Authors' recommendations: • International practice guidelines recommend prescribing anticoagulants to patients with non-valvular atrial fibrillation and a CHA2DS2-VASc score ≥ 2 (for men) and score ≥ 3 (for women). For a CHA2DS2-VASc score = 0 they are best not prescribed. In patients with a CHA2DS2-VASc = 1 (men) and = 2 (women), the risk of stroke due to atrial fibrillation (AF) is of the same order of magnitude as the risk of cerebral haemorrhage induced by anticoagulants. Hence, it is not certain whether the latter patients all benefit from treatment with anticoagulants. Therefore we conclude that these medications are not recommended for them. • Based on randomised clinical trials (RCTs) in which NOACs are compared with VKAs, we can conclude that these two classes of medications are equivalent in preventing ischemic strokes in patients with atrial fibrillation (AF). • In the area of adverse effects, the NOACs score better than VKAs in terms of tenths of a percent: the risk of cerebral bleeding is slightly (but statistically significantly) lower (-0.20 to -0.31% per year). However, NOACs appear to cause a higher number of gastrointestinal bleedings (+0.51 to -0.10% per year). It should be noted that these RCTs may favour the NOACs due to bias. • With NOACs there are clearly fewer laboratory checks needed than with VKAs. • The long-term effects of NOACs are not yet known, although these medications must be taken by some patients for 10 or 20 years or more. • From the Belgian economic models it appears that NOACs extend the life of patients by 1 to 4 quality adjusted life months if their effects are extrapolated over an entire lifespan. The accumulated cost over this period are only slightly higher than those of the VKAs. For that reason these models consider the NOACs cost-effective compared to the VKAs. This holds only on the assumption that the anticoagulants are used in daily practice as they are in the RCTs, and that the results of the RCTs remain applicable in the long term. • In contrast to what was expected, the persistence of patients on a NOAC is no better than with a VKA, although NOACs do not require regular blood testing. According to the figures of the IMA/AIM, 20 to 30% on anticoagulation treatment interrupt their treatment; these figures hold for both classes of medications. • Due to the short duration of action of the NOACs, compliance is even more important with them than with the VKAs. A patient who forgets to take a NOAC only once already runs a higher risk of stroke, which is not the case for VKAs. • In daily practice, a substantial number (43%) of Belgian patients have a lower dose prescribed than that in the RCTs. Moreover, even in the strict framework of the RE-LY study (dabigatran) it turned out that 20% of the patients who took the dose assigned to them fell outside the optimal therapeutic serum levels. That is a problem, because for NOACs it is impossible to check whether the patient is getting/taking the adequate dose with coagulation tests. It cannot be excluded that some patients on a NOAC are no better protected than on a VKA. • The NOACs are a good therapeutic choice for patients in whom a stable INR is difficult to achieve on a VKA, or for whom regular blood tests are problematic. The prerequisite is that these patients have the proper NOAC dose prescribed and show adequate persistence/compliance.

Project Status: Completed

Year Published: 2017

URL for published report: https://doi.org/10.57598/R279C

English language abstract: An English language summary is available

Publication Type: Not Assigned

Country: Belgium

DOI: 10.57598/R279C

Organisation Name: Belgian Health Care Knowledge Centre

Contact Address: Administrative Centre Botanique, Doorbuilding (10th floor), Boulevard du Jardin Botanique 55, B-1000 Brussels, Belgium tel: +32 2 287 33 88 fax: +32 2 287 33 85

Contact Name: info@kce.fgov.be

Contact Email: info@kce.fgov.be

Copyright: Belgian Health Care Knowledge Centre (KCE)