Correspondence

Thromboembolism after Major Trauma

N Engl J Med 1995; 332:1448-1450May 25, 1995

Article

To the Editor:

The study by Geerts et al. (Dec. 15 issue)1 emphasizes once again the substantial incidence of thromboembolism after major trauma. We were surprised, however, that such a prospective study was undertaken in a group of high-risk patients and that no prophylaxis against thromboembolism was used. As the authors themselves have noted, high rates of deep-vein thrombosis in patients with multiple trauma have been described previously.2 The National Institutes of Health Consensus Development Conference also recognized this fact and recommended prophylaxis for such patients three years before Geerts et al. started their prospective study.3 Why was no prophylaxis used in their patients?

Contrary to the implication of the authors' concluding statement, safe and effective regimens against thromboembolism are already available. In a recent study of 183 patients with multiple trauma who were admitted to our surgical intensive care unit, all of whom had received either subcutaneous heparin or external pneumatic compression, the incidence of proximal deep-vein thrombosis was only 9 percent.4 This is half the rate reported by Geerts et al. and demonstrates, in a comparable group of patients, that effective prophylaxis is available and beneficial in this population.

We also take exception to the authors' lack of enthusiasm for the use of venous duplex ultrasonography in the detection of deep-vein thrombosis. We have found it to have an accuracy of 91 percent.5 It has been easily used at the bedside for the detection of deep-vein thrombosis in patients with trauma.4 On the other hand, the fact that Geerts et al. were able to obtain adequate venograms in less than half of their eligible patients (349 of 716, or 49 percent) underscores the limitations of venography under these circumstances. Impedance plethysmography was performed every other day on their patients. What was their accuracy with this less invasive technique?

Finally, we wonder what “clinical characteristics” of deep-vein thrombosis the authors were evaluating, since only 1.5 percent of their patients with proved deep-vein thrombosis had such characteristics? Although they are certainly not very specific findings, pain or swelling was present in 39 percent of our patients and was associated with a significantly higher likelihood of deep-vein thrombosis. Our biggest problem was not the lack of symptoms but their low predictive value. Therein lies the appeal of a noninvasive screening technique such as duplex ultrasonography.

John Blebea, M.D.
Gregory Strothman, M.D.
University of Cincinnati Medical Center, Cincinnati, OH 45267-0558

5 References

1. 1

   Geerts WH, Code KI, Jay RM, Chen E, Szalai JP. A prospective study of venous thromboembolism after major trauma. N Engl J Med 1994;331:1601-1606
   Full Text | Web of Science | Medline

2. 2

   Shackford SR, Moser KM. Deep venous thrombosis and pulmonary embolism in trauma patients. J Intensive Care Med 1988;3:87-98
   CrossRef

3. 3

   Consensus Development Panel. Prevention of venous thrombosis and pulmonary embolism. JAMA 1986;256:744-749
   CrossRef | Web of Science

4. 4

   Meyer CS, Blebea J, Davis K Jr, Fowl RJ, Kempczinski RF. Surveillance venous scans for deep venous thrombosis in multiple trauma patients. Ann Vasc Surg 1995;9:109-114
   CrossRef | Web of Science | Medline

5. 5

   Patterson RB, Fowl RJ, Keller JD, Schomaker W, Kempczinski RF. The limitations of impedance plethysmography in the diagnosis of acute deep venous thrombosis. J Vasc Surg 1989;9:725-730
   CrossRef | Web of Science | Medline

To the Editor:

The study by Geerts et al. was a cohort study with no control arm, and the venographic interpretations were therefore subject to observer bias. Kappa values measuring observer agreement range from 0.63 to 0.70 even in randomized, controlled trials in which bias is less likely.1,2 In a study without a control arm and in which the observers expect to see a substantial rate of thrombosis, there is a chance of overestimating the rates of venous thrombosis. Were any measures taken to minimize the risk of observer bias? Furthermore, given that the thrombi in asymptomatic patients are substantially smaller and less extensive and more often nonocclusive than those in symptomatic patients,3,4 it would be interesting to know further details on the characteristics of the thrombi: how many were nonocclusive? How many were small and confined to single venous segments? This information would be useful, because many of the reports of smaller thrombi being detected by screening asymptomatic patients with venography have come from clinical trials in which prophylaxis against deep-vein thrombosis was used; prophylaxis was not used in the study by Geerts et al.

The very low rate of fatal pulmonary emboli (0.4 percent) despite the high rate of deep-vein thrombosis is revealing. Although Geerts et al. argue that the rate of fatal pulmonary emboli would have been higher if they had not screened patients with impedance plethysmography and venography, they do not provide any data on false positive rates or on the number of patients whose thrombi were actually detected by impedance plethysmography. Therefore, it is impossible to determine whether screening with impedance plethysmography is useful. Previous reports have suggested that impedance plethysmography has limited value as a screening test in asymptomatic patients.5 We agree that the number of fatalities may have been decreased by treating the cases detected by venography, but wonder whether impedance plethysmography is really necessary.

Philip Wells, M.D.
University of Ottawa, Ottawa, ON K1Y 4E9, Canada

David R. Anderson, M.D.
Victoria General Hospital, Halifax, NS B3H 2Y9, Canada

5 References

1. 1

   Borris LC, Lassen MR. Venography in deep venous thrombosis: postoperative screening of patients in prophylaxis studies. Haemostasis 1993;23:Suppl 1:80-84
   Medline

2. 2

   Couson F, Bounameaux C, Didier D, et al. Influence of variability of interpretation of contrast venography for screening of postoperative deep venous thrombosis on the results of a thromboprophylactic study. Thromb Haemost 1993;70:573-575
   Web of Science | Medline

3. 3

   Wells PS, Lensing AWA, Anderson DR, Hirsh J. The distribution of deep vein thrombosis in asymptomatic post-operative patients and the implications for screening. Blood 1992;80:Suppl:316-316 abstract.
   

4. 4

   Cogo A, Lensing AW, Prandoni P, Hirsh J. Distribution of thrombosis in patients with symptomatic deep vein thrombosis: implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med 1993;153:2777-2780
   CrossRef | Web of Science | Medline

5. 5

   Cruickshank MK, Levine MN, Hirsh J, et al. An evaluation of impedance plethysmography and ¹²⁵I-fibrinogen leg scanning in patients following hip surgery. Thromb Haemost 1989;62:830-834
   Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: The National Institutes of Health Consensus Development Conference stated, “The risks of thromboembolic diseases and subsequent PE [pulmonary embolism] have not been specifically defined for the general trauma population.”1 The purpose of our study was to provide these data.

Prophylaxis was not used in our study for several reasons. Previous reports did not quantitate the overall risk of thrombosis in patients with trauma, and there were large gaps in knowledge of trauma subgroups. When our study was initiated, thromboprophylaxis was not used in our trauma unit and there were no published trials of prophylaxis in trauma. Finally, we thought that plethysmographic and venographic screening provided a margin of patient safety not present before the study.

Although safe and effective methods of thromboprophylaxis may be available, none have been adequately tested using methodologically rigorous trials in trauma. The study cited by Blebea and Strothman was a retrospective review of nonconsecutive patients that was not designed to test prophylaxis and that used a diagnostic test (venous ultrasonography) that has not been validated for this purpose. It is possible that the two interventions used in their study are of some benefit in patients with trauma, but this is not the type of study that should result in widespread adoption of these methods.

Blebea and Strothman misrepresent our position on the diagnostic usefulness of venous ultrasonography. Although we agree that this method is accurate for the diagnosis of proximal deep-vein thrombosis in symptomatic patients, there is considerable evidence that it has substantially reduced sensitivity when used in high-risk, asymptomatic patients.2-4 The benefit of venous ultrasonography has never been validated in proper prospective studies in which venograms were used as a reference in large numbers of patients with trauma. The study cited supporting the high degree of accuracy of ultrasonography was retrospective, and ultrasonography was performed only in symptomatic patients, of whom only 3 percent underwent venography.

A clinical suspicion of deep-vein thrombosis in our study was aroused by attending trauma surgical staff in the context of swelling and pain frequently present with leg injuries. We therefore confirm that clinical characteristics have poor positive and negative predictive value for thrombosis in trauma.

We agree with Wells and Anderson that the interpretation of venograms is subject to bias. We therefore adopted conservative criteria for the diagnosis of thrombosis.

William H. Geerts, M.D.
Richard M. Jay, M.D.
Sunnybrook Health Science Centre, Toronto, ON M4N 3M5, Canada

4 References

1. 1

   Consensus Development Panel. Prevention of venous thrombosis and pulmonary embolism. JAMA 1986;256:744-749
   CrossRef | Web of Science

2. 2

   Ginsberg JS, Caco CC, Brill-Edwards PA, et al. Venous thrombosis in patients who have undergone major hip or knee surgery: detection with compression US and impedance plethysmography. Radiology 1991;181:651-654
   Web of Science | Medline

3. 3

   Davidson BL, Elliott CG, Lensing AW. Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. Ann Intern Med 1992;117:735-738
   Web of Science | Medline

4. 4

   Elliott CG, Suchyta M, Rose SC, et al. Duplex ultrasonography for the detection of deep vein thrombi after total hip or knee arthroplasty. Angiology 1993;44:26-33
   CrossRef | Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   Gregory Strothman, John Blebea, Richard J. Fowl, Gary Rosenthal. (1995) Contralateral duplex scanning for deep venous thrombosis is unnecessary in patients with symptoms. Journal of Vascular Surgery 22:5, 543-547
   CrossRef