Impact of antiplatelet and anticoagulant drugs on postoperative bleeding in reverse total shoulder arthroplasty

Background: The decision to withdraw or continue antiplatelet and anticoagulant drugs would be balanced on the bleeding risk and the cardiovascular risk. The purpose of this study was to investigate perioperative bleeding in reverse shoulder arthroplasty (RSA) to determine the impact of oral antiplatelet and anticoagulant medications in the absence of drug withdrawal. The hypothesis was that the continuation of antiplatelet and anticoagulant drugs would increase postoperative bleeding but to a limited extent.

Methods: During the study period, RSA cases were prospectively included and retrospectively reviewed. Cases of revision for arthroplasty, proximal fracture, and cemented stem fixation were excluded. Cases with dual therapy of aspirin and clopidogrel and cases with heparin bridges were also excluded. Age, gender, height, weight, American Society of Anesthesiologists physical status anesthesia risk (ASA), smoking status, preoperative diagnosis, preoperative blood sampling data within one month of surgery (estimated glomerular filtration rate (eGFR), Prothrombin Time-International Normalized Ratio (PT-INR), and activated partial thromboplastin time (APTT)), surgical time, stem length (standard or short stem), and use of tranexamic acid were identified. The presence of diabetes mellitus, rheumatoid arthritis, and hypertension were confirmed as comorbidities. Hemoglobin (Hb) and hematocrit (Ht) values were recorded preoperatively and on the third postoperative day, and estimated blood loss was calculated by Gross formula using the change in preoperative Ht and postoperative day 3 Ht values. Multiple regression models were used to determine predictors of the estimated blood loss. P values less than .05 were considered statistically significant.

Results: 315 RSA cases were analyzed (172 females, a mean age of 77.2 years). The estimated blood loss was 819 ml. The regression equation was statistically significant (P < .0001, R²=0.21), and aspirin, direct oral anticoagulants (DOAC), tranexamic acid, height, and surgery time were statistically significant explanatory variables, showing the predicted changes in bleeding volume for each drug were +283 ml for aspirin, +180 ml for DOAC, and -237 ml for tranexamic acid, respectively. Warfarin and cilostazol were not employed due to P values and Akaike's Information Criterion. Two blood transfusions (without antiplatelet or anticoagulant medications) were administered. One patient on aspirin had a cardiovascular event.

Conclusions: The current study showed that the use of aspirin and DOAC were significant predictors of increased perioperative blood loss, but the amount of increase was in the acceptable range, suggesting that continued antiplatelet and anticoagulant were relatively safe in primary RSA. Administration of tranexamic acid was shown to reduce bleeding significantly.