Clinical Trial Analyzing the Efficacy of Oral Administration of Tranexamic Acid in Spine Surgery.

The study design is a prospective randomized control trial done in a single institution from 2021 to 2022. Approval from the institutional ethics committee was obtained for the study (number 04/21). Major surgery was defined as "surgeries involving the intervention of more than 4 levels of the thoracolumbar spine, and with a diagnosis of failed back surgery or, deformities in the spine". Patients excluded were those with low preoperative hemogram values (Hb <10mg/dL, low platelet counts (<100 x 109/L), bleeding disorders, coagulopathies, intake of contraceptives or anticoagulant medication, active thromboembolic disease (deep venous thromboembolism, chronic venous insufficiency, chronic thromboembolism), fibrinolytic disorders secondary to consumption coagulopathy, history of thromboembolic or coronary disease, history of seizure, liver failure, dyslipidemia, congenital coagulopathies (Von Willebrand Disease, Hemophilia A and B), acquired coagulopathy (Vitamin K deficiency, disseminated intravascular coagulation) and, thrombocytopenic purpura. Patients in whom postsurgical hemoglobin was not available, or patients who accidentally removed the drain or had leaks were eliminated from the study.

General anesthesia was given for all the procedures and surgery was performed by two seniors' surgeons with over 27 years of experience in spine surgery, standard posterior approach was utilized in all cases, and all the cases were treated with pedicle screw fixation, at least one osteotomy and posterolateral or interbody fusion were performed in all the cases. All cases have been randomly divided into 2 groups.

Interventions:

1. Control group (n=30 cases): 100ml of water, and 3 tablets of placebo pills were administered 2 hours prior to surgery.
2. PO TXA group (n=30 cases): Single dose of 1950 mg (3 tablets of 650 mg) was administered orally 2 hours before the surgical incision. The dose was administered by the nurse on duty who provided 100ml of water to swallow the pills.

For both groups estimation of intraoperative bleeding was calculated by the anesthesiologist at the end of the surgery, anesthesiologist and surgeons in charge were blinded to the patient group, the need for transfusion during surgery and the amount of volume to be transfused was determined by the anesthesiologist. A suction drain was placed deep into the fascia before closure; the amount of total drainage at 36 hours was recorded. The Suction drain was constantly kept in suction mode, and it was removed when the output was less than 100ml/24-hour period after postoperative day 1.

Post-operatory transfusion (within the first 48 hours after surgery) was carried out in patients with hemoglobin <8mg/dL and for patients with clinical symptoms of anemia such as hypotension and tachycardia.

Clinical data include age, height, weight, BMI, operative time, hemoglobin levels preoperatively and 24 hours postoperatively, number of spine levels operated, the volume of blood loss, calculated as intraoperative blood loss plus drainage collection total bleeding, number of transfusions received (units of transfusions received in the trans operative and postoperative period), length of stay and, the complication was also analyzed.

Statistical Analysis The qualitative data were presented as numbers and percentages. Kolmogorov-Smirnov test determined normality of the data. The parametric quantitate data were presented as mean and standard deviation. Comparison between groups was analyzed by using the x2 test or Fisher exact test for qualitative data, and the unparaided t-test for quantitative data. A logistic regression model that included the variables use of oral TXA, sex, BMI, age, number of levels operated, complication, and time of surgery (mean 230min) was performed.

A multivariate forward stepwise logistic regression model was then used to measure the adjusted association of these variables with intraoperative transfusion, with significance set at p<0.05. Data was analyzed using SPSS version 24.0 (IBM COrp, Armonk, NY).