(Chest. 2004;125:1595-1596.)
© 2004
American College of Chest Physicians
Prevention of Fatal Pulmonary Embolism in the Hospital
Dominick A. Rascona, MD, FCCP
United States Naval Medical Center Portsmouth, VA
Correspondence to: Dominick A. Rascona, MD, FCCP, Pulmonary and Critical Care Medicine, US Naval Medical Center, Portsmouth, VA 23708; e-mail: darascona{at}mar.med.navy.mil
To the Editor:
A recent discussion with resident physicians regarding the relative benefits of various VTE prophylaxis regimens prompted a review of the article in CHEST by Goldhaber et al1 entitled, "New Onset of Venous Thromboembolism Among Hospitalized Patients at Brigham and Women’s Hospital Is Caused More Often by Prophylaxis Failure Than by Withholding Treatment" (December 2000). To my knowledge, there has been no specific correspondence regarding this study, which concluded with the statement that "most deaths from pulmonary embolism (PE) among patients hospitalized for other conditions occurred in the setting of failed prophylaxis rather than omitted prophylaxis." Several explorations would seem warranted regarding this conclusion.
Goldhaber et al1 described 384 patients who developed venous thromboembolism (VTE) either in the hospital (211 patients) or within 30 days of prior hospital discharge (173 patients). Among this pooled set of patients, 201 had received some form of VTE prophylaxis. Details other than the type of modality employed (eg, the dose of unfractionated heparin or the duration of prophylaxis prior to diagnosis) are not provided.
By the definitions used for failure, an assumption is made that the patients who presented within 30 days of hospital discharge (173 of the total of 384 patients) would have developed their deep vein thrombosis (DVT)/VTE while they were hospitalized and receiving some sort of prophylaxis. If this assumption is rejected, then fully 45% of the patients described cannot be considered as having failed. It is more conceivable that the 173 patients who developed VTE after hospital discharge simply presented with a manifestation of their underlying disease processes while no longer receiving VTE prophylaxis.2
The main conclusion itself, that "most deaths from pulmonary embolism among patients hospitalized for other conditions occurred in the setting of failed prophylaxis rather than omitted prophylaxis" is intriguing but bears closer scrutiny. The authors themselves state that, because the patients they described represented < 0.5% of all hospital admissions, VTE prophylaxis was "quite possibly almost always successful." Since the most important failure of VTE prophylaxis should be considered fatal PE, an attempt to critically understand the conditions surrounding such failures would be useful. However, some of the data that could contribute to such an understanding are confusing:
- Among the patients whose death was attributed to PE and "failure of prophylaxis," it is not clear that these patients were receiving adequate prophylaxis. Inadequate application of an accepted regimen should not be interpreted as that regimen having failed.
- It is not clear exactly how the course to diagnosis of VTE proceeded in the patients who died, as follows:
- • "... 112 patients are described as having PE (62 alone and 50 in association with DVT)... " (in "Results" section of article).
- • "PE was diagnosed primarily by high-probability lung scan (38 patients), intermediate lung scan with high clinical suspicion (50 patients) using revised criteria of the Prospective Investigation of Pulmonary Embolism Diagnosis, or positive pulmonary angiography (33 patients)" (total, 121 patients [in "Materials and Methods" section]).
- • In the 13 patients whose deaths were attributed to PE, that diagnosis was made at autopsy in 8 patients, clinically alone in 2 patients, and by a combination of clinical findings and ultrasound in 3 patients (in Table 6 of the article by Goldhaber et al1). These data would seem to imply that in none of the fatalities was the diagnosis suspected sufficiently to employ one of the three primary diagnostic modalities.
- Seven percent of the patients in whom VTE was diagnosed (26 of 384 patients) in this study did not receive any treatment for the disorder. It is not clear that the patients who died were not overly represented within this group that did not receive treatment.
- It is not possible to discern from the data presented whether, in the cases of fatal PE, any of the patients who failed VTE prophylaxis received diagnoses tardily and/or were treated inappropriately once the suspicion of PE was raised. Because hospitalized patients are likely to be diagnosed and treated promptly for any acute clinical deterioration, this was probably not the case. However, such reassurance is not provided.
It is generally accepted that, once diagnosed, acute nonfatal VTE is at best an uncommon cause of death, recurring in only a small minority of patients, with most deaths actually attributable to underlying diseases.34 Therefore, absent the data mentioned, I wonder whether standard VTE prophylaxis, while not 100% successful in preventing all cases of VTE, might actually come quite close to that percentage in preventing fatal PE. Indeed, a more recent article by Arnold et al5 suggests this may well be the case. In that article, the majority of nonpreventable thromboses (ie, thromboses that occurred despite adequate prophylactic regimens) were lone distal DVT.
In summary, the following three points deriving from a critical appraisal of the study by Goldhaber et al1 are likely to be important when considering the concept of VTE prophylaxis failure:
- It does not follow that patients who go without VTE prophylaxis for any time should be considered as having failed VTE prophylaxis. It is not clear that pooling patients, as was done in the study by Goldhaber et al,1 is valid. Further illustrative of this point, as an example, is data showing that extended (ie, out-of-hospital) VTE prophylaxis is useful in hip surgery patients but not necessarily in those undergoing knee surgery.6 Such a distinction further underscores the importance of unique patient characteristics or disease-associated alterations in coagulability.
- Unique and possibly peculiar individual patient characteristics may be important factors with respect to our evolving understanding of VTE prophylaxis failure. Clinical signs and symptoms, exact methods of prophylaxis, and time to diagnosis and treatment would be useful information that should be available for hospitalized patients. Because hospitalized patients are likely to be diagnosed and treated promptly for any acute clinical deterioration, it is possible that routine standard VTE prophylaxis may be close to 100% effective in the prevention of fatal PE in the hospital.
- I believe there is persistent contention regarding whether subcutaneous unfractionated heparin, administered two or three times daily in doses of either 5,000 or 7,500 U, is adequate prophylaxis in any population of hospitalized patients. No conclusion from this study can be drawn regarding the utility of this inexpensive drug.
References
- Goldhaber, SZ, Dunn, K, MacDougall, RC (2000) New onset of venous thromboembolism among hospitalized patients at Brigham and Women’s Hospital is caused more often by prophylaxis failure than by withholding treatment. Chest 118,1680-1684[Abstract/Free Full Text]
- Schafer, AI Venous thrombosis as a chronic disease. N Engl J Med 1999;340,955-956[Free Full Text]
- Carson, JL, Kelley, MA, Duff, A, et al The clinical course of pulmonary embolism. N Engl J Med 1992;326,1240-1245[Abstract]
- Douketis, JD, Kearon, C, Bates, S, et al Risk of fatal pulmonary embolism in patients with treated venous thromboembolism. JAMA 1998;279,458-462[Abstract/Free Full Text]
- Arnold, DM, Kahn, SR, Shrier, I Missed opportunities for prevention of venous thromboembolism: an evaluation of the use of thromboprophylaxis guidelines. Chest 2001;120,1964-1971[Abstract/Free Full Text]
- Pineo GF. Prevention of venous thromboembolic disease. UpToDate Version 11.3. www.uptodate.com. Topic updated July 16, 2003. Accessed October 10, 2003
This article has been cited by other articles:
|
|
|
|
 
C. G. Elliott, S. Z. Goldhaber, and R. L. Jensen
Delays in Diagnosis of Deep Vein Thrombosis and Pulmonary Embolism
Chest,
November 1, 2005;
128(5):
3372 - 3376.
[Abstract]
[Full Text]
[PDF]
|
|
|
