Influence of Direct Oral Anticoagulant Levels and Thrombin Generation on Postoperative Bleeding [SONAR]: A Nested Case-Control Study.

A direct oral anticoagulant (DOAC) concentration threshold above which an impact on surgical hemostasis starts to occur is unknown. Thrombin generation assays (TGAs) provide a measure of the coagulation phenotype. This study aimed to determine whether preoperative TGA parameters are associated with postoperative bleeding, and whether this is partly due to residual DOAC levels.We conducted a nested case-control study using samples from apixaban/rivaroxaban-treated patients with atrial fibrillation from the PAUSE (Perioperative Anticoagulation Use for Surgery Evaluation) perioperative study. Cases were participants with postoperative major or clinically relevant nonmajor bleeding; controls were participants without bleeding. DOAC levels were measured using a chromogenic anti-Xa assay (BIOPHEN DiXaI; rivaroxaban/apixaban calibrators). TGA parameters were measured using calibrated automated thrombography.Generalized linear mixed models and causal mediation analyses were used to evaluate the relationship between DOAC levels, TGA parameters, and bleeding.Forty eight cases were matched to 474 controls. Residual DOAC levels were higher in cases than controls ( = 0.03) and each TGA parameter was correlated with residual DOAC levels (<0.05). A longer lag time (LT; odds ratio [OR] = 1.319 per minute [95% confidence interval [CI]: 1.077-1.617]) and time-to-peak (TTP; OR = 1.154 per minute [95% CI: 1.028-1.296]) were associated with an increased odds of bleeding; higher peak (OR = 0.994 per nM [95% CI: 0.989-0.998]) and mean velocity rate index (mVRI; OR = 0.986 per nM/min [95% CI: 0.976-0.996]) were associated with a lower odds of bleeding. The effect of apixaban/rivaroxaban levels on bleeding was mediated by altered TGA parameters (LT, TTP, peak, mVRI).These findings support a measurable effect from low residual DOAC levels on thrombin generation and suggest a causal contribution of both toward bleeding.