Despite the latest advancements in surgical procedures and anesthesiology, perioperative blood loss is still one of the most important complications of the TKA. Over 1790cc blood may be lost perioperatively by TKA surgery. Blood transfusion, and the subsequent increasing cost and complications such as infections or allergic reactions, along with prolonged hospital stay are a number of possible consequences associated with perioperative blood loss.
TXA as an anti-fibrinolytic agent, is used in TKA surgeries in order to reduce the blood loss and transfusion requirement. Different methods of administration have been introduced and investigated, but the most efficient method is still a matter of debate. A number of RCTs and meta-analyses have reported better results (among IA, IV and combined IA/IV) using the combined IV/IA method[8, 16–18], whereas Meshram  and Lee  studies suggest no significant difference between IA and combined IA/IV use[10, 19]. One of the reasons for this discrepancy, may stem from the PBL measuring method. It is demonstrated that the nadir of postoperative Hb occurs 3–4 days after the TKA[20, 21]. Thus, many studies measure the postoperative Hb on day 3 after surgery in order to subtract it from preoperative Hb and calculate the PBL according to the Nadler formula[16, 21–25]. However, Meshram  and Lee  studies used day 5 after surgery to analyze the perioperative blood loss. This may lead to underestimation of blood loss and the effect of TXA on perioperative blood conservation. Our study also shows that the postoperative day with which preoperative Hb is compared, is important. As demonstrated, no significant Hb loss was noticed between IA, IV and combined IA/IV administration on postoperative day 1. However, significantly less Hb loss was detected for combined IA/IV method compared to the other two methods (IA, IV) on postoperative 2 and 3.
Schnettler found that the combined use of tourniquet and intravenous TXA may be associated with more blood loss compared to the intravenous TXA alone suggesting that using tourniquet may eliminate the effect of intravenous TXA by precluding its access to the site of hemorrhage. Therefore, accompanied use of intraarticular TXA in these circumstances has the rationale of enhancing the availability of the drug intraoperatively.
To the best of our knowledge, no study has compared the three methods (IV, IA, combined IV/IA) simultaneously in those undergoing TKA with tourniquet placement and no use of drainage catheter. Theoretically, use of no suction tube may cause hematoma formation and subsequently increase the postoperative pain. On the other hand, it may eliminate the postoperative blood loss due to tampon effect. Interestingly, recent randomized trial in patients receiving Intravenous TXA during knee replacement, did not show any significant increase in postoperative pain when no drain is used except in the immediate 6 hrs. after the operation . Furthermore, no difference regarding perioperative blood loss was detected between those with and individuals without drain. It may merit special attention, that suction drain might theoretically eliminate the TXA injected intrarticularly and as a result might decrease the possible effect of intraarticular TXA.
Our study demonstrated no significant difference regarding perioperative blood loss between IA and IV administrations. This is in accordance with Chen et. al, study capturing similar result in the presence of drainage system. Furthermore, in our study, the combined IV/IA method had superior effect in declining PBL. Several other studies using drainage catheter arrived at similar conclusion[8, 9]. So, placing drainage catheter, if clamp for enough time after the TXA intraarticular administration, may not affect the efficacy of intraarticular TXA.
PBL in our study in the group using 15 mg/kg dose of TXA (via IV route) was similar to the lee study using 10 mg/kg of the drug intravenously . This is in accordance with Xianhua Ye study which also found no statistically significant difference between 10 and 15 mg/kg IV administration of TXA, either in the amount of intraoperative blood loss or postoperative blood transfusion.
We had no case needing blood transfusion postoperatively. This is in contrast to Yuan et. al. demonstrating 17–39% of participants needing blood transfusion depending on the route of TXA administration postoperatively. Unlike our study in which all TKAs were performed under spinal anesthesia, Yuan et. al. used general anesthesia for the surgeries. Thus, the more percentage of transfusion in that study may be a result of the type of anesthesia. Previous study , also noted that spinal anesthesia compared with general anesthesia may be associated with less blood loss perioperatively.
Our patients were followed for 2 years after the surgery. However, most previous papers had a follow up ranging from 7 days to 11 months[8–10, 16–19]. Similar to Iseki, Karampinas  and Nielsen studies, We detected no thromboembolism after the surgery. In other studies, the rate of thromboembolism was reported 0–4%, 0–2%, 0.3-2% percent using IA, IV and combined IA/IV tranexamic acid respectively during TKA [10, 17, 19]. No statistically significant difference has been demonstrated between TXA administration and the control groups or between different TXA usage methods in many studies[8–10, 16–19]. A population based study on 872416 patients in 510 US hospitals compared the rate of postoperative inpatient complications after total hip or knee replacement between those received TXA perioperatively and those managed without TXA. They found that TXA administration was associated with less incidence of many complications postoperatively including blood transfusion, thromboembolic complications and intensive care unit admission. However, further studies may be needed to evaluate the risk of thromboembolism after TXA consumption in high risk patients (e.g. those with a history of thromboembolism, active cancer, etc.)
Among study limitations, we can mention to the fact that patients undergoing bilateral TKA, general anesthesia or surgical drain use were not included in this paper. Therefore, the results might not be applicable to these subgroups. Our sample size was similar to many studies published in the literature [8, 16]. However, further larger sample size studies are needed to determine the likelihood of rare side effects following TXA use.