Download Deep Venous Thrombosis Caused by Femoral Venous Catheters in

Deep Venous Thrombosis Caused by
Femoral Venous Catheters in Critically
Ill Adult Patients*
Gavin M. Joynt, MBBCh; Jacqueline Kew, MBBCh; Charles D. Gomersall, MBBS†;
Vivian Y. F. Leung, PDDR, MPhil; and Eric K. H. Liu, PDDR, MPhil
Study objectives: To determine the frequency of and potential risk factors for catheter-related
deep venous thrombosis (DVT) in critically ill adult patients.
Design: Prospective, controlled, observational cohort study.
Setting: A mixed medical and surgical ICU in a university hospital.
Patients: All adult patients undergoing femoral vein catheterization.
Interventions: None.
Measurements: ICU diagnosis, underlying disease, demographic data, type of catheter, complications during cannulation, use of anticoagulants, coagulation status, medications infused, and
duration of catheterization were recorded. Compression and duplex Doppler ultrasound studies
of both femoral veins were performed prior to insertion, at 12 h after insertion, and daily until
catheter removal. Follow-up investigation was performed at 24 h and 1 week after removal.
Results: Of 140 cases entered into the study, 124 were evaluated. Fourteen patients developed
iliofemoral vein DVTs. Two were clinically obvious. Twelve (9.6%) were line related (uncannulated leg normal) and two (1.6%) occurred only in the uncannulated leg (p 5 0.011; relative risk,
6.0; confidence interval, 1.5 to 23.5). Line-related DVT can occur any time from the day after
insertion to 1 week after removal. The incidence of catheter-related DVT was unrelated to
number of insertion attempts, arterial puncture or hematoma, duration of catheterization,
coagulation status, or type of infused medications. No other predisposing or protective factors
were identified. Three of the 12 patients with catheter-related DVT died. In no patient was
clinical pulmonary embolus suspected.
Conclusion: Although the femoral route is convenient and has potential advantages, the use of
femoral lines increases the risk of iliofemoral DVT. Catheter-related DVT may occur as soon as
1 day after cannulation and is usually asymptomatic. This increased risk should be carefully
considered when the femoral route of cannulation is chosen.
(CHEST 2000; 117:178 –183)
Key words: catheterization; central venous; complications; critically ill; femoral vein; thrombosis
Abbreviations: DVT 5 deep venous thrombosis; PT 5 prothrombin time; PTT 5 partial thromboplastin time
venous catheterization is one of the most
C entral
commonly used invasive procedures in critically
ill patients. The most frequent sites of cannulation
are the internal jugular and subclavian veins.1 Complications include arterial puncture with the internal
jugular approach (0.1 to 4%) and pneumothorax or
*From the Department of Anaesthesia and Intensive Care (Drs.
Joynt and Gomersall) and the Department of Diagnostic Radiology and Organ Imaging (Dr. Kew, Ms. Leung, and Mr. Liu),
Chinese University of Hong Kong, Prince of Wales Hospital,
Shatin, Hong Kong.
†Currently at St. Mary’s Hospital, London, UK.
Manuscript received October 23, 1998; revision accepted May
28, 1999.
Correspondence to: Gavin M. Joynt, MBBCh, Department of
Anaesthesia and Intensive Care, Prince of Wales Hospital, Shatin
N.T., Hong Kong, PRC; e-mail: [email protected]
hemothorax with the subclavian approach (1 to
5%).1–3 Other possible complications include air
embolism, venous thrombosis, thoracic duct laceration, phrenic or recurrent laryngeal nerve damage,
and brachial plexus injury.2
Central venous cannulation via the femoral route
has several potential advantages. There is no risk of
pneumothorax, the site is directly compressible
should bleeding occur, and nerve damage is unlikely.
More sites of access are available for use in patients
requiring multiple cannulae or with restricted access
to the veins of the upper extremities. Femoral
venous catheters can be inserted with relative ease
and have been used successfully in acute resuscitation4,5 and severely burned patients.6 Recent studies
in pediatric7 and adult8 critically ill patients have
178
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
Clinical Investigations in Critical Care
concluded that the femoral vein is a useful site for
central venous cannulation, and these studies quoted
few complications.
Earlier reports of catheter-induced deep venous
thrombosis (DVT) and other complications have
resulted in the restricted use of the femoral route for
central venous catheterization.9,10 However, because
of changes in catheter materials (modern polyurethane catheters are less thrombogenic than polyvinyl
and polyethylene catheters) and improved sterile
techniques in recent years, interest in femoral catheterization has increased.11 As a result, the utility of
femoral venous catheters and the associated complications in critically ill patients are being reevaluated.12–15
The aim of this study was to prospectively evaluate
the incidence and risk factors for catheter-related
DVT associated with the use of femoral catheters in
critically ill patients.
Materials and Methods
The study was performed in a 16-bed, mixed medical and
surgical ICU at a university teaching hospital. Approval for the
study was obtained from the Clinical Research Ethics Committee
of the Chinese University of Hong Kong. Informed consent was
obtained from a senior relative when the patient’s condition
precluded provision of his or her own consent. All adult patients
undergoing femoral-route central venous catheterization between January 1996 and February 1998 were recruited for the
study. The following patients were excluded from the study:
patients with infection or inflammation at the site of insertion;
existing or previous DVT; recent pelvic or abdominal trauma;
lower extremity ischemia; documented hypercoagulable state
(protein C or S deficiency, antithrombin deficiency, or lupus
anticoagulant); prior femoral catheterization; and survival , 24 h
after catheter insertion. Attempts were made not to cannulate the
contralateral femoral vein; where this was unavoidable, the
patient was withdrawn from the study.
Catheters were inserted by intensive care specialists, intensive
care trainees, or senior anesthesiology trainees. All catheters were
inserted using the Seldinger technique.16 The skin was prepared
with an iodine solution and chlorhexidine in alcohol solution.
Strict aseptic conditions were followed, with the operator wearing
mask, gloves, and a surgical gown. Triple-lumen catheters,
introducer sheaths, and dialysis catheters were included in the
study. After insertion, catheters were sutured to the skin and
covered with gauze and an occlusive dressing.
At catheter insertion the following were recorded: Acute
Physiology and Chronic Health Evaluation II score; patient age;
weight; diagnosis; catheter position; catheter type, size, and
length; number of attempts at insertion; arterial puncture; hematoma; and local bleeding. During the catheterization period, local
complications monitored included presence of leg swelling/
discoloration and insertion site infection (erythema or purulent
discharge). Catheter tips were aseptically removed and cultured
by the roll-plate method. Colonization was considered present if
a growth of $ 15 colony-forming units was reported. Catheterrelated bacteremia was considered present if the same organism
was grown from the blood at a peripheral site at the time of
catheter removal. The following were recorded daily: prothrombin time (PT) and partial thromboplastin time (PTT); platelet
count; drugs infused through the femoral line; and systemic drug
therapy. Routine DVT prophylaxis is not practiced in the ICU.
The use of anticoagulants in special circumstances or for any
other purpose, along with the use of other drugs, was recorded.
Investigation and diagnosis of DVT was made by ultrasound
examination. Ultrasound examination was performed with the
patient in the supine position before catheter insertion, within
12 h after insertion, and daily until discharge. A follow-up
ultrasound examination was performed at 24 h and 1 week after
catheter removal. The deep venous systems of both lower
extremities were examined, from the external iliac veins proximally to the popliteal veins distally. The veins were evaluated in
the transverse and longitudinal planes using the compression
technique of Cronan et al,17 supplemented with duplex and color
Doppler capability.18,19 DVT was diagnosed when there was
visualization of thrombus, noncompressibility of the vein, no
spontaneous Doppler flow, lack of phasicity and augmentation,
and Doppler and color spectral flow void. The site of DVT was
recorded. Scans were performed with a diagnostic ultrasound
system (Brüel and Kjaer Medical; Gentoffe, Denmark), using a
high-resolution 5-MHz curved linear array probe. In each patient, the uncannulated contralateral femoral vein served as a
control. All examinations were recorded and reviewed by a
radiologist who was unaware of the patient’s state of coagulation.
Because the catheter is usually clearly visible when present, only
those investigations performed after removal of the catheter
could be adequately blinded.
A power analysis indicated that approximately 140 patients
would be required to show a difference at a 5% significance level
with 80% power, assuming an incidence of 5% (0 to 10%) in the
control leg and 15% (10 to 20%) in the cannulated leg.20 Results
are presented as mean (SD) or median (range), as appropriate.
The Student’s t test, x2 test, and Fisher’s Exact Test were used
where appropriate. A p value of , 0.05 was considered significant.
Results
Over a 25-month period, 140 patients were entered into the study. Sixteen patients were withdrawn after entry into the study: 1 patient was found
to be lupus anticoagulant–positive (this patient developed a line-related DVT); 6 patients survived
, 24 h after insertion of the femoral catheter; 3
patients had attempted cannulation of the contralateral vein; and 6 patients required cannulation of the
contralateral femoral vein before the follow-up period was complete. Thus, 124 patients were evaluated. Twelve patients developed an iliofemoral DVT
on the same side as the femoral catheter. Two
patients developed an iliofemoral DVT in the control
leg (p 5 0.011; relative risk, 6.0; confidence interval,
1.5 to 23.5). All catheters were used for fluid or drug
therapy; some were also used for parenteral nutrition. The right femoral vein was successfully cannulated in 105 patients and the left in 19 patients. The
incidence of catheter-related DVT on the right side
(9/105) and left side (3/19) was similar (p 5 0.3).
Demographic data and catheter characteristics are
presented in Table 1. The type, diameter, and length
of catheter used was not standardized in order to
CHEST / 117 / 1 / JANUARY, 2000
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
179
Table 1—Patient Demographics and Catheter Characteristics*
Characteristics
Age, yr
Female sex, No. (%)
Weight, kg
APACHE II score, median (range)
Diagnostic category, No. (%)
General surgical
Neurosurgical
Medical
Trauma
Burns
Presence of malignancy, No. (%)
Coagulation status†
PT, s
Preinsertion
At 24 h
PTT, s
Preinsertion
At 24 h
Platelet count, 3 109/mL
Preinsertion
At 24 h
Line characteristics
Length, cm
Size, F
Heparin-coated, No. (%)
Days in situ
Median (range)
Patients Without
Catheter-Related
DVT (n 5 112)
Patients With
Catheter-Related
DVT (n 5 12)
53 (18)
41 (37)
66 (11)
21 (9–38)
54 (15)
6 (50)
65 (17)
20 (12–29)
28 (25)
15 (13)
57 (51)
9 (8)
3 (2)
13 (12)
2 (17)
—
9 (75)
1 (8)
—
1 (8)
16 (6)
17 (6)
14 (3)
15 (3)
41 (14)
44 (15)
37 (10)
39 (7)
174 (112)
170 (129)
146 (103)
144 (116)
24 (11)
7 (1)
90 (80)
5 (3)
6 (2–16)
27 (10)
7 (0)
9 (75)
4 (2)
4 (2–7)
*Data expressed as mean (SD) unless otherwise indicated. There were no significant differences between patients with and without
catheter-related DVT; APACHE 5 Acute Physiology and Chronic Health Evaluation.
†Coagulation was assessed immediately before insertion and 24 h after insertion.
reflect the normal clinical setting. Of the 124 patients assessed, cannulation was performed with
triple-lumen catheters in 113 patients, with introducer sheaths in 7 patients, and with dialysis catheters in 4 patients. All DVTs occurred in patients
cannulated with triple-lumen catheters. Neither increased length nor diameter appeared to increase
the rate of thrombosis (Table 1). Heparin impregnation of catheters did not appear to protect against
thrombosis. While the majority of catheters used
were heparin bonded, no increase in the rate of
thrombosis was seen in nonbonded lines (Table 1).
Other possible factors that might have predisposed
patients to the development of DVT, such as the
infusion of hypertonic or irritant medications
through the catheter, were not identified. The following were individually assessed; mannitol, thiopentone, b-lactam antibiotics, erythromycin, metronidazole, potassium, and anticoagulants.
General catheter-related complications are presented in Table 2. All superficial hematomas or hemorrhage responded to local pressure dressing. No surgical intervention or transfusion was required for
therapy. Similarly, no serious or long-term sequelae
resulted from inadvertent arterial puncture; all cases
were successfully treated with digital pressure followed
by a pressure dressing. Catheter colonization occurred
in nine patients (7%). Colonizing organisms were Acinetobacter baumannii (in two patients), Pseudomonas
aeruginosa (one patient), Candida albicans (one patient), and Staphylococcus epidermidis (eight patients).
Three additional cases (2.4%) were associated with
bacteremia/fungemia involving C albicans (one patient)
and S epidermidis (two patients). All cases responded
to line removal and appropriate antibiotic therapy. No
patient exhibited clinical evidence suggesting pulmonary embolism. The causes of death in the two nonsurvivors with confirmed iliofemoral DVT were fulminant
septic shock and multiple organ dysfunction in one
patient and respiratory failure from ARDS in the other
patient.
Discussion
This prospective study demonstrates a consistent
and clinically important increase in iliofemoral DVT
180
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
Clinical Investigations in Critical Care
Table 2—Complications of Femoral Venous
Catheterization*
Complications
Patients Without
DVT
(n 5 112)
Patients With
DVT
(n 5 12)
Insertion attempts, No.
Mean (SD)
Median (range)
Hematoma, No. (%)
Catheter colonization, No. (%)
Catheter bacteremia, No. (%)
Arterial puncture, No. (%)
Leg swelling, No. (%)
Clinical pulmonary embolism
1.6 (0.9)
1 (1–4)
9 (8)
8 (7)
2 (1.7)
8 (7)
1 (1)
—
1.8 (0.6)
2 (1–3)
1 (8)
1 (8)
1 (8)
—
2 (16)
—
*There were no significant differences between patients with and
without catheter-related DVT.
following the use of femoral venous catheters. Patients with femoral lines in situ can expect to have,
on average, a sixfold increased risk of iliofemoral
DVT. Despite this being the largest study to date,
the confidence intervals are wide and the risk may be
as low as twofold or as high as 20-fold the baseline
risk of iliofemoral DVT in noncatheterized limbs.
Other studies performed to specifically examine
iliofemoral catheter-related DVT have reached varying conclusions regarding the increased risk from
catheterization.13,14,21,22 Using venous duplex ultrasound, Meredith et al22 identified a 12% incidence of
line-related iliofemoral DVT (vs 4% in control limb)
in acute trauma patients following the use of 8.5F
catheters for resuscitation, and concluded that routine or indiscriminate use of femoral lines should be
reconsidered. Trottier et al,21 also using duplex
ultrasound, showed a proximal lower limb (not
clearly defined iliofemoral region) DVT rate of 25%
(incidence in control limbs, 0%) following the use of
femoral triple-lumen and dialysis catheters. They
concluded that physicians should be aware of the risk
of the femoral route and consider ultrasound screening of the lower limb after catheter removal. Durbec
et al14 found a 6.6% incidence of iliofemoral DVT
(3% in control limbs) using phlebography and concluded, in contrast to others, that this incidence was
acceptable. In a second study, Durbec et al13 found
an incidence of iliofemoral thrombosis of 8.5%, and
again concluded that this was acceptable. The studies did examine slightly different patient groups and
used different techniques of detecting DVT; however, the results presented above appear remarkably
consistent. These studies, if combined, would produce an increased risk of approximately seven times
the expected baseline risk, similar to that found in
our study. In keeping with the results of other
studies,13,22 the incidence of iliofemoral DVT in the
control limb was low. The few studies that have
examined the incidence of iliofemoral DVT in intensive care patients indicate that the incidence of DVT
in the iliofemoral region, in the absence of any
attempt at femoral vein catheterization of either leg,
is about 2% (range, 0 to 4%).14,21,23 The attributable
increase in risk can therefore be expected to be in
the region of 10 to 15%. This is a risk of clinical
significance and should be carefully considered
whenever femoral catheterization is indicated. The
lack of clinical features of DVT in these patients—
only 2 of 12 patients had evidence of leg swelling
(Table 2)—may produce a false sense of security and
makes knowledge of the risk of this complication
more important.
A unique aspect of this study is that patients were
examined prior to catheter insertion and at least daily
while the catheter was in situ. Absence of DVT prior
to catheterization was thus documented, and the
postinsertion time at which the DVT developed
could be clearly demonstrated. Although it has been
previously suggested that there may be a safe duration of catheterization,13,21 our data clearly show that
catheter-related DVT may occur at any time, from
day 1 to 1 week after removal of the catheter (Fig 1),
and no “safe” period for catheterization exists.
In accordance with preexisting intensive care protocols at the time of the study, routine DVT prophylaxis was not used. This is because the incidence of
DVT in Chinese patients has been reported to be
relatively low.24,25 The results of this study, without
prophylaxis, remain similar to those found in other
studies in which either some or all patients received
DVT prophylaxis.13,14,21,22 The study design does not
permit an answer to the important question of
whether the use of optimal DVT prophylaxis will
reduce the overall risk of catheter-related DVT.
Autoanticoagulation, commonly seen in critically ill
Figure 1. Time of onset of DVT in critically ill patients after
insertion of the femoral catheter. Day 1 and Week 1 refer to the
diagnosis of DVT at 24 h after removal of the catheter and at 1
week of follow-up, respectively.
CHEST / 117 / 1 / JANUARY, 2000
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
181
patients, might have been expected to protect patients against DVT. Surprisingly, the PT and PTT
did not differ in those patients who developed
line-related DVT vs those who did not, suggesting
that anticoagulation is not significantly protective
(Table 1).
The major life-threatening complication of DVT is
pulmonary embolism. The frequency of pulmonary
embolism following catheter-related DVT is unknown. No recent study of catheter-related DVT
following the use of femoral lines has systematically
searched for evidence of pulmonary embolism. Recent studies (representing a total of 743 patients)
evaluating the complications and utility of femoral
lines have reported no cases of clinically evident
pulmonary embolism.4,8,14,21,22 Although this is not
strong evidence of the absence of pulmonary embolism, the likelihood of a high incidence of clinically
important pulmonary embolus following femoral
catheterization is low.
Compression and color duplex Doppler ultrasonography was chosen as the diagnostic modality for this
study because it allows repeat examinations in the ICU,
it is safe, and it has an acceptably high sensitivity and
specificity for detecting iliofemoral DVT.26 Distal and
calf vein DVT was not assessed, as it is unlikely that
distal and calf vein DVT could be caused by the
presence of a femoral catheter. A recent study confirmed that distal vein thrombosis occurs at similar
rates, regardless of the site of catheter insertion,14 and
that the occurrence of distal thrombosis is not associated with femoral vein cannulation.
A disappointing aspect of the study was the inability to identify factors that might be associated with
the occurrence of catheter-related DVT. Reasons for
this failure may include the additive nature of risk
factors for DVT, the small number of patients who
developed DVT, or a failure to document and record
important potential factors. Therefore, on the basis
of this study, no recommendations can be made that
might help identify high-risk patients, improve technique, or guide catheter choice.
When clinicians are deciding whether to use the
thoracic or femoral route, the following important
issues need to be considered. Insertion complications such as arterial puncture (0.1 to 4%) and
pneumothorax or hemothorax (1 to 5%) associated
with the thoracic route should be weighed against
the risk of arterial puncture (8%) when the femoral
route is used (in the event of bleeding, the femoral
artery is readily compressible). The risk of infection
is thought to be greater when the femoral route is
used, but there are no good clinical data in intensive
care patients supporting this view; the one study
comparing catheters in the radial and femoral artery
showed no difference in infectious complications.27
The risk of catheter-related DVT in thoracic veins
(approximately 10%)28 –30 is similar to that found in
femoral veins in this study. The risk of pulmonary
embolus as a result of DVT of the proximal lower
limbs is thought to be higher than from thoracic
vessel DVT31,32; however, we can find no good
evidence to support this in catheter-associated DVT
or critically ill patients. As discussed above, the
incidence of clinically observed pulmonary embolism
in patients undergoing femoral catheterization is
low. The decision must, therefore, be one of clinical
judgment, depending on the benefit-risk ratio in
individual circumstances.
In conclusion, there is a sixfold increased risk of
iliofemoral DVT following femoral vein catheterization. Catheter-related DVT occurrence is not related
to the duration of catheterization, and it can occur at
any time during the course of catheterization or after
removal. There are currently no identified factors
that predict increased risk for developing catheterrelated DVT. In clinical practice, clinicians need to
be aware of an increased risk of iliofemoral DVT, but
the femoral route of venous access remains an
important alternative to the upper extremities.
References
1 Agee KR, Balk RA. Central venous catheterization in the
critically ill patient. Crit Care Clin 1992; 8:677– 686
2 Hagley MT, Martin B, Gast P, et al. Infectious and mechanical complications of central venous catheters placed by
percutaneous venipuncture and over guidewires. Crit Care
Med 1992; 20:1426 –1430
3 Seneff MG. Central venous catheters. In: Rippe JM, Irwin
RS, Fink MP, et al, eds. Intensive care medicine. Boston,
MA: Little, Brown, 1996; 15–35
4 Mangiante EC, Hoots AV, Fabian TC. The percutaneous
common femoral vein catheter in critically injured patients.
J Trauma 1988; 28:1644 –1649
5 Swanson RS, Uhlig PN, Gross PL, et al. Emergency intravenous access through the femoral vein. Ann Emerg Med 1984;
13:244 –247
6 Purdue GF, Hunt JL. Vascular access through the femoral
vessels: indications and complications. J Burn Care Rehabil
1986; 7:498 –500
7 Kanter RK, Zimmerman JJ, Strauss RH, et al. Central venous
catheter insertion by femoral vein: safety and effectiveness for
the pediatric patient. Pediatrics 1986; 77:842– 847
8 Williams JF, Seneff MG, Friedman BC, et al. Use of femoral
venous catheters in critically ill adults: prospective study. Crit
Care Med 1991; 19:550 –553
9 Moncrief JA. Femoral catheters. Ann Surg 1957; 147:166 –
169
10 Bansmer B, Keith D, Tesluk H. Complications following use
of indwelling catheters of inferior vena cava. JAMA 1958;
167:1606 –1611
11 Smyrnios NA, Irwin RS. The jury on femoral vein catheterization is still out. Crit Care Med 1997; 25:1943–1946
12 Joynt GM, Gomersall CD, Buckley TA, et al. Comparison of
intrathoracic and intra-abdominal measurements of central
venous pressure. Lancet 1996; 347:1155–1157
182
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
Clinical Investigations in Critical Care
13 Durbec O, Viviand X, Potie F, et al. A prospective evaluation
of the use of femoral venous catheters in critically ill adults.
Crit Care Med 1997; 25:1986 –1989
14 Durbec O, Viviand X, Potie F, et al. Lower extremity deep
vein thrombosis: a prospective randomized, controlled trial in
comatose or sedated patients undergoing femoral vein catheterization. Crit Care Med 1997; 25:1982–1985
15 Ho KM, Joynt GM, Tan P. A comparison of central venous
pressure and common iliac venous pressure in critically ill
mechanically ventilated patients. Crit Care Med 1998; 26:
461– 464
16 Seldinger SI. Catheter replacement of the needle in percutaneous arteriography. Acta Radiol 1953; 39:368 –376
17 Cronan JJ, Dorfman GS, Scola FH, et al. Deep venous
thrombosis: US assessment using vein compression. Radiology 1987; 162:191–194
18 Cronan JJ, Froehlich JA, Dorfman GS. Image-directed
Doppler ultrasound: a screening technique for patients at
high risk to develop deep vein thrombosis. J Clin Ultrasound
1991; 19:133–138
19 Girasole GJ, Cuomo F, Denton JR, et al. Diagnosis of deep
vein thrombosis in elderly hip-fracture patients by using the
duplex scanning technique. Orthop Rev 1994; 23:411– 416
20 Campbell MJ, Julious SA, Altman DG. Estimating sample
sizes for binary, ordered categorical, and continuous outcomes in two group comparisons. BMJ 1995; 311:1145–1148
21 Trottier SJ, Veremakis C, O’Brien J, et al. Femoral deep vein
thrombosis associated with central venous catheterization:
results from a prospective randomized trial. Crit Care Med
1995; 23:52–59
22 Meredith JW, Young JS, O’Neil EA, et al. Femoral catheters
and deep venous thrombosis: a prospective evaluation with
venous duplex sonography. J Trauma 1993; 35:187–191
23 Hirsch DR, Ingenito EP, Goldhaber SZ. Prevalence of deep
venous thrombosis among patients in medical intensive care.
JAMA 1995; 274:335–337
24 Nandi P, Wong KP, Wei WI, et al. Incidence of postoperative
deep vein thrombosis in Hong Kong Chinese. Br J Surg 1980;
67:251–253
25 Nandi PL, Li WS, Leung R, et al. Deep vein thrombosis and
pulmonary embolism in the Chinese population. Hong Kong
Med J 1998; 4:305–310
26 Cronan JJ. Venous thromboembolic disease: the role of US.
Radiology 1993; 186:619 – 630
27 Russell JA, Joel M, Hudson RJ, et al. Prospective evaluation
of radial and femoral artery catheterization sites in critically ill
patients. Crit Care Med 1983; 11:936 –939
28 Brismar B, Hardstedt C, Jacobson S. Diagnosis of thrombosis
by catheter phlebography after prolonged central venous
catheterization. Ann Surg 1981; 194:779 –783
29 Bozzetti F, Scarpa D, Terno G, et al. Subclavian venous
thrombosis due to indwelling catheters: a prospective study
on 52 patients. JPEN J Parenter Enteral Nutr 1983; 7:560 –
562
30 Becker DM, Philbrick JT, Walker FB 4th. Axillary and
subclavian venous thrombosis: prognosis and treatment. Arch
Intern Med 1991; 151:1934 –1943
31 Hirsh J, Hoak J. Management of deep vein thrombosis and
pulmonary embolism: a statement for healthcare professionals—Council on Thrombosis (in consultation with the Council
on Cardiovascular Radiology), American Heart Association.
Circulation 1996; 93:2212–2245
32 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
CHEST / 117 / 1 / JANUARY, 2000
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21938/ on 04/29/2017
183