vein Thrombosis (deep vein thrombosis) FREE Steering Committee. Upper-extremity deep vein thrombosis: a prospective registry of 592 patients. Circulation. 2004;110:1605–1611
   • View In Article
   • CrossRef
3. Heit JA, O’Fallon WM, Petterson TM, et al. Relative impact of risk factors for deep vein thrombosis and pulmonary embolism: a population-based study. Arch Intern Med. 2002;162:1245–1248
   • View In Article
   • MEDLINE
   • CrossRef
4. Bern MM, Lokich JJ, Wallach SR, et al. Very low doses of warfarin can prevent thrombosis in central venous catheters: a randomized prospective trial. Ann Intern Med. 1990;112:423–428
   • View In Article
   • MEDLINE
5. Heaton DC, Han DY, Inder A. Minidose (1 mg) warfarin as prophylaxis for central vein catheter thrombosis. Intern Med J. 2002;32:84–88
   • View In Article
   • MEDLINE
   • CrossRef
6. Masci G, Magagnoli M, Zucali PA, et al. Minidose warfarin prophylaxis for catheter-associated thrombosis in cancer patients: can it be safely associated with fluorouracil-based chemotherapy?. J Clin Oncol. 2003;21:736–739
   • View In Article
   • MEDLINE
   • CrossRef
7. Monreal M, Alastrue A, Rull M, et al. Upper extremity deep venous thrombosis in cancer patients with venous access devices: prophylaxis with a low-molecular weight heparin (Fragmin). Thromb Haemost. 1996;75:251–253
   • View In Article
   • MEDLINE
8. Buller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease. Chest. 2004;126:401S–428S
   • View In Article
   • MEDLINE
   • CrossRef
9. Prandoni P, Polistena P, Bernardi E, et al. Upper-extremity deep vein thrombosis: risk factors, diagnosis, and complications. Arch Intern Med. 1997;157:57–62
   • View In Article
   • MEDLINE