0
Original Investigation |

The Outpatient Bleeding Risk Index:  Validation of a Tool for Predicting Bleeding Rates in Patients Treated for Deep Venous Thrombosis and Pulmonary Embolism FREE

Philip S. Wells, MD; Melissa A. Forgie, MD; Maureen Simms, BSc; Alison Greene; Donna Touchie, BSc; Geoffrey Lewis, MSc; Jodi Anderson; Marc A. Rodger, MD
[+] Author Affiliations

From the Departments of Medicine and Epidemiology, University of Ottawa, Ottawa Health Research Institute (Drs Wells, Forgie, and Rodger, and Mss Simms, Greene, Touchie, and Anderson), and the Department of Pharmacy, Ottawa Hospital (Mr Lewis), Ottawa, Ontario. The authors have no relevant financial interest in this article.


Arch Intern Med. 2003;163(8):917-920. doi:10.1001/archinte.163.8.917.
Text Size: A A A
Published online

Background  Long-term anticoagulation prevents recurrent thrombosis in patients with idiopathic deep venous thrombosis or pulmonary embolism, but with a risk of clinically important so-called major bleeding. Physician- and patient-based decisions on the optimal duration of therapy are sensitive to the bleeding risk. The Outpatient Bleeding Risk Index potentially provides a means of calculating the potential risk of bleeding using easily elicited clinical findings, but, to our knowledge, the authors of the index have provided the only published validation of it. We sought to determine the accuracy of the index in our population of patients.

Methods  We prospectively applied the Outpatient Bleeding Risk Index to consecutive patients in our clinic who had been objectively diagnosed as having pulmonary embolism or deep venous thrombosis and who were about to undergo standard therapy. Standard therapy consisted of a minimum of 5 days of low-molecular-weight heparin therapy overlapped with warfarin sodium therapy, and continuation of warfarin therapy for at least 3 months, with a target international normalized ratio of 2.5. Patients were placed in 3 risk groups (low, moderate, or high), as defined by the index. The survival curves of the groups, using major hemorrhages as the events, were then compared by the log-rank test.

Results  Bleeding rates were lower than expected, but the index did discriminate between low- and moderate-risk groups (P = .03, log-rank test). The rate of major hemorrhage per 100 person-years was 0% (95% confidence interval, 0%-2.8%) in the low-risk group and 4.3% (95% confidence interval, 1.1%-11.1%) in the moderate-risk group. The rate in the high-risk group could not be defined because only 2 patients were at high risk.

Conclusion  The Outpatient Bleeding Risk Index discriminates between low- and moderate-risk patients, and could be used to guide decisions on the optimal duration of anticoagulant therapy.

Figures in this Article

IT IS APPARENT that long-term anticoagulation for patients with idiopathic deep venous thrombosis or pulmonary embolism prevents recurrent thrombosis. Recent studies suggest that the recurrent thrombosis rate may be as high as 27% per year, but the risk of major hemorrhage is 3.7% per year.1 In patients untreated, the major hemorrhage rate was 0%. Clearly, the trade-off to ongoing anticoagulation is the risk of major bleeding. Recent decision analyses2,3 suggest that rates of major hemorrhage have a significant impact on determining the optimal duration of anticoagulant therapy, from a patient and a physician perspective. The decision analysis performed by our group demonstrated that patient decisions are most sensitive to the rate of major bleeding and the burden of anticoagulant therapy. If major bleeding rates are less than approximately 2.5% per year, lifelong therapy is favored in patients with idiopathic deep venous thrombosis. A valid predictive index to determine bleeding risks would, therefore, be of benefit.

To estimate the risk for major bleeding during anticoagulant therapy, a few prediction rules have been published. One of the more promising indexes was based on a study by Landefeld et al4 that identified 5 independent risk factors for major bleeding—65 years or older, history of stroke, history of gastrointestinal (GI) tract bleeding, a serious comorbid condition (recent myocardial infarction, renal insufficiency, or severe anemia), and atrial fibrillation. Beyth et al5 developed a modified Outpatient Bleeding Risk Index, and found that cumulated rates of major bleeding at 12 months were 3%, 12%, and 48% in patients considered low, moderate, and high risk, respectively, in their derivation data set. The validation study indicated the rates were 3%, 8%, and 30%, respectively, over 12 months. An index that can accurately determine the risk of bleeding may help to make decisions on anticoagulant use. Those patients identified as high risk may benefit from closer monitoring or shorter duration of their anticoagulant therapy. To assess the potential benefit of using the Outpatient Bleeding Risk Index and to determine its accuracy in our patients, we applied the rule to patients at our institution.

We prospectively applied the Outpatient Bleeding Risk Index of Beyth et al5 to patients in our clinic diagnosed as having a pulmonary embolism or deep venous thrombosis, treated with low-molecular-weight heparin therapy followed by prolonged warfarin sodium therapy. Our thrombosis unit serves as a secondary and tertiary care center, with most patients in a 700 000-population region receiving treatment from our unit. Patients were enrolled in treatment studies approved by the local research ethical board, and they signed informed consent forms. Major bleeding in this group of patients was defined as bleeding that led to a loss of 2 U of blood in a 7-day period or bleeding that was otherwise life threatening (ie, the same definition used by Beyth et al). All other hemorrhages were considered minor. Deep venous thrombosis was diagnosed objectively with ultrasonography or phlebography. Pulmonary emboli were diagnosed by a high-probability ventilation-perfusion lung scan with a moderate to high pretest clinical probability, a spiral computed tomographic scan with contrast medium, or pulmonary angiography.

Using the Outpatient Bleeding Risk Index, 1 point was scored for each of the following: (1) 65 years or older, (2) history of GI tract bleeding, (3) history of stroke, and (4) one or more comorbid conditions (recent myocardial infarction, anemia [hematocrit, <30%], renal impairment [creatinine level, >1.5 mg/dL {>133 µmol/L}], or diabetes mellitus). The patient was low risk if the score was 0, moderate risk if the score was 1 or 2, and high risk if the score was 3 or more. Many patients were part of other studies, so exclusions for our study included surgery in the prior 4 days, GI tract bleeding in the previous 14 days, recent stroke, platelet count less than 75 × 103/µL, or active bleeding. Therefore, in general, our patients may have been at lower risk than the patients enrolled in the trial of Beyth et al5; however, all of the risk points used in the index could still be applied to our patients.

The bleeding rates in the 3 groups were compared by using survival curve analysis, with the 3 groups compared by the log-rank test (performed with Statistical Product and Service Solutions for Windows, version 10.0; SPSS Inc, Chicago, Ill).

We observed 222 patients for a mean of 18.5 months (SD, 14.9 months; range, 3-70 months). There were 342 patient-years of observation. The mean age of the patients was 58.4 years; 35.1% of the patients had cancer, 58.1% had idiopathic venous thromboembolism, and 6.8% had venous thromboembolism due to transient risk factors. There were 95 women and 127 men. Table 1 summarizes the frequency of each point in the index according to risk category. The most frequent point by far was 65 years or older. Of the 222 patients, 128 (57.7%) were low risk. This group had 7 minor hemorrhages (5.5%), but no major hemorrhages. Overall, 41.4% (92/222) of the patients were at moderate risk, and 5 (5.4%) (95% confidence interval, 1.8%-12.2%) had a major hemorrhage over a mean follow-up of 15.2 months, for an annualized risk of 4.3% (95% confidence interval, 1.1%-11.1%), or 4.3 events per 100 patient-years of observation. In the moderate-risk group, 74 patients (33.3%) had a score of 1. This group had 4 major hemorrhages (5.4%; 95% confidence interval, 1.5%-13.3%) and 5 minor bleeding events (6.7%) over a mean follow-up of 17 months. There were an additional 18 (8.1%) of the 222 patients at moderate risk, but they had a score of 2. One had a major hemorrhage (5.6%; 95% confidence interval, 1.4%-27.0%) and no minor bleeding events over a mean follow-up of 14 months. Only 2 patients of the 222 followed up were at high risk. Only 1 of the 2 had a minor hemorrhage after a mean ± SD follow-up of 12 ± 6 months. Bleeding events by score are summarized in Table 2. The differences between the low- and moderate-risk groups were statistically significant (P = .03, log-rank test) (Figure 1). The 5 major hemorrhages varied in the time of occurrence; they occurred after 1, 3, 11, 12, and 36 months of treatment. All the major bleeding events were in the GI tract, with 2 in the upper GI tract and 3 in the lower GI tract (Table 3).

Table Graphic Jump LocationTable 1. Frequency of Each Risk Factor According to Risk Category*
Table Graphic Jump LocationTable 2. Number of Bleeding Events per Risk Category
Place holder to copy figure label and caption

Time to major hemorrhage by the Outpatient Bleeding Risk Index score.

Graphic Jump Location

To estimate the risk for major bleeding during anticoagulant therapy, a few prediction rules have been published. Landefeld and Beyth6 reported that the average annual frequencies of fatal, major, and minor bleeding during warfarin therapy were 0.6%, 3.0%, and 9.6%, respectively. These frequencies are approximately 5 times those expected without warfarin therapy. It was suggested that an individual patient's risk for major anticoagulant-related bleeding could be estimated based on specific risk factors, such as the intensity of the anticoagulant effect achieved and the presence of serious comorbid diseases. They also concluded that during the first month of warfarin therapy, the risk for major bleeding is approximately 10 times the risk after the first year of therapy, but this is of little help in patients who require at least 3 months of oral anticoagulation. Beyth et al5 have more recently promoted the Outpatient Bleeding Risk Index. We believed the major bleeding event rates they reported were much higher than those we observed in our patients, so we conducted a prospective evaluation of the index. Major bleeding event rates using the Outpatient Bleeding Risk Index in our patient population were lower than previously published rates, with no major bleeding in the low-risk group and 4.3 events per 100 patient-years of observation in the moderate-risk group. These rates are much lower than those in the validation study published by Beyth et al. It is possible our rates were lower because we have an anticoagulation service that provides care similar to that described recently by Beyth et al.7 In that study, the rates were almost half those previously published. Regardless, we have demonstrated that the index did discriminate between low- and moderate-risk groups (P = .03, log-rank test). Given that the major bleeding event rate was greater than 4% per year of patient observation in this group, the Outpatient Bleeding Risk Index should be a useful tool to use in decisions regarding the optimal duration of anticoagulation therapy. The definition of major bleeding used by Beyth et al is not that used in many trials evaluating the treatment of patients with venous thromboembolic diseases. However, using other definitions did not change the results (data not shown); thus, our results are probably generalizable to centers similar to ours.

Nieuwenhuis et al8 recognized 2 independent prognostic factors for major bleeding during heparin treatment: the World Health Organization performance score and the total body surface area. They found that the scoring system, as described by Landefeld et al,9 did not identify patients with major bleeding in their study, leading them to develop a new scoring system. Nieuwenhuis et al determined World Health Organization performance status, history of a bleeding diathesis, recent trauma or surgery, and body surface area less than 2 m2 to be risk factors. A score of 5 or greater identified 44% of the patients with major bleeding, with a false-positive rate of 9%. Patients with World Health Organization grades 3 and 4 experienced minor hemorrhages more often. Major hemorrhages occurred in 6% of the patients classified as low risk, compared with 17% of the patients at high risk for bleeding (P = .04). We did not apply the scale of Nieuwenhuis et al to all our patients, but in a recent randomized treatment trial,10 the scale was not predictive of bleeding. Van der Meer et al11 have reported that age, type of coumarin used, and achieved intensity of anticoagulation are important predictors of major hemorrhage. We could not evaluate the influence of different coumarin derivatives because in Canada, most patients (all patients in our study) receive warfarin. Also, we have not looked at achieved intensity of anticoagulation because all patients had a target international normalized ratio of 2.0 to 3.0.

Because the major complication of anticoagulant therapy is the risk of bleeding, the Outpatient Bleeding Risk Index could prove useful in assessing the potential risk to patients commencing or currently undergoing anticoagulation therapy. Patient risk could be assessed in conjunction with the potential benefit of preventing recurrent thrombosis. In addition, knowledge of the risk category could indicate those in whom closer monitoring of anticoagulant therapy, including vigilant observation for signs and symptoms of bleeding, may be of benefit. The ability to classify patients with venous thrombosis as at greater risk for a hemorrhage could be beneficial in guiding anticoagulant therapy and could be incorporated in decision aids. Beyth et al7 recently demonstrated lower bleeding rates in patients who were better educated and more closely observed during anticoagulant therapy.

At our institution, patients commencing or considering anticoagulation therapy are assessed for their risk of major hemorrhage using the Outpatient Bleeding Risk Index. The risk is assessed at 4.3% per year for moderate-risk patients and considered to be higher if patients score high risk by the index. The risk is factored into decisions on the optimal duration of anticoagulant therapy. We are evaluating whether the Outpatient Bleeding Risk Index can be used as a preventive measure for patients who have a greater risk for a major hemorrhage by subjecting them to more intense monitoring of the oral anticoagulation therapy. The Outpatient Bleeding Risk Index could prove effective in decreasing the rates of major hemorrhage and death in patients undergoing or commencing anticoagulant therapy.

Corresponding author: Philip S. Wells, MD, Ottawa Hospital Civic Campus, 1053 Carling Ave, Suite F6-47, Ottawa, Ontario, Canada K1Y 4E9 (e-mail: pwells@ohri.ca).

Accepted for publication June 28, 2002.

Kearon  CGent  MHirsh  J  et al.  Extended anticoagulation compared to placebo after three months of therapy for a first episode of idiopathic venous thromboembolism. N Engl J Med. 1999;340901- 907
Forgie  MANichol  GWells  PS Duration of oral anticoagulation in first episode idiopathic deep vein thrombosis: a Markov decision analysis of patient preferences [abstract]. Thromb Haemost. 2001;86OC1728
van den Belt  AGMHutten  BAPrins  MHBossuyt  PMM Duration of oral anticoagulant treatment in patients with venous thromboembolism and deficiency of antithrombin protein C or protein S: a decision analysis. Thromb Haemost. 2000;84758- 763
Landefeld  CSRosenblatt  MWGoldman  L Bleeding in outpatients treated with warfarin: relation to the prothrombin time and important remediable lesions. Am J Med. 1989;87153- 159
Beyth  RJQuinn  LMLandefeld  CS Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med. 1998;10591- 99
Landefeld  CSBeyth  RJ Anticoagulant-related bleeding: clinical epidemiology, prediction, and prevention. Am J Med. 1993;95315- 328
Beyth  RJQuinn  LLandefeld  CS A multicomponent intervention to prevent major bleeding complications in older patients receiving warfarin: a randomized, controlled trial. Ann Intern Med. 2000;133687- 695
Nieuwenhuis  HKAlbada  JBanga  JDSixma  JJ Identification of risk factors for bleeding during treatment of acute venous thromboembolism with heparin or low molecular weight heparin. Blood. 1991;782337- 2343
Landefeld  CSMcGuire  EIRosenblatt  MW A bleeding risk index for estimating the probability of major bleeding in hospitalized patients starting anticoagulant therapy. Am J Med. 1990;89569- 578
Wells  PSRodger  MForgie  MA  et al.  A randomized trial comparing two low molecular weight heparins for the outpatient treatment of deep vein thrombosis or pulmonary embolism [abstract]. Thromb Haemost. 2001;8678
Van der Meer  FJMRosendaal  FRVandenbroucke  JPBriet  E Assessment of a bleeding risk index in two cohorts of patients treated with oral anticoagulants. Thromb Haemost. 1996;7612- 16

Figures

Place holder to copy figure label and caption

Time to major hemorrhage by the Outpatient Bleeding Risk Index score.

Graphic Jump Location

Tables

Table Graphic Jump LocationTable 1. Frequency of Each Risk Factor According to Risk Category*
Table Graphic Jump LocationTable 2. Number of Bleeding Events per Risk Category

References

Kearon  CGent  MHirsh  J  et al.  Extended anticoagulation compared to placebo after three months of therapy for a first episode of idiopathic venous thromboembolism. N Engl J Med. 1999;340901- 907
Forgie  MANichol  GWells  PS Duration of oral anticoagulation in first episode idiopathic deep vein thrombosis: a Markov decision analysis of patient preferences [abstract]. Thromb Haemost. 2001;86OC1728
van den Belt  AGMHutten  BAPrins  MHBossuyt  PMM Duration of oral anticoagulant treatment in patients with venous thromboembolism and deficiency of antithrombin protein C or protein S: a decision analysis. Thromb Haemost. 2000;84758- 763
Landefeld  CSRosenblatt  MWGoldman  L Bleeding in outpatients treated with warfarin: relation to the prothrombin time and important remediable lesions. Am J Med. 1989;87153- 159
Beyth  RJQuinn  LMLandefeld  CS Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med. 1998;10591- 99
Landefeld  CSBeyth  RJ Anticoagulant-related bleeding: clinical epidemiology, prediction, and prevention. Am J Med. 1993;95315- 328
Beyth  RJQuinn  LLandefeld  CS A multicomponent intervention to prevent major bleeding complications in older patients receiving warfarin: a randomized, controlled trial. Ann Intern Med. 2000;133687- 695
Nieuwenhuis  HKAlbada  JBanga  JDSixma  JJ Identification of risk factors for bleeding during treatment of acute venous thromboembolism with heparin or low molecular weight heparin. Blood. 1991;782337- 2343
Landefeld  CSMcGuire  EIRosenblatt  MW A bleeding risk index for estimating the probability of major bleeding in hospitalized patients starting anticoagulant therapy. Am J Med. 1990;89569- 578
Wells  PSRodger  MForgie  MA  et al.  A randomized trial comparing two low molecular weight heparins for the outpatient treatment of deep vein thrombosis or pulmonary embolism [abstract]. Thromb Haemost. 2001;8678
Van der Meer  FJMRosendaal  FRVandenbroucke  JPBriet  E Assessment of a bleeding risk index in two cohorts of patients treated with oral anticoagulants. Thromb Haemost. 1996;7612- 16

Correspondence

CME
Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Submit a Comment

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 60

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Topics
PubMed Articles
JAMAevidence.com

Users' Guides to the Medical Literature
Deep Venous Thrombosis

Users' Guides to the Medical Literature
Scenario 2