2.1. Literature searchy
2.2. Article selection
2.3. Critical appraisal and data extraction (see appendix 4 page 129)
2.4. Conclusions and levels of evidence
2.5.1. Economic considerations
2.5.2. General considerations:
2.6. Peer-review process
This clinical practice guideline methodology is based on Clinical Practice Guidelines (CPG)
method with the following steps:
- – literature exhaustive review and selection;
- – critical appraisal;
- – data extraction;
- – conclusions and recommendations writing;
- · peer-review process.
The different steps of the methodology are shown in appendix 5 page 133. This method is fully
detailed below and in appendix 6 page 135.
A literature search for all studies published between January 1996 and January 2011 was
performed using the Medline® database and the following subject headings: cancer, venous thromboembolism, and anticoagulant drugs. A prospective follow-up of the literature was continued up to June 2011. Members of the working group also added other references: those not found by the bibliographic search. In addition, when available, data previously extracted by the working group members from former guidelines or meta-analyses [LYMAN20007] [FARGE2008] [NOBLE2008] [GEERST2008] [DEBOURDEAU2009] were added into the analysis, even if the original paper had been published before 1996. National guidelines and several sites of Evidence-Based Medicine were also consulted (see list of URL link visited in appendix 3 page 128).
The literature search was limited to publications in English or in French.
Meta-analyses, systematic reviews, randomized clinical trials, or non-randomized prospective or retrospective studies in the absence of randomized clinical trials, were included in the analysis. Editorials, letters to the editor, case reports, publications without an abstract, press releases and animal studies were excluded. Abstracts were included only if a full paper was accepted in a peer-reviewed medical journal.
In the absence of specific studies on patients with cancer, we also analyzed studies performed in the general population of VTE patients, which also included patients with cancer. In this case, results were extrapolated to cancer patients and methodological biases were taken into account.
The terms used in the literature search for the treatment of VTE were: treatment, anticoagulation, low-molecular-weight heparin, unfractionated heparin, and therefore included new oral anticoagulants, The bibliographic search strategy is shown in appendix 2 on page 122.
Major bleeding was defined as [SCHULMAN2005] [SCHULMAN2010]:
- · fatal bleeding,
- · bleeding into a critical organ (brain, intra-abdominal bleeding, gastrointestinal tract)
- · associated with a decrease in hemoglobin level of more than 2 g/dL
- · leading to the transfusion of two or more units of blood.
Minor bleeding was defined as all other bleeds.
The quality of the studies was evaluated in a double-blind manner by 2 reviewers (PD, MB) by means of systematic completion and combined analysis and review of:
- · methodological appraisal grids,
- · clinical relevance grids.
Data were then extracted and constructed in evidence tables (see Tables 3 – 38), which were subsequently validated by the whole working group members.
First, for each question, results of the literature analysis were summarized and discussed by the whole working group taking into account critical appraisal and data extraction grids.
Second, for each clinical question, conclusions were formulated on the basis of:
- · the corresponding pooled questions results and conclusions,
- · the degree of agreement between the studies.
The “Grading of Recommendations Assessment Development and Evaluation” (GRADE) scale [GUYATT2008] [GUYATT2008A] allowed distinction of 4 levels of evidence for each conclusion:
- · High: Further research is very unlikely to change our confidence in the estimate of effect.
- · Moderate: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
- · Low: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate
- · Very low: Any estimate of effect is very uncertain.
These levels of evidence are attributed to each conclusion according to (see Table 2):
- · the study design
- · the study limitations, inconsistency, indirectness, imprecision and publication bias.
Following the conclusions for each clinical question, recommendations were established taking into account the following parameters:
- · Conclusion level of evidence (high, moderate, low, very low): the higher the quality of the evidence, the higher the likelihood that a strong recommendation is warranted;
- · Balance between desirable and undesirable effects: the larger the difference between the desirable and undesirable effects, the higher the likelihood that a strong recommendation is warranted. The smaller the net benefit and the lower certainty for that benefit, the more likely that a weak recommendation is warranted;
- · Values and preferences: the more values and preferences vary, or the greater the uncertainty in values and preferences, the higher the likelihood that a weak recommendation is warranted;
- · Costs (resource allocation): the higher the costs of an intervention (i.e., the greater the resources consumed), the lower the likelihood that a strong recommendation is warranted.
- GRADE scale [GUYATT2008] [GUYATT2008 A] allows distinction of 2 levels for each recommendation, corresponding to “the degree of confidence that the desirable effects of adherence to a recommendation outweighs the undesirable effects.”
- · Strong: the panel is confident that the desirable effects of adherence to a recommendation outweigh the undesirable effects.
- · Weak: the panel concludes that the desirable effects of adherence to a recommendation probably outweigh the undesirable effects, but is not confident.
For previously selected questions without any clear scientific evidence or any data available after an extensive search and analysis of the literature, the working group decided to add a third level of recommendation, defined as “Best Clinical Practices”. Elaboration of the Best Clinical Practices recommendation had to be based on the consensus of the international working group experts after discussion, taking into account the professional experience and all experts’ opinion.
The following economic and general considerations were taken into account to elaborate the recommendations.
- · The price of a drug varies different countries and in different regions of the world.
- · In the case of a strong recommendation, the benefit to the patient outweighs health economics considerations.
- · In the case of a weak recommendation, health economics aspects were considered.
- · Costs of anticoagulants are negligible compared with the cost of cancer treatment.
For each patient and before applying a recommendation, the working group advocates a careful evaluation of:
- · Contraindications to anticoagulation,
- · Creatinine clearance,
- · Risk of hepatic dysfunction ,
- · Bleeding risk,
- · Benefit-risk ratio of each treatment,
- · Patient preferences.
These guidelines were peer-reviewed in February 2012 by 43 independent experts in different medical specialties involved in VTE management (internal medicine = 5, vascular medicine =13, hematology = 10, oncology = 7, anticoagulation clinical nurse specialist = 2, pharmacist = 1;pneumology = 1, anesthesiology = 1, palliative medicine = 1), working in public and private institutions.
Reviewers were sollicited among “International Scientific Societies”, “National experts”, “International patients’ associations”, etc.
A grid was proposed allowing partial and general appreciation of the document via a quantitative (quotation) and a qualitative evaluation (comments). All the comments were examined and discussed by the members of the working group and were then integrated in the final version of the document in April 2012. The members who took part in this peer-review process are named in the document as reviewers (see appendix 1, page 121).
Any discrepancies in opinion between reviewers and members of the working group were resolved by consensus during the last meeting.