SEARCH

SEARCH BY CITATION

Keywords:

  • deep vein thrombosis;
  • duration of anticoagulant therapy;
  • pulmonary embolism;
  • venous thromboembolism;
  • vitamin K antagonist

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Abstract. Pinede L, Duhaut P, Cucherat M, Ninet J, Pasquier J, Boissel JP (Hôpital Edouard Herriot and Unité de Pharmacologie Clinique, Lyon, France). J Intern Med 2000; 247: 553–562.

Objective. To assess the length of oral anticoagulant therapy (short versus long duration) after a first episode of venous thromboembolism (VTE).

Design. Meta-analysis of randomized controlled trials, comparing two durations of anticoagulation, identified in 1999 by a computerized search of the Cochrane Controlled Trial Register, Medline and Embase, completed by an extensive review of the references of pertinent articles.

Setting and subjects. The meta-analysis was performed on literature data. Seven published controlled trials were included. Relative risks with 95% confidence intervals were computed using the relative risk logarithm method. Statistical significance was set up at 0.01 for the test of association.

Main outcome measures. Outcomes are major haemorrhage and recurrence after a 12-month follow-up.

Results. For the recurrence end-point (sample size of 2304 patients), a duration treatment of 12–24 weeks seems preferable to a 3–6 week regimen, with a relative risk (RR) of 0.60 (95% CI: 0.45–0.79, P < 0.001). For the major haemorrhage end-point (1823 patients), the RR is not significantly different from 1 (RR = 1.43, 95% CI: 0.51–4.01, P = 0.5). The results were similar for the subgroup ‘permanent risk factors’ or ‘idiopathic VTE’ (RR for recurrence = 0.48, 95% CI: 0.34–0.68, P < 0.001). The tendency was similar, although not reaching statistical significance, for the ‘temporary risk factors’ subgroup (RR for recurrence = 0.34, 95% CI: 0.13–0.93, P = 0.035).

Conclusions. After a first episode of VTE, a long-term treatment regimen allows a significant reduction in the incidence of recurrences without increasing the incidence of bleeding events.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Venous thromboembolism (VTE) is a very common disease: its prevalence has been estimated at 48/100 000, and the prevalence of pulmonary embolism (PE) alone at 23/100 000 [12].

Heparin therapy and the moment at which to introduce oral anticoagulant therapy in the acute phase of VTE disease have been well defined [3–7]. However, the duration of oral anticoagulant therapy is still a controversial question [8–11]. The current consensus recommendations are a 4–12 week treatment for isolated calf deep vein thrombosis (DVT), and a 12–24 week treatment for proximal DVT or for DVT associated with PE. However, only a few randomized controlled trials with enough statistical power have been performed in this field. Therefore, the present state of knowledge is likely to be modified.

The choice of the duration of oral anticoagulant therapy is likely to depend on several factors:

  • 1
    The pre-existence of risk factors for thrombosis: (i) temporary risk factors, such as surgery, trauma, plaster, puerperium or immobilization for medical conditions; (ii) permanent risk factors, such as obesity, varicosity, heart failure, bedridden status, malignancy, congenital or acquired deficiencies or gene mutations of the coagulation factors; (iii) idiopathic VTE is defined by the absence of known risk factors.
  • 2
    The estimation of the incidence of thromboembolic recurrences and haemorrhagic iatrogenic complications. (i) The annual cumulative incidence of recurrences is about 5–8% [1213]. In a prospective study [13], it has been estimated at 4.9% at 3 months, 8.6% at 6 months, 17.5% at 2 years, 24.5% at 5 years and 30.3% at 8 years. Several controlled trials [121415] have shown that the incidence of recurrence increases in the weeks following the cessation of oral anticoagulant therapy and that the incidence is greater for idiopathic VTE or for VTE with permanent risk factors than for VTE with temporary risk factors. (ii) Two prospective studies [1617] and a literature review of randomized trials or cohort studies [18] demonstrated that the annual incidence of haemorrhagic iatrogenic complications varies from 7.6 to 16.5%, of which 0.25–0.64% were fatal haemorrhages, 1.1–2.7% were major haemorrhages and 6.2–13.8% were qualified as minor. This haemorrhagic risk is dependent [17–20] on the intensity or variability of anticoagulation, treatment compliance, occurrence of drug interactions, duration of treatment, age of the patient and medical history or present status.

Recent recommendations [8–10] proposed a short anticoagulant course in patients with temporary risk factors (low risk of recurrence) and a long course of oral anticoagulant therapy in patients with permanent risk factors or idiopathic VTE (high risk of recurrence). In practice [11], it is also essential to balance the desired effect of the anticoagulants in reducing recurrences against the risk of bleeding.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Objective

The objective was to perform a meta-analysis, based on a systematic review of the randomized controlled trials, on the optimal duration of oral anticoagulant therapy after a first episode of DVT of the lower limbs and/or PE.

Data collection

We identified trials through an extensive Embase and Medline search, without any language restriction, using as keywords and text words the following: ‘randomised controlled trials’, ‘clinical trials’, ‘prospective studies’, ‘thrombophlebitis’, ‘deep vein thrombosis’, ‘pulmonary embolism’, ‘thromboembolism’, ‘venous thromboembolism’, ‘warfarin’, ‘antivitamin K or vitamin K antagonists’, ‘oral anticoagulants or anticoagulant’, ‘anticoagulant treatment’, ‘oral anticoagulant therapy’, ‘duration of anticoagulant’, ‘duration of anticoagulant therapy’. We also consulted the Cochrane Controlled Trials Register.

We contacted the principal investigators of each of the published trials in order to get updated data and check the published data.

We reviewed all the references of recent and general reviews, synthesis articles and editorials dealing with the management of VTE and the duration of anticoagulant therapy [3–79–11212227].

Eligibility criteria

Prospective randomized trials comparing two durations of oral anticoagulant therapy after a first episode of DVT and/or PE were eligible.

We found eight published randomized controlled trials [12141523–27]. The trial most recently published [27] cannot be included: the lengths of treatment or follow-up and the methodology are very different from those used in other trials; more particularly, the randomization is performed after 3 months of anticoagulant therapy. Thus, the end-points were not analysed during the same period, especially the 3 months after the initial episode of thrombosis.

The total sample size is equal to 2304 patients and the individual data of each trial are summed up in Table 1.

Table 1.  Randomized controlled trials on the optimal duration of anticoagulant therapy after a first episode of venous thromboembolism
Trial no./yearAuthor [ref.] (location)No. of patientsSite of initial thrombosisReported baseline characteristics of included patientsShort OATG (no. of patients randomized)Long OATG (no. of patients randomized)Follow-upOutcome
  • R, recurrence; maj H, major haemorrhage; OATG, oral anticoagulant treatment groups; TRF, only patients with temporary risk factors; surg., postsurgery; preg., pregnancy; oral C., oral contraceptives; PRF, only patients with permanent risk factors; immob., immobilization; v. insuf., venous insufficiency; SLE, systemic lupus erythematosus; PE, pulmonary embolism; P. DVT, proximal deep vein thrombosis; C. DVT, calf deep vein thrombosis.

  • a

    The data on the 12-month follow-up have been kindly provided by the author.

1/1972O’Sullivan [23] (Australia)18663% P. or C. DVT 37% DVT + PE 50% surg., 6% preg, 10% oral C., 2% cancer 6 weeks (94)24 weeks (92)12 monthsR
2/1985Holmgren et al.[25] (Sweden)13583% P. DVT 17% C. DVT 10% surg., 10% oral C., 9% cancer, 5% immob.4 weeks (69)24 weeks (66)12 monthsR
3/1985Schulman et al.[24] (Sweden)
TRF2013 P. DVT 7 C. DVT Surg. or trauma or oral C. or transient immob.6 weeks (10)12 weeks (10)12 and 24 monthsR
PRF4028 P. DVT 12 C. DVT V. insuf. or cancer or prolonged immob.12 weeks (20)24 weeks (20)12 and 24 monthsR
4/1985Fennerty et al.[26] (UK)100P. or C. DVT ± PEAre excluded: preg. or cancer or prolonged immob.3 weeks (49)6 weeks (51)12 monthsR
5/1992British Thoracic Society [14] (UK)71231% PE alone 50% P. or C. DVT 19% DVT + PE Are excluded: idem trial no. 4 16% surg. 84% medical patients 4 weeks (358)12 weeks (354)12 monthsR/maj H
6/1995Levine et al.[15] (Canada)214P. DVT alone21% cancer 31% orthopaedic surg. and 37% TRF/63% PRF 4 weeks (109)12 weeks (105)12 monthsR/maj H
7/1995Schulman et al.[12] DURAC study (Sweden)89712% PE ± DVT 49% P. DVT 39% C. DVT 62% PRF (cancer, SLE, v. insuf.) or idiopathic DVT 38% TRF: surg., preg., oral C., immob., travel6 weeks (443)24 weeks (454)24 monthsaR/death/maj H
8/1999 Kearon et al.[27] LAFIT trial (Canada) 162 P. DVT isolated (75%) + PE (25%) Idiopathic DVT but 36% had biological abnormalities (factor II or V gene mutation, antiphospholipid)12 weeks (83) 2 years (89)10 months (mean)R/maj H

End-points

As shown in Tables 1 and 2, the end-points were (i) the recurrence of VTE, and (ii) the occurrence of iatrogenic haemorrhagic complications, since these are the two main clinical outcomes likely to influence the duration of oral anticoagulant therapy.

Table 2.  Data used in the meta-analysis. The relative weighting factor is related to the sample size of each study and to the incidence of major haemorrhages and recurrences.
Number of patientsRecurrenceMajor haemorrhage
Trial no./[ref]Long OATShort OATLong OATShort OATLong OATShort OATRelative weighting factor
  1. OAT, oral anticoagulant treatment groups; TRF, patients with temporary risk factors; PRF, patients with permanent risk factors.

1/[23]92949 (9.8%)7 (7.5%)≈ 9%
2/[25]666910 (15.2%)10 (14.5%)≈ 12%
 TRF101001 (10%)≈ 0.5%
 PRF20202 (10%)1 (5%)≈ 2%
4/[26]51496 (11.8%)5 (10.2%)≈ 20%
5/[14]35435814 (4%)28 (7.8%)4 (1.1%)5 (1.4%)≈ 20%
6/[15]1091057 (6.4%)12 (11.4%)1 (0.9%)0≈ 10%
7/[12]45444326 (5.7%)63 (14.2%)5 (1.1%)1 (0.2%)≈ 40%
Total11561148 100%

Data on VTE recurrences were available in all trials (2304 patients), but in three trials [142326], we considered all reported recurrent thromboembolic events, despite the fact that a proportion of these were not systematically verified by objective methods (ultrasonography, venography, angiography, perfusion ventilation lung scan). These differences between the methods of diagnosis do not substantially alter the estimation of the treatment effect. This treatment effect is assessed as the ratio between the frequencies (relative risk) of recurrence of the two groups, where the same procedure is used. Even if the crude frequency of recurrence varies across the studies, due to the difference in the criteria of diagnosis, the ratio can be assumed to be constant.

Data on major haemorrhagic complications have only been reported for three trials [121415] (1823 patients); the missing data were not available directly from the other trials’ authors. A haemorrhage was defined as major when requiring hospitalization or transfusion; when it was intracranial, intraocular or retroperitoneal; or when the haemoglobin level fell by 20 g L–1 or more.

Qualitative analysis

The compared lengths of treatment differed from one study to another (Table 1). The respective trials are variable in terms of the number of included patients, the baseline characteristics of included patients (Table 1), and the eligibility criteria or the validation of end-points; initial and/or recurrence thromboembolic events were not systematically verified by objective methods, except in trials 2, 3, 6 and 7. However, the global methodology, the patient population, the treatment regimens and the length of follow-up were similar in the different trials.

The randomization procedure is not given in the papers for trials 1 and 5; sealed envelopes were used in trials 2 and 3, a random number table in trial 4, and computer-generated randomized arrangements in trials 6 and 7.

In all trials but one [15], recurrences were diagnosed by the investigators, who were not blinded to the treatment duration. However, in three of the seven trials [121415], recurrences were assessed afterwards by an independent committee of experts, who were unaware of the duration of treatment received.

Information about loss of follow-up was given in three of the seven studies, amounting to 8% of the included patients in the first study [14], 5% in the second study [12] and 6% in the third [15].

In all trials, the patients were treated with warfarin. The information on the treatment intensity (international normalized ratio [INR] range or prothrombin time ratio value) and its comparison between the different treatment groups are not given in the paper for trial 2. For trials 1 and 4, the authors report that the anticoagulation ‘in the majority of cases, was achieved with very little difficulty’ or that ‘no difference is observed in the degree of anticoagulation between treatment groups’. For the other trials (3, 5, 6 and 7), oral anticoagulation was efficient in 59–68% of patients and is similar amongst those with short or long duration of treatment. The target INR (or prothrombin times for old studies) is similar for all trials (between 2 and 3 for INR).

Quantitative analysis

In this meta-analysis, we compared the short arm (the group with the shortest duration of oral anticoagulant therapy) with the long arm (the group with the longest duration of treatment). The short arm was arbitrarily considered as the control group. The homogeneity of the relative risks was checked amongst the trials.

Outcomes were assessed after a 12-month follow-up in six trials; in one trial (no. 7), data on the 12-month follow-up were not available in the published article [12]. They were kindly provided by the principal investigator.

The meta-analysis was performed on a sample of 2304 patients for the ‘recurrence’ end-point, and one of 1823 patients for the ‘haemorrhage’ end-point. Several statistical methods have been used (Peto, Mantel Haenszel, Cochran, odds ratio logarithm, relative risk logarithm, random relative risk, percentage differences, method of Dersimonian and Laird), based on both fixed and random models, and using odds ratio, relative risk or risk difference. In the absence of significant heterogeneity with the fixed model, the random model was discarded. Amongst the fixed model-based methods, we retained the relative risk logarithm method, which was the most conservative.

The meta-analysis calculations were performed using the ‘Easy MA’ software [28]. To avoid calculation problems with the zero cells, a constant of 0.25 was added to all cells in the 2 × 2 tables.

The level of significance for the chi-square test for association was set at 0.01. A homogeneity test was performed, and the level of significance for the chi-square test for homogeneity was set at 0.1.

Subgroup analysis

Risk factors for thrombosis were studied, taking into consideration the four trials in which they were detailed (Table 1: trials 3, 5, 6 and 7); temporary or permanent risk factors and idiopathic VTE were considered, as defined above. The meta-analysis of ‘permanent’ risk factors for VTE or idiopathic VTE (Table 3) included 1323 patients, and the analysis of ‘temporary’ risk factors for VTE (mainly recent surgery) included 560 patients (Table 4). The same statistical methods as described above were applied. In these subgroups, only the recurrence end-point was considered; the data did not permit us to use the haemorrhage end-point. Unfortunately, we cannot, with the published data, distinguish proximal and calf DVTs in our meta-analysis.

Table 3.  Data used in the meta-analysis for patients (n = 1323) with permanent risk factors or idiopathic VTE – only the recurrence end-point was considered
Number of patientsRecurrence
Trial no./[ref]Long OATShort OATLong OATShort OATRelative weighting factor
  1. OAT, oral anticoagulant treatment groups; PRF, patients with permanent risk factors.

3/[24] (PRF)20202 (10%)1 (5%)≈ 2%
5/[14]29829814 (4.7%)27 (9.1%)≈ 30%
6/[15]66687 (10.6%)10 (14.7%)≈ 14.5%
7/[12]28726622 (7.7%)52 (19.6%)≈ 53.5%
Total671652 100%
Table 4.  Data used in the meta-analysis for patients (n = 526) with temporary risk factors – only the recurrence end-point was considered. The relative weighting factor is related to the sample size of each study and to the incidence of recurrences
Number of patientsRecurrence
Trial no./[ref]Long OATShort OATLong OATShort OATRelative weighting factor
  1. OAT, oral anticoagulant treatment groups; TRF, patients with temporary risk factors.

3/[24] (TRF)101001 (10%)≈ 5%
5/[14]566001 (1.7%)≈ 6%
6/[15]433702 (5.4%)≈ 6%
7/[12]1671774 (2.4%)11 (6.2%)≈ 83%
Total276284 100%

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The 12-month incidence weighted mean of thromboembolic recurrences in the seven trials is equal to 6.4% in the long treatment groups, and 11.1% in the short treatment groups. When the patients with permanent risk factors or idiopathic thrombosis are considered alone, this incidence rises to 6.7% in the long treatment groups and 13.8% in the short. For patients with temporary risk factors, the incidences of recurrence are smaller, being 1.4 and 5.3% in the long and short treatment groups, respectively. The incidence weighted mean of major bleeding is 1.1% in the long treatment regimen, and 0.7% in the short one.

Global meta-analysis

For the ‘major bleeding’ end-point, there is no statistically significant difference between the two regimens of oral anticoagulant therapy, whatever the statistical method used. Using the logarithm of the relative risk method (Fig. 1), the relative risk (RR) is equal to 1.43 with a 95% confidence interval (95% CI) lying between 0.51 and 4.01; the test of association is not significant (P = 0.5). There is no significant heterogeneity between the two treatment regimens, and the studies can be considered as homogeneous (P = 0.34).

image

Figure 1. Global results of the meta-analysis: long-term versus short-term regimen; end-point = major haemorrhage (n = 1823).

Download figure to PowerPoint

When the ‘recurrence’ end-point is considered, the long-term treatment regimen is superior to the short-term one, whatever the statistical method used (the results are presented with the logarithm of the RR method). For this end-point, we first performed the meta-analysis restricted to the three most recent and adequately designed trials (trials 5, 6 and 7; Table 1), and then performed it again with all seven studies:

Three most recent trials – the RR is 0.45 (95% CI: 0.32–0.63, P < 0.001). There is no substantial heterogeneity (P = 0.73).

The seven trials (Fig. 2) – the RR is 0.60 (95% CI: 0.45–0.79). The test of association is also significant (P < 0.001), while the test for heterogeneity is not (P = 0.16).

image

Figure 2. Global results of the meta-analysis: long-term regimen versus short-term regimen; end-point = recurrence (n = 2304).

Download figure to PowerPoint

We conclude that adding the results of the four earlier studies does not change the conclusions of our analysis significantly.

Subgroup meta-analysis (Fig. 3)

image

Figure 3. Results of the meta-analysis for subgroups: long-term regimen versus short-term regimen; end-point = recurrence. Patients with permanent risk factors = 1323; patients with temporary risk factors = 560.

Download figure to PowerPoint

Only the recurrence end-point was considered.

For patients with a permanent risk factor or with idiopathic VTE, the long-term regimen of anticoagulant therapy is superior to the short-term one: the RR is 0.48 (95% CI: 0.34–0.68) and the test of association is significant (P < 0.001). There is no significant heterogeneity between the two treatment regimens (P = 0.39).

For patients with a temporary risk factor, there is no statistically significant difference between the long-term regimen of oral anticoagulant therapy and the short-term regimen, using the strict end-points defined above: RR = 0.34 (95% CI: 0.13–0.93), P = 0.035 (> 0.01). The test of heterogeneity is not significant (P = 0.91). However, there is no statistically significant heterogeneity between the two treatment regimens, suggesting that the lack of significance for the association test in the temporary risk factor subgroup could be related to lack of power.

Patients from trial 7 [12] account for one-third of the patients included in the global meta-analysis and for about one-half of the patients included in the subgroup analysis. The influence of this trial in the results we report has to be acknowledged (Table 24).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The optimal duration of oral anticoagulant therapy is that which would allow a significant reduction in the incidence of recurrences without increasing the incidence of iatrogenic haemorrhagic complications.

After a first episode of VTE, data from only eight controlled trials are available [12141523–27]; moreover, these published data have limitations. The lengths of treatment compared were different from trial to trial. The power of some of them [23–26] is weak, and the single end-point analysed is the ‘recurrence’ of thromboembolic events. The validation of either DVT or PE, at diagnosis and during follow-up, has not always been assessed by objective procedures [142326] (very often, the diagnosis of VTE has been made on clinical criteria only the previous years [26]). The results of the first four trials [23–26] argued in favour of the short treatment regimen since there was no significant difference between the short and long treatments.

Thus, in 1988, a meta-analysis from available data [29] showed no difference between short courses (3, 4 or 6 weeks) and longer courses (6–24 weeks) of anticoagulant therapy; however, the 95% CI of the difference was very wide.

Four other trials [12141527], more recently published, are methodologically stronger and two of them [1214] have included a large number of patients. Three of these studies [121415] concluded that: (i) after a proximal DVT or PE, the duration of oral anticoagulant therapy should be at least 3 months (12 weeks), and perhaps 6 months (24 weeks); (ii) a separate analysis was performed for subgroups of patients, and those with temporary risk factors should be treated with a short-duration regimen, and those with permanent risk factors or idiopathic thrombosis with a longer duration regimen. A recent double-blind randomized study [27] has compared 3 vs. 27 months of anticoagulant therapy (3 months of warfarin then placebo vs. 27 months of warfarin) in 162 patients with a first episode of idiopathic VTE (LAFIT trial). The authors confirm that the patients with ‘idiopathic’ VTE should be treated for longer than 3 months. However, for methodological reasons already detailed, the data of this trial cannot be included in a meta-analysis with the other trials. Nevertheless, we have included in the analysis the seven other trials. In fact, when repeating the global meta-analysis for only those studies in which the methodology is rigorous (n = 3), the results and the point estimates were almost identical.

Our meta-analysis confirms that a long course of oral anticoagulant therapy (12 or 24 weeks) is superior to a short course (3, 4 or 6 weeks) for the whole series of patients and in the subgroup of patients with permanent risk factors and/or idiopathic VTE, when ‘12-month recurrence’ is considered as the end-point. However, for patients with temporary risk factors, the difference seems to be less significant (P = 0.035), according to the strict criteria we requested (P < 0.01). For these patients, recurrences are rare events, the sample size is small and our analysis might not have been powerful enough.

Major bleeding is a rare, but severe, event and our meta-analysis did not show a significant difference between the two treatment regimens. However, major haemorrhage might be more frequent in the long-term treatment groups, and the sample size (1823 patients) would not allow us to show this difference to be significant. If the observed RR of 1.43 really describes the reality, a sample size of 16 200 patients in each arm would be needed to show a significant difference for the haemorrhage end-point (alpha error being set at 0.05, and beta at 0.20) [30]. Nevertheless, the risk of bleeding must not be neglected in clinical practice, especially in patients receiving prolonged oral anticoagulant therapy.

After a first episode of idiopathic VTE, or of VTE with permanent risk factors, the optimal duration of treatment seems to be that of the long-term regimen, and could be 12 or 24 weeks. Whether 12 weeks (3 months) or 24 weeks (6 months) represents the optimal duration remains uncertain. The results of the DURAC study [12] and of our meta-analysis favour a 6-month duration. To confirm this fact, we are conducting a multicentre, controlled randomized trial in France (the DOTAVK study), which compares a 3-month with a 6-month regimen for patients with proximal DVT and/or PE (and a 6-week with a 12-week regimen for patients with isolated calf DVT) [31].

For patients with temporary risk factors, we still have no answer to the following question: is a shorter treatment preferable, as has been suggested by recently published trials, or is it necessary to maintain a long-term treatment (12 or 24 weeks) as has been suggested by our analysis? Further studies are needed in order to reach sufficient statistical power.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

In the specific medical field of oral anticoagulant therapy after a first episode of VTE, our meta-analysis suggests that prolonged therapies are to be preferred. However, the optimal duration still has not been determined and this is not possible with the available data. A large randomized controlled trial is needed to answer this question, as well as a meta-analysis on individual data, which would draw more precise conclusions from these data. These studies will be of particular importance to allow a clear identification of the subgroups of patients who could benefit from a long course of oral anticoagulant therapy (high risk of recurrence) or a short anticoagulant course (high risk of haemorrhage and low risk of recurrence).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We are indebted to Mrs Alison Foote for revising the manuscript, to Mrs Marie Laure Boncors, Sylvie Delair and Magali Hervé for their technical assistance.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • 1
    Anderson FA, Wheeler HB, Goldberg RJ et al. A population based perspective of the hospital incidence and case fatality rates of deep vein thrombosis and pulmonary embolism. Arch Intern Med 1991; 151: 93338.
  • 2
    Lindblad B, Sternby NH, Bergqvist D. Incidence of venous thromboembolism verified by necropsy over 30 years. Br Med J 1991; 302: 70911.
  • 3
    Hirsh J, Salzman EW, Marder Vj. Treatment of venous thromboembolism. In: ColmanRW, HirshJ, MarderVJ, SalzmanEW, eds. Hemostasis and Thrombosis Basic Principles and Clinical Practice, 3rd edn. Philadelphia: J.B.Lippincott, 1994.
  • 4
    Hyers TM, Hull RD, Weg Gj. Antithrombotic therapy for venous thromboembolic disease. Chest 1998; 114: 56178S.
  • 5
    Ten Cate JW, Koopman MM, Prins MH, Buller Hr. Treatment of venous thromboembolism Thromb Haemost 1995; 74: 197203.
  • 6
    Hirsh J, Hoak J. Management of deep vein thrombosis and pulmonary embolism. A statement for healthcare professionals. Circulation 1996; 93: 221245.
  • 7
    Ginsberg Js. Management of venous thromboembolism. New Engl J Med 1996; 335: 181628.
  • 8
    Hirsh J. The optimal duration of anticoagulant therapy for venous thrombosis. New Engl J Med 1995; 332: 171011.
  • 9
    Diuguid Dl. Oral anticoagulant therapy for venous thromboembolism. New Engl J Med 1997; 336: 43334.
  • 10
    Hirsh J, Kearon C, Ginsberg J. Duration of anticoagulant therapy after first episode of venous thrombosis in patients with inherited thrombophilia. Arch Int Med 1997; 157: 217476.
  • 11
    Bounameaux H, De Moerloose P, Sarasin Fp. Optimal duration of oral anticoagulant therapy following deep vein thrombosis of lower limbs. Blood Coag Fibrinol 1996; 7: 50714.
  • 12
    Schulman S, Rhedin AS, Lindmarker P, Carlsson A, Larfars G, Nicol P and the Duration of Anticoagulation Trial Study Group. A comparison of six weeks with six months of oral anticoagulant therapy after a first episode of venous thromboembolism. New Engl J Med 1995; 332: 166165.
  • 13
    Prandoni P, Lensing AW, Cogo A et al. The long term clinical course of acute deep venous thrombosis. Ann Intern Med 1996; 125: 17.
  • 14
    Research Committee of the British Thoracic Society (Campbell IA). Optimum duration of anticoagulation for deep vein thrombosis and pulmonary embolism. Lancet 1992; 340: 87376.
  • 15
    Levine MN, Hirsh J, Gent M et al. Optimal duration of oral anticoagulant therapy: a randomized trial comparing four weeks with three months of warfarin in patients with proximal deep vein thrombosis. Thromb Haemost 1995; 74: 60611.
  • 16
    Palareti G, Leali N, Coccheri S, et al. on behalf of the Italian Study on Complications of Oral Anticoagulant Therapy. Bleeding complications of oral anticoagulant treatment: an inception cohort prospective collaboration study (ISCOAT). Lancet 1996; 348: 42328.
  • 17
    Van Der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briet E. Bleeding complications in oral anticoagulant therapy. An analysis of risk factors. Arch Intern Med 1993; 153: 155762.
  • 18
    Landefeld CS, Beyth Rj. Anticoagulant related bleeding: clinical epidemiology, prediction, and prevention. Am J Med 1993; 95: 31528.
  • 19
    Fihn SD, Mc Donell M, Martin D et al. Risk factors for complications of chronic anticoagulation. A multicenter study. Ann Intern Med 1993; 118: 51120.
  • 20
    Gitter MJ, Jaeger TM, Petterson TM, Gersh BJ, Phil D, Silverstein Md. Bleeding and thromboembolism during anticoagulant therapy. A population based study in Rochester Minnesota. Mayo Clin Proc 1995; 70: 72533.
  • 21
    Pineo GF, Hull Rd. Prevention and treatment of venous thromboembolism. Drugs 1996; 52: 7192.
  • 22
    Kearon C. Drug trials that have influenced our practice in the treatment of venous thromboembolism. Thromb Haemost 1997; 78: 55357.
  • 23
    O'sullivan Ef. Duration of anticoagulant therapy in venous thromboembolism. Med J Aust 1972; 2: 1104107.
  • 24
    Schulman S, Lockner D, Juhlin Dannfelt A. The duration of oral anticoagulation after deep vein thrombosis: a randomized study. Acta Med Scand 1985; 217: 54752.
  • 25
    Holmgren KA, Andersson G, Fagrell B. One-month versus six months therapy with oral anticoagulants after symptomatic deep vein thrombosis. Acta Med Scand 1985; 218: 27984.
  • 26
    Fennerty AG, Dolben J, Thomas P et al. A comparison of 3 and 6 weeks anticoagulation in the treatment of venous thromboembolism. Clin Lab Haematol 1987; 9: 1721.
  • 27
    Kearon C, Gent M, Hirsh J et al. A comparison of three months of anticoagulation with extended anticoagulation for a first episode of idiopathic venous thromboembolism. New Engl J Med 1999; 340: 901907.
  • 28
    Cucherat M, Boissel JP, Leizorovicz A, Haugh Mc. Easy MA: a program for the meta-analysis of clinical trials. Computer Methods Prog Biomed 1997; 53: 18790.
  • 29
    Fennerty AG, Campbell IA, Routledge Pa. Anticoagulants in venous thromboembolism. Guidelines for optimum treatment. Br Med J 1988; 297: 128588.
  • 30
    Fleiss Jl. Determining sample sizes needed to detect a difference between two proportions.In: Statistical Methods for Rates and Proportions, 2nd edn. New York: Wiley, 1981; 3349
  • 31
    Pinède L, Ninet J, Boissel JP, Pasquier J. Traitement anticoagulant pour maladie thrombo-embolique veineuse: durée optimale de prescription des Antivitamines-K. Revue de la littérature. Presse Med 1994; 23: 181720.