Abstract
INTRODUCTION: Current guidelines suggest preoperative direct oral anticoagulant levels of < 30-50 ng/ml. However, there is limited evidence to guide this expert consensus. Reviewing assay titres and clinical outcomes may be able to inform perioperative care of the anticoagulated patient. This review aimed to determine whether DOAC assay plasma concentrations are associated with bleeding or systemic embolic events to better appreciate a possible therapeutic or hazardous reference range. METHODS: Systematic search, performed by an information specialist using a peer-reviewed search. Main search concepts were direct oral anticoagulant therapy for atrial fibrillation or venous thromboembolism. Data synthesised in narrative and tabular format whilst data that could be pooled was subjected to meta-analysis, using a random effects model. Meta regression was conducted for DOAC peak levels and clinical events. PRISMA guidelines were adhered to. RESULTS: Of 6717 retrieved publications, a total of 17 studies were included in the systematic review and 14 in the meta-analysis/regression. Studies report clinical outcome follow up ranging from 28 to 128 weeks. For every 10 ng/ml increase in DOAC assay trough and peak levels, the mean number of bleeding cases increases by 0.03(95 %CI: -0.32 -0.38, P = 0.84) and 0.09(95 %CI: -3.4 -5.3, P = 0.55) respectively, the mean number of major bleed cases increases by 0.01(95 %CI: -0.05 -0.07, P = 0.62) and 0.011(95 %CI: -0.32 -0.34, P = 0.74) respectively and the mean number of systemic embolic event cases decreases by 0.00039(95 %CI: -0.06 -0.0054, P = 0.88) and 0.04(95 %CI: -0.56 -0.48, P = 0.77) respectively. CONCLUSION: There exists no significant, independent relationship, as determined by a univariate meta regression, between DOAC assay concentrations and a patient's risk of bleeding or systemic embolic embolism. This review also highlights the possibility of an absolute, patient specific DOAC assay concentration that may indicate adequate anticoagulation, above which further increases do not confer an increased risk of bleeding. However, further research to characterise this and its utility in the perioperative setting is required.