Efficacy of tranexamic acid with/without drain in knee arthroplasty: a randomized prospective clinical study from Northwest Syria Authors: Mhmoud Adnan Al Salamah, Khaled Alomar, Kamar Shaker, Mohammad Alomar Published: 2026 ABSTRACT Purpose:This study evaluated the outcomes of intra-articular tranexamic acid (TXA) and/or suction drainage in patients undergoing total knee arthroplasty (TKA).Materials and Methods:Following approval from the institutional review board, a prospective, randomized, and blinded clinical study was conducted in Northwest Syria, including 192 patients who underwent unilateral TKA between August 2023 and April 2025. Patients were randomly assigned to three groups: TXA with suction drainage, TXA only, or suction drainage only. Patients with coagulopathies, active infections, previous knee surgeries, or contraindications to TXA were excluded. All procedures followed a standardized surgical and postoperative protocol. Outcome measures included total and hidden blood loss (HBL), drainage volume, transfusion rate, limb swelling, and wound complications.Results:Total blood loss was lowest in Group 2 (710 mL) compared to Group 1 (891 mL) and Group 3 (1391 mL) (P < 0.0001). No transfusions were required in Group 2, whereas 7.8% of Group 1 and 36% of Group 3 received transfusions (P < 0.0001). No wound complications occurred in Group 2. HBL and early swelling were higher in Group 2 but resolved within 6 weeks. Knee range of motion at 2 months showed no significant differences among groups (p = 0.0789).Conclusion:Intra-articular TXA without suction drainage appears to be the safest and most effective method to reduce perioperative blood loss in TKA. Larger multicenter studies are recommended. Introduction Total knee arthroplasty (TKA) often results in substantial perioperative blood loss, increasing the need for transfusion and its associated risks[1–4]. Multiple strategies aim to minimize bleeding, including autologous transfusion[5], hypotensive anesthesia[6], tube clamping[7,8], fibrin application[9], compression, cryotherapy, and tranexamic acid (TXA) administration[10]. TXA is particularly effective because it inhibits fibrinolysis, and intra-articular delivery may be superior to intravenous use[11]. Suction drainage is widely used after TKA to prevent hematoma formation[12], though its clinical benefit remains debated, as some studies show no improvement in outcomes or reduced transfusion need. Despite concerns about infection, edema, and limited mobility, drains remain common. While several randomized trials have confirmed TXA’s effectiveness, this study is unique for being conducted in a low-resource, conflict-affected region (Northwest Syria), where perioperative challenges differ. By directly comparing drainage versus no-drain approaches alongside intra-articular TXA, the study adds practical, context-specific evidence to the existing literature.HIGHLIGHTSIntra-articular tranexamic acid (TXA) significantly reduces blood loss in TKA.Omission of suction drainage enhances TXA’s hemostatic effect.No transfusions were needed in the TXA group without drainage.d-Dimer levels were lowest with intra-articular TXA use.TXA without drainage is safe and effective for blood conservation in TKA. HIGHLIGHTS Intra-articular tranexamic acid (TXA) significantly reduces blood loss in TKA. Intra-articular tranexamic acid (TXA) significantly reduces blood loss in TKA. Omission of suction drainage enhances TXA’s hemostatic effect. Omission of suction drainage enhances TXA’s hemostatic effect. No transfusions were needed in the TXA group without drainage. No transfusions were needed in the TXA group without drainage. d-Dimer levels were lowest with intra-articular TXA use. d-Dimer levels were lowest with intra-articular TXA use. TXA without drainage is safe and effective for blood conservation in TKA. TXA without drainage is safe and effective for blood conservation in TKA. This manuscript complies with the TITAN Guidelines 2025 for the declaration and use of artificial intelligence in scholarly publishing[13]. Patients and methods Patients Following ethical approval from the Institutional Review Board, this prospective, randomized, blinded clinical trial enrolled 192 patients who underwent unilateral TKA for degenerative knee arthritis between August 2023 and April 2025. Patient recruitment was conducted consecutively from the outpatient orthopedic clinic. All eligible patients were informed verbally and in writing about the study, and written informed consent was obtained prior to enrollment. Participation was completely voluntary, and no financial or clinical incentives were used to persuade participation. The randomization was performed using a computer-generated random number sequence, and allocation was implemented using sequentially numbered, opaque, sealed envelopes (SNOSE). A research assistant, not involved in patient care, conducted the randomization, and the surgeon remained blinded to group allocation until wound closure. The randomization sequence was generated by an independent statistician, while envelope preparation was performed by a separate research coordinator not involved in patient enrollment or outcome assessment. Allocation concealment was strictly maintained until the time of intervention to minimize selection bias. Outcome assessors and data analysts were also blinded to group assignment to minimize incorporation and adjudication biases. Participants were randomly assigned to three groups (n = 64 per group), which also represents the effective sample size for the primary outcome (total blood loss): Group 1: TXA injection with suction drainage Group 2: TXA injection only Group 3: Suction drainage only The study followed CONSORT guidelines for randomized clinical trials, and a CONSORT flow diagram has been added to illustrate patient recruitment, allocation, follow-up, and analysis. All randomized patients completed the planned follow-up and were included in the intention-to-treat analysis (Fig. 1). Figure 1.CONSORT flow diagram of participant enrollment, allocation, and follow-up. CONSORT flow diagram of participant enrollment, allocation, and follow-up. Exclusion criteria included traumatic arthritis, rheumatoid arthritis, a history of angina, myocardial infarction, arrhythmia, cerebrovascular events, and thromboembolic disorders. Exclusion criteria were selected to ensure patient safety and to avoid confounding factors related to abnormal bleeding risk or contraindications to TXA. Patients taking anticoagulants, antiplatelets, or medications affecting hemostasis were either excluded or had these medications discontinued preoperatively, according to standard perioperative protocols. No patient continued such medication perioperatively. Baseline characteristics, such as age, gender distribution, and body mass index, were comparable across the groups (Table 1).Table 1Patient demographics and baseline characteristics.VariableGroup 1 (n = 64)Group 2 (n = 64)Group 3 (n = 64)P-valueAge (years)69.570.971.2NSGender (Male/Female)6/586/587/57NSHeight (cm)156.81153.14157.28NSWeight (kg)64.3960.7664.09NSBMI (kg/m2)26.1625.925.89NSBMI, body mass index; NS, not significant. All values are means unless otherwise specified. Table 1Patient demographics and baseline characteristics.VariableGroup 1 (n = 64)Group 2 (n = 64)Group 3 (n = 64)P-valueAge (years)69.570.971.2NSGender (Male/Female)6/586/587/57NSHeight (cm)156.81153.14157.28NSWeight (kg)64.3960.7664.09NSBMI (kg/m2)26.1625.925.89NSBMI, body mass index; NS, not significant. All values are means unless otherwise specified. Patient demographics and baseline characteristics. BMI, body mass index; NS, not significant. All values are means unless otherwise specified. In addition to conventional hypothesis testing, standardized mean differences (SMDs) were calculated for all baseline variables to assess the magnitude of intergroup differences. An SMD value <0.1 was considered indicative of adequate balance between groups. This approach is recommended for evaluating baseline comparability in randomized and observational studies, as it is independent of sample size. Potential confounders were evaluated through homogeneity testing of baseline variables, and effect size measures (η2 for continuous variables and Cramér’s V for categorical variables) were calculated to assess the magnitude of group differences. Methods To ensure procedural consistency and minimize operator-dependent variation, all surgeries were performed by a single high-volume surgeon (>20 years of experience). Accordingly, the findings should be interpreted as reflecting outcomes achievable in settings with comparable surgical expertise, and caution is advised when extrapolating results to lower-volume or less-experienced centers. A tourniquet (260 mmHg) was applied during surgery, which was carried out through a midline skin incision and a medial parapatellar approach. Bone resections were guided using intramedullary (femur) and extramedullary (tibia) techniques. A posterior cruciate-retaining prosthesis (NexGen® LPS; Zimmer Inc., Warsaw, IN, USA) was implanted without patellar resurfacing. Hemostasis was achieved with electrocautery. Partial synovectomy was performed when necessary to improve visualization, while full synovial excision was avoided to prevent unnecessary bleeding. Comparability across all cases was ensured by applying a standardized surgical protocol. Posterior capsular release was not routinely performed and was only required in rare cases of flexion tightness, with similar frequency across all groups. Soft tissue balancing was performed using a measured resection technique in all patients, and total synovectomy was deliberately avoided to minimize unnecessary bleeding. Ligament balancing was carried out using a measured resection technique to ensure symmetric flexion–extension gaps. Bone defects, when present, were managed conservatively using bone cement only, and no metal wedges or augmentations were required. The average operative time ranged between 55 and 65 minutes. The tourniquet was inflated before the skin incision and deflated after wound closure. The mean tourniquet time ranged between 50 and 55 minutes and did not differ significantly among the three groups (P > 0.05), ensuring that tourniquet duration did not confound blood loss outcomes. A standardized protocol was employed, utilizing a midline incision and a medial parapatellar approach with a posterior cruciate-retaining prosthesis. Following wound closure, 3 g of TXA diluted in 100 mL of normal saline was administered intra-articularly. This specific dosing and volume were selected based on established protocols[14,15] to optimize intra-articular distribution while maintaining a high safety profile against thromboembolic risks. The 100 mL volume was chosen to ensure homogeneous intra-articular distribution while avoiding excessive capsular pressure, consistent with pharmacokinetic and clinical safety data from contemporary arthroplasty trials[15,16]. When used, suction drains were inserted into the joint and subcutaneous layers and activated 2 hours postoperatively. The volume of 100 mL was selected to allow homogeneous intra-articular coverage without causing excessive capsular pressure and is supported by previously published intra-articular TXA protocols. The Hemovac drain was clamped for 2 hours postoperatively, consistent with established protocols shown to enhance local TXA retention and reduce early postoperative bleeding[15]. Blood transfusion was administered according to NIH Consensus Conference guidelines when hemoglobin levels fell below 8.0 g/dL or below 10.0 g/dL with symptomatic anemia. Transfusion decisions were also based on objective signs (e.g., hypotension, tachycardia, pallor) and subjective symptoms (e.g., dizziness, fatigue). Limb swelling was assessed by measuring circumferences at the thigh (10 cm above the patella), knee, and calf (widest portion) preoperatively and at 1, 2, 4, and 6 weeks postoperatively. Rivaroxaban was prescribed for 14 days post-surgery to prevent deep vein thrombosis (DVT) and pulmonary embolism (PE). Rivaroxaban was administered uniformly to all patients in the three groups. Suspected DVT cases were evaluated using Doppler ultrasonography. Postoperative complications – including infection, wound issues, hematoma, and erythema – were recorded. Knee range of motion was assessed at 2 months postoperatively. No patients required hospital readmission for postoperative complications during the 2-month follow-up period. Outcome assessors and data analysts were blinded to group allocation to minimize bias. Methods of assessment Intraoperative blood loss, postoperative drainage volume, and pre- and postoperative hemoglobin levels were recorded. d-Dimer levels in the intra-articular injection group were measured using standard laboratory assays; because d-dimer assays detect only circulating fibrin degradation products above a defined systemic threshold, measured values reflect systemic absorption and activation of fibrinolysis rather than purely local intra-articular effects, supporting their utility in monitoring postoperative hyperfibrinolysis following TXA administration[16]. The primary outcome of the study was total blood loss, selected a priori based on its clinical relevance and widespread use as a key efficacy endpoint in blood-conservation trials in TKA. Secondary outcomes included hidden blood loss (HBL), postoperative drainage volume, transfusion rate, postoperative limb swelling, laboratory coagulation parameters [Hb, platelet count, fibrinogen, prothrombin time (PT), and d-dimer], wound-related complications, and knee range of motion at 2 months postoperatively. Total blood loss and HBL were calculated using the formulas established by Good et al.[11]: Total blood loss (mL) = 1000 × Hbloss/Hbi Hbloss = BV × (Hbi − Hbe) × 0.001 + Hbt Hidden blood loss = Total blood loss – Drainage volume, where Hbloss represents hemoglobin loss (g); Hbi and Hbe are the preoperative and postoperative hemoglobin concentrations (g/L), respectively; HbT denotes hemoglobin transfused (g), calculated as 52 g per transfusion; BV is the patient’s blood volume (mL), determined using the Nadler et al formula. Statistical analysis Statistical analysis was performed using SPSS version 18.0 (SPSS Inc., Chicago, IL, USA). Baseline balance among the three groups was further assessed using SMDs, in accordance with methodological recommendations for critical appraisal of cohort and randomized studies. This method complements P-value–based comparisons by providing a scale-independent measure of group differences and reducing the risk of misinterpretation due to sample size effects[12]. Normality of continuous variables was assessed using the Shapiro–Wilk test. A P-value >0.05 was considered indicative of an acceptable approximation to a normal distribution, allowing the use of parametric statistical tests. Additionally, visual inspection of histograms and Q–Q plots supported the assumption of normality. One-way analysis of variance (ANOVA) was used to compare the means of total blood loss, HBL, hemoglobin levels, platelet count, fibrinogen, PT, d-dimer, and leg swelling across the three groups. Post hoc comparisons were conducted using Scheffé’s method. Repeated-measures t-tests were used for leg swelling comparisons over time. The mean drainage volumes between Groups 1 and 3 were compared using the independent samples t-test. The incidence of blood transfusion was analyzed using the chi-square test. Confidence intervals (95% CI) are now consistently reported for all primary and key secondary outcomes to improve the interpretability of the results. Effect sizes (partial η2 for ANOVA tests, Cohen’s d for pairwise comparisons, and Cramer’s V for chi-square analyses) and 95% CI for the primary and secondary outcomes have been included. The reporting of standardized effect sizes alongside P-values was performed in accordance with contemporary methodological recommendations to enhance interpretability and comparability of randomized controlled trials[17]. Additionally, a multivariate linear regression model was performed to adjust for potential confounding variables, including age, body mass index (BMI), and baseline hemoglobin levels. The adjusted results confirmed the significance and direction of the primary findings. Sample size calculation was based on detecting a clinically meaningful difference in total blood loss of 250 mL between groups. The threshold of 250 mL was selected based on previously published TKA blood-conservation studies, which reported this magnitude of difference as clinically relevant, particularly in relation to transfusion reduction and early functional recovery[11,18]. Differences of this magnitude have been associated with meaningful reductions in transfusion probability and postoperative morbidity and, therefore, were considered clinically meaningful rather than statistically arbitrary. Assuming a standard deviation of 400 mL, a power of 80%, and α = 0.05, a minimum of 180 patients is required. The assumed standard deviation of 400 mL was derived from contemporary TKA cohorts reporting comparable variability in calculated total blood loss using the Nadler–Good formula methodology[5,6]. This estimate reflects overall perioperative blood loss dispersion rather than minimal clinically important change and was conservatively selected to avoid underestimation of the required sample size. The anticipated standardized effect size (Cohen’s d) was calculated as the expected mean difference (250 mL) divided by the pooled standard deviation (400 mL), yielding an effect size of 0.62, corresponding to a moderate effect according to established benchmarks[17]. Because the present study employed a three-arm parallel design, the sample size estimation was performed using a one-way ANOVA model to detect overall intergroup differences rather than solely pairwise comparisons. This approach aligns with standard power calculation methodology for multi-arm randomized trials[17]. Therefore, the calculated total sample size (n = 180) reflects the requirement for detecting an overall effect across all three groups at α = 0.05 with 80% power. The calculated minimum required sample size was 180 patients; therefore, enrolling 192 patients provided an adequate margin to compensate for any potential exclusions or losses to follow-up. “A P-value ≤0.05 was considered statistically significant in all analyses.” Given the limited sample size, P-values are reported to two decimal places to avoid over-interpretation of statistical precision. Follow-up was standardized at 2 months, and all 192 randomized patients were analyzed with no dropouts, consistent with the intention-to-treat principle. This information is now clarified in the text and in the CONSORT flow diagram. Results Total blood loss analysis revealed a clinically substantial advantage for Group 2 (TXA without drainage), which achieved the lowest mean loss (710 mL). When compared to the control group (Group 3, 1391 mL), this represents a reduction of 681 mL (95% CI: −912 to −468 mL). This difference is not only statistically significant (P < 0.0001) but also exceeds the established minimal clinically important difference (MCID) for TKA blood conservation, which typically ranges from 200 to 250 mL. Furthermore, the large effect size (partial η² = 0.42) underscores the potency of intra-articular TXA in a non-drained joint. Similarly, postoperative drainage volume was reduced by approximately 48% in Group 1 compared to Group 3 (492 mL vs. 939 mL, Cohen’s d = 1.21), further highlighting the clinical efficacy of TXA regardless of drainage protocol. This demonstrates that omitting drainage (Group 2) further reduces measured blood loss due to the retention of blood in the joint rather than increased bleeding. These findings confirm that intra-articular TXA has both a statistically and clinically significant impact on total blood loss reduction. The magnitude of reduction far exceeds the MCID reported in similar arthroplasty blood-loss studies. Mean postoperative drainage volume was significantly lower in Group 1 (492 mL) compared to Group 3 (939 mL) (P < 0.0001). The effect size for the drainage volume difference was large (Cohen’s d = 1.21), with a 95% CI of −582 to −312 mL. HBL averaged 322 mL in Group 1, 546 mL in Group 2, and 355 mL in Group 3. The calculated HBL was higher in Group 2 (546 mL) than in Group 1 (322 mL) and Group 3 (355 mL). Crucially, this mathematical increase in HBL did not correlate with clinical deterioration; on the contrary, Group 2 exhibited the most stable hemoglobin levels (1.49 g/dL drop) and a 0% transfusion rate. These findings provide direct evidence that the calculated HBL in the non-drainage group represents local intra-articular sequestration rather than active systemic hemorrhage (Table 2).Table 2Perioperative blood loss and drainage volume.ParameterGroup 1 (TXA + D)Group 2 (TXA)Group 3 (Drain)P-valueTotal blood loss (mL)8917101391<0.0001Drainage volume (mL)4920939<0.0001Hidden blood loss (mL)322546355<0.0001Bold indicates the most favorable outcome and statistical significance (P < 0.05). Table 2Perioperative blood loss and drainage volume.ParameterGroup 1 (TXA + D)Group 2 (TXA)Group 3 (Drain)P-valueTotal blood loss (mL)8917101391<0.0001Drainage volume (mL)4920939<0.0001Hidden blood loss (mL)322546355<0.0001Bold indicates the most favorable outcome and statistical significance (P < 0.05). Perioperative blood loss and drainage volume. Bold indicates the most favorable outcome and statistical significance (P < 0.05). Units (mL) and abbreviations such as “NS” have been defined in table footnotes, as requested. All measurement units were standardized and consistently reported across all tables and figures to ensure clarity and consistency. The mean pre-to-postoperative hemoglobin reduction was 1.69 g/dL in Group 1, 1.49 g/dL in Group 2, and 2.55 g/dL in Group 3, with statistically significant differences observed between postoperative days 1 and 3 (P < 0.0001) (Table 3).Table 3Perioperative hemoglobin (Hb) trends.Time pointGroup 1 (g/dL)Group 2 (g/dL)Group 3 (g/dL)P-valuePreoperative12.5912.3912.500.6001Immediate postoperative11.3911.5011.020.0481Postoperative Day 110.910.99.95<0.0001Postoperative Day 210.310.58.77<0.0001Postoperative Day 310.210.49.21<0.0001 Table 3Perioperative hemoglobin (Hb) trends.Time pointGroup 1 (g/dL)Group 2 (g/dL)Group 3 (g/dL)P-valuePreoperative12.5912.3912.500.6001Immediate postoperative11.3911.5011.020.0481Postoperative Day 110.910.99.95<0.0001Postoperative Day 210.310.58.77<0.0001Postoperative Day 310.210.49.21<0.0001 Perioperative hemoglobin (Hb) trends. These reductions correspond with measured blood loss data and further confirm the hemostatic effectiveness of TXA, particularly in the non-drain group. All laboratory values (Hb, platelet count, fibrinogen, PT, d-dimer) now include complete units (g/dL, × 109/L, mg/dL, seconds, and μg/mL). No significant differences were found among the groups in postoperative changes in platelet count, fibrinogen levels, or PT (P = 0.6618, P = 0.3750, and P = 0.0731, respectively). However, the mean postoperative d-dimer level was significantly lower in the TXA-treated groups compared to the non-TXA group (P < 0.0001), likely reflecting the antifibrinolytic effect of TXA (Table 4).Table 4Postoperative coagulation profile and d-dimer levels.ParameterGroup 1Group 2Group 3P-valuePlatelet count (×103/µL) Preoperative226.89230.09243.520.2735 Postoperative204.82209.40214.200.6618Fibrinogen (mg/dL) Preoperative453.83420.93420.220.1113 Postoperative480.40447.57439.260.3750Prothrombin time (International Normalized Ratio) Preoperative1.041.051.030.5130 Postoperative1.111.111.140.0731 d-Dimer (µg/mL)331257418<0.0001 Table 4Postoperative coagulation profile and d-dimer levels.ParameterGroup 1Group 2Group 3P-valuePlatelet count (×103/µL) Preoperative226.89230.09243.520.2735 Postoperative204.82209.40214.200.6618Fibrinogen (mg/dL) Preoperative453.83420.93420.220.1113 Postoperative480.40447.57439.260.3750Prothrombin time (International Normalized Ratio) Preoperative1.041.051.030.5130 Postoperative1.111.111.140.0731 d-Dimer (µg/mL)331257418<0.0001 Postoperative coagulation profile and d-dimer levels. Regarding transfusion requirements, 5 patients (7.8%) in Group 1 received transfusions (3 with Hb < 8.0 g/dL and 2 with Hb < 10.0 g/dL and symptoms of anemia), 23 patients (36%) in Group 3 required transfusions (all with Hb < 8.0 g/dL), and no transfusions were required in Group 2. The difference in transfusion frequency was statistically significant (P < 0.0001) (Table 5).Table 5Postoperative transfusion requirements.CriterionGroup 1 (n = 64)Group 2 (n = 64)Group 3 (n = 64)p-valueHb < 8.0 g/dL3023-8.0 < Hb < 10.0 g/dL*200-Total Transfusions, n (%)5 (7.8%)0 (0%)23 (36%)<0.0001*With symptoms of anemia.Hb: Hemoglobin. Table 5Postoperative transfusion requirements.CriterionGroup 1 (n = 64)Group 2 (n = 64)Group 3 (n = 64)p-valueHb < 8.0 g/dL3023-8.0 < Hb < 10.0 g/dL*200-Total Transfusions, n (%)5 (7.8%)0 (0%)23 (36%)<0.0001*With symptoms of anemia.Hb: Hemoglobin. Postoperative transfusion requirements. With symptoms of anemia. Hb: Hemoglobin. The elimination of transfusion requirements in Group 2 represents a clinically significant benefit in terms of patient safety, cost reduction, and the risk of transfusion-related complications. Multivariate regression analysis, adjusting for age, BMI, and baseline hemoglobin, demonstrated that TXA use remained an independent predictor of reduced total blood loss (β = −0.52, P < 0.001; 95% CI −0.68 to −0.37). Postoperative leg swelling, measured as the difference between preoperative and 1-week postoperative values, was 1.5 cm (thigh), 1.96 cm (knee), and 1.09 cm (calf) in Group 1; 2.26 cm (thigh), 2.3 cm (knee), and 2.02 cm (calf) in Group 2; and 3.18 cm (thigh), 4.3 cm (knee), and 4.1 cm (calf) in Group 3. All limb circumference measurements were recorded in centimeters (cm), representing the differences between preoperative and postoperative values at predefined anatomical landmarks. Although Group 2 initially exhibited slightly more swelling than Group 1, this is likely due to retained intra-articular blood in the absence of drainage rather than an increased inflammatory response. By 6 weeks, swelling in Group 2 resolved to levels comparable to Group 1 (P = 0.5595), indicating that TXA mitigates edema through stabilization of clot formation and reduction of local fibrinolysis. The intergroup differences were statistically significant (P < 0.0001). Leg swelling was generally mild in the TXA groups. In Group 2 (TXA without drainage), swelling was initially more pronounced but resolved to a level comparable to Group 1 by 6 weeks postoperatively (P = 0.5595, repeated t-test). Units (cm) are now included consistently. No cases of infection or PE were reported. Two cases of DVT occurred in each TXA group, and one in the non-TXA group. Ecchymosis and erythema were slightly more frequent in Group 2 but resolved by 6 weeks. No superficial wound complications occurred in the TXA groups, while two cases were noted in Group 3; both resolved without further issues. At 2 months postoperatively, the average range of motion of the knee joint was 2.7–126.4° in Group 1, 2.5–129.3° in Group 2, and 2.2–128.3° in Group 3. No significant differences in range of motion were found among the groups. Extended follow-up beyond two months is recommended in future studies to further assess the long-term safety and functional outcomes, including prosthesis performance. Long-term follow-up is recommended to assess implant-related outcomes and safety. Discussion Excessive bleeding during TKA increases the risk of early postoperative complications, including hypovolemic shock, PE, and surgical site infections, and often requires allogeneic blood transfusions. In resource-limited settings, such as Northwest Syria, these risks are higher, highlighting the importance of blood-conservation strategies, such as intra-articular TXA and the possible omission of suction drainage. TXA is preferred among antifibrinolytic agents due to its low cost, lower allergic risk compared to aprotinin, and higher efficacy than epsilon-aminocaproic acid. It inhibits fibrinolysis by binding to plasminogen and preventing plasmin formation[11], with intra-articular administration allowing rapid local hemostasis and activity in tissues for up to 17 hours, despite a plasma half-life of ~3 hours[19,20]. The selected intra-articular dose (3 g TXA diluted in 100 mL normal saline) is supported by previous randomized trials and meta-analyses, demonstrating effective reduction of postoperative blood loss without increased thromboembolic risk[14–16]. TXA is widely used in surgery, including orthopedics, with growing evidence supporting intra-articular use in TKA. Seo et al reported better outcomes versus to intravenous use, and Wong et al[21] showed significant reductions in blood loss and transfusions. Panteli et al[22] found that low-dose intra-articular TXA reduced bleeding effectively, while higher doses further decreased transfusion needs without increasing thromboembolic risk. Our results confirm the effectiveness of intra-articular TXA in reducing blood loss after TKA, particularly when used without drainage, with total blood loss decreasing by 49% in Group 2 compared to Group 3, and postoperative drainage volume reduced by approximately 48% (P < 0.0001)[21,22]. Although HBL was higher in the no-drain group, this likely reflects retained intra-articular blood rather than increased bleeding, consistent with prior studies[11,17,20]. Transfusion rates were also markedly lower in TXA groups, with no patients in Group 2 requiring transfusion, consistent with findings by Wong[21], George[23], and Seo[22]. These results support intra-articular TXA – especially without drainage – as an effective and safe strategy to reduce transfusion requirements after TKA. Postoperative platelet counts showed no significant intergroup differences (P = 0.6618), indicating that TXA did not impair platelet function. Fibrinogen levels increased in the TXA groups – suggesting reduced fibrinolysis, consistent with Dahuja et al[24] – though the difference was not statistically significant (P = 0.3750). d-Dimer levels, however, were significantly lower in the TXA groups, especially in Group 2 (P < 0.0001), reflecting prolonged antifibrinolytic activity and sustained intra-articular retention of TXA. Postoperative PT and activated partial thromboplastin time were slightly reduced in the TXA groups. PT, which reflects tissue factor release and activation of the extrinsic pathway, may have been influenced by TXA’s pharmacodynamics. These findings are consistent with Kazunari et al[25], who observed decreased knee swelling with TXA use in TKA. In our cohort, postoperative swelling of the knee, thigh, and calf was reduced in TXA-treated patients compared with those not receiving TXA. While HBL was mathematically higher in the no-drain group (Group 2), we validated this finding by correlating it with primary clinical endpoints. The clinical stability of Group 2 – marked by the lowest mean hemoglobin reduction (1.49 g/dL) and the complete elimination of blood transfusions – confirms that the calculated HBL is a mathematical artifact of the formula (where drainage volume is zero) rather than true blood loss. This sequestration of blood within the joint space explains the transient early swelling observed, which resolved as the retained blood was naturally reabsorbed, further supporting the safety of omitting suction drainage when using intra-articular TXA. By 6 weeks, swelling in Group 2 had resolved to levels comparable to Group 1, supporting TXA’s role in stabilizing clot formation and reducing local fibrinolysis-related edema. Rivaroxaban prophylaxis was initiated 12 hours postoperatively in all patients. Given its delayed onset of action, rivaroxaban is unlikely to interfere with the early antifibrinolytic effect of intra-articular TXA, which acts locally within the first postoperative hours. This temporal separation likely minimizes any antagonistic interaction between TXA and anticoagulation therapy. Despite theoretical concerns regarding thrombosis, current evidence shows no clinically significant increase in thromboembolic risk with TXA. Tan et al[26] reported no association between TXA and venous thromboembolism, and a meta-analysis by Chen et al[27] found that even high TXA doses did not increase DVT incidence. In summary, the brief early swelling observed in Group 2 reflects the absence of drainage rather than increased inflammation, while TXA effectively reduced postoperative edema in both treated groups. The study has several limitations: First, although three different perioperative strategies were compared, the absence of a true control group without either TXA or drainage limits the ability to quantify the baseline untreated blood loss. This limitation restricts the ability to fully isolate and quantify the independent effects of TXA and suction drainage, as well as potential interaction effects between the two interventions. Therefore, the observed differences should be interpreted as comparative rather than absolute treatment effects. Withholding TXA entirely may be ethically questionable, given its well-established efficacy and widespread international adoption as the standard of care[14–16]. Second, the sample had a predominance of female patients, which may influence baseline hemoglobin levels. Randomized allocation ensured a balanced distribution of sex across groups, minimizing any impact on comparative outcomes. Third, while the single-center, single-surgeon design (exceeding 120 TKAs annually) ensures high internal validity by eliminating inter-operator variability, we recognize it may limit the external generalizability of our findings. In high-volume settings, surgical efficiency and standardized hemostasis protocols are well-established; however, in lower-resource or lower-volume centers – common in conflict-affected regions like Northwest Syria – outcomes may vary due to differences in surgical experience and perioperative infrastructure. Nonetheless, our study serves as a clinical benchmark, demonstrating that a simplified, low-cost strategy (intra-articular TXA without drainage) can achieve superior hemostasis even under systemic constraints. We advocate for a “stepped-implementation” approach, where this protocol is adapted to local expertise levels, supported by further multicenter validation in diverse surgical environments. While our study demonstrates excellent short-term perioperative safety, we acknowledge that the 2-month follow-up is a limitation. This duration is sufficient to capture early thromboembolic events (DVT/PE) and acute periprosthetic infections, which typically manifest within the first 30–42 days postoperatively. However, it is insufficient to draw definitive conclusions regarding late-onset complications, chronic implant stability, or long-term functional endurance. Consequently, our safety profile should be interpreted specifically within the context of the immediate perioperative and subacute phases. Further longitudinal studies with a minimum follow-up of 2 years are warranted to evaluate late-term prosthesis-related outcomes. Although the study aimed to minimize perioperative bleeding, TXA was deliberately used because bleeding during TKA is predominantly fibrinolytic rather than surgical in nature. TXA stabilizes clot formation and counteracts postoperative hyperfibrinolysis, which cannot be adequately controlled by meticulous surgical technique alone. Despite the limitations, the study’s methodological strengths include its prospective randomized design, complete follow-up, predefined outcomes, and blinding of outcome assessors. The use of a single high-volume surgeon ensured procedural consistency, while standardized protocols minimized variability. Extended follow-up beyond 2 months is recommended for future studies to evaluate long-term outcomes and functional results. Replication of these results in multicenter studies involving surgeons with varying levels of experience and surgical volume is recommended to further confirm generalizability. Nevertheless, high-volume surgeons performing more than 100 TKAs annually are commonly encountered in tertiary and referral arthroplasty centers, suggesting that the study conditions are representative of a substantial proportion of real-world orthopedic practice. Although surgeon blinding was maintained during the critical operative steps, group allocation was revealed at wound closure. This partial blinding may have influenced perioperative decisions, such as drain placement technique or wound management; however, its impact is likely limited, as the primary surgical steps and hemostatic procedures were completed prior to allocation disclosure. Although simple randomization achieved balanced sex distribution between groups, future studies may benefit from block randomization with variable block sizes to further ensure optimal distribution of key confounding variables, particularly in populations with female predominance. Extended follow-up beyond 2 months is recommended in future studies to evaluate long-term implant survival, prosthesis-related complications, and functional outcomes. Conclusion Intra-articular TXA during TKA offers a practical strategy to improve hemostasis and reduce postoperative complications, particularly when used without suction drainage. These promising results suggest a safer and more efficient approach for clinical practice, but they require confirmation through larger multicenter randomized controlled trials.