The impact of tranexamic acid on surgical efficiency and visualization in arthroscopic rotator cuff repair: a systematic review and meta-analysis

Article information

Clin Shoulder Elb. 2026;29(1):28-43

Publication date (electronic) : 2025 December 18

doi : https://doi.org/10.5397/cise.2025.00780

Carlos Acosta-Olivo, Gregorio Villarreal-Villarreal, Ricardo Salinas-Garza, Víctor Peña-Martínez, Francisco Arrambide-Garza, Mario Simental-Mendía

Orthopedic and Trauma Service, Hospital Universitario Dr. José Eleuterio González, Universidad Autónoma de Nuevo León, Monterrey, Mexico

Corresponding author: Mario Simental-Mendía, Orthopedic and Trauma Service, Hospital Universitario Dr. José Eleuterio González, Universidad Autónoma de Nuevo León, Monterrey, Ave Madero y Gonzalitos S/N, Mitras Centro, Monterrey, Mexico Tel: +52-81-8347-6698, Email: mario.simentalme@uanl.edu.mx

Received 2025 July 2; Revised 2025 September 10; Accepted 2025 October 5.

Abstract

INTRODUCTION

Tranexamic acid (TXA) is widely used across multiple surgical specialties, including cardiac, neurosurgical, and orthopedic procedures. It is most frequently administered intravenously in the preoperative setting, although the topical and oral routes have been explored and produced comparable outcomes. TXA acts as an antifibrinolytic agent, stabilizing fibrin clots to reduce perioperative bleeding and transfusion requirements [1,2].

In orthopedic surgery, TXA has been shown to reduce blood loss and transfusion rates without increasing the risk of venous thromboembolism, particularly during total hip and knee arthroplasty [3]. Intravenous administration remains the most common and effective route, given its rapid systemic distribution and ability to reach large joints efficiently [2]. Topical application has demonstrated efficacy similar to intravenous delivery in terms of blood loss and transfusion rates [4,5]. Oral administration has also been reported to provide comparable safety, with additional benefits of reduced transfusion rates, cost-effectiveness, and ease of administration for total hip and knee replacement, as well as for intertrochanteric fracture surgery [6,7].

In shoulder surgery, intravenous TXA has been associated with reduced perioperative blood loss, lower postoperative pain in the early recovery period, and decreased hematoma formation [8]. A meta-analysis of patients undergoing total shoulder arthroplasty, reverse shoulder arthroplasty, or arthroscopic rotator cuff repair demonstrated that TXA reduced blood loss and might have contributed to lower postoperative pain and shorter operation times [9]. Another meta-analysis confirmed its safety profile, showing no increased risk of complications or thromboembolic events; however, it found that TXA did not reduce intraoperative bleeding sufficiently to improve visualization during arthroscopic rotator cuff repair [10]. Those findings are clinically relevant because several factors—such as low perioperative hemoglobin, procedures for periprosthetic joint infection, and fracture-related surgeries—have been linked to increased risk of transfusion during shoulder surgery [11].

Several systematic reviews and meta-analyses have specifically examined the role of TXA in arthroscopically treated rotator cuff tears [10,12-14]. The most recent systematic review on this topic was conducted by Jain et al. [15]. However, the interpretation of results across those studies is limited by methodological heterogeneity, particularly the lack of pre- and postoperative values for key clinical outcomes. This absence of standardized data complicates the assessment of TXA efficacy—whether administered intravenously or intra-articularly—and hinders reliable interpretation. Therefore, we conducted this systematic review to evaluate the efficacy of TXA in primary arthroscopic rotator cuff repair, with specific attention to operative field visualization, operation time, mean arterial pressure (MAP), hospital length of stay, postoperative pain and function, and complications including thromboembolic events.

METHODS

This review adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for conducting systematic reviews and meta-analyses [16]. This study did not involve human participants, and therefore ethical approval and consent to participate were not applicable.

Data Synthesis

This review presents a summary of each included study, detailing the intervention type, target population, and outcomes of interest (Table 1). Key findings and conclusions from each study were extracted, examined alongside articles reporting similar outcomes of interest, and organized into distinct sections as below. Due to the diversity of the evaluated outcomes and inconsistencies in data reporting, forest plot construction across studies was limited to assessing the level of pain reported on a visual analog scale (VAS). The mean score was the single-group summary measure for aggregating results separately for control and intervention groups. To account for inter-study variability, a random-effects model was used. Study weights were assigned based on inverse variance. Plot construction was conducted using the meta package in R version 4.2.2 (R Foundation for Statistical Computing).

Table 1.

Characteristics of the included studies

The intervention effect on functional outcomes (American Shoulder and Elbow Surgeons [ASES] score) was determined through the mean difference (MD) and 95% CI. The mean change from baseline in both the intervention and control groups was used for analysis. The standard deviation of the MD was computed using the intervention-specific standard deviations and an imputed correlation coefficient of 0.5 [18]. A meta-analysis was conducted using a fixed-effects model and the generic inverse variance method. The Cochrane's Q statistic test was applied to assess consistency and heterogeneity among studies, with a significance threshold set at P<0.05. Additionally, the I² statistic was used, with 0%−25% heterogeneity categorized as unimportant, >25%–50% as moderate, and >50% as substantial. Our meta-analysis was conducted in Review Manager statistical software version 5.4.1.

RESULTS

Search Strategy and Study Selection

The screening process for study selection is summarized in the PRISMA flow diagram (Fig. 1). We identified 298 records in the database search. After removing duplicate records, 130 studies remained for screening. Ninety-eight records were excluded during title and abstract screening. Full-text retrieval was attempted for 32 reports, but three could not be accessed. Among the 29 full-text reports assessed for eligibility, 17 were excluded due to inappropriate study design (10 were not RCTs), intervention (5 did not involve TXA), or population differences (2 included a different patient group). Ultimately, 12 studies met the inclusion criteria and were included in this review [19-30].

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram. RCT: randomized clinical trial.

Quality Assessment

The risk of bias assessment demonstrated variability across the included studies (Fig. 2). Among them, some demonstrated an overall low risk of bias [20,23,24,30], and others raised concerns in specific domains. The primary issues identified were deviations from the intended interventions, missing outcome data, and concerns about the measurement and selection of the reported results. Although no study was found to have a high risk of bias, potential biases due to methodological limitations could influence the overall reliability of the evidence.

Fig. 2.

Risk of bias in the included studies.

Summary of Outcomes

Visualization

Eight studies evaluated visual clarity during surgery [19-21,23,25-27,29]. Four of them were based on a 10-point scale [19,21,25,29], three used a 3-point scale [23,26,27], and one used a 5-point scale [20]. All but two studies evaluated visualization globally. Takahashi et al. [27] evaluated the percentage of visual clarity every 15 minutes during the surgical procedure, based on a 3-point scale, and Shin et al. [26] evaluated visual clarity on a 3-point scale in four stages during surgery (stage I: soft tissue, intra-articular procedures, stage II: acromioplasty, stage III: bursectomy, and stage IV: greater tuberoplasty). In three studies [20,21,29], the TXA (intravenous or intra-articular) group showed better visualization of the surgical field than the saline solution group. In the study of Shin et al. [26], the TXA group showed better visualization only during the intra-articular soft tissue procedure, and Takahashi et al. [27] reported better visualization during most of the surgical time. Only one study compared the use of intra-articular TXA with that of intra-articular epinephrine, and those results did not show a difference in favor. The heterogeneity in visualization assessment scales made it difficult to determine whether TXA has a clear effect during shoulder arthroscopy. However, based on the analyzed results, TXA does not appear to improve visualization of the surgical field (Table 2).

Table 2.

Visual clarity of the surgical field

Pain

Seven studies evaluated pain related to the use of TXA [20,22-25,27,30]. Three studies evaluated the preoperative and postoperative pain [24,27,30]. The rest of the included studies reported pain evaluation in several settings of follow-up, from 0 hours to 52 weeks. We constructed a forest plot using the studies that reported pain scores on a VAS at similar time points [20,22-25] to provide a visual summary of the outcome distribution across studies. In the TXA group (Fig. 3), the mean pain score was 2.74 at 8 hour (n=2; 95% CI, 1.74–3.74; I²=83.4%), 2.91 at 24 hour (n=4; 95% CI, 1.89–3.92; I²=96.4%), and 2.29 at 48–72 hour (n=2; 95% CI, 0.13–4.44; I²=97.5%). In the control group (Fig. 4), the mean pain score was 3.42 at 8 hour (n=2; 95% CI, 3.0–3.83; I²=33.4%), 3.4 at 24 hour (n=4; 95% CI, 2.39–4.41; I²=94.1%), and 2.57 at 48–72 hour (n=2; 95% CI, 0.02–5.11; I²=97%). In the control group (Fig. 4), the mean pain score was 3.42 at 8 hours (n=2; 95% CI, 3.0–3.83; I²=33.4%), 3.4 at 24 hours (n=4; 95% CI, 2.39–4.41; I²=94.1%), and 2.57 at 48–72 hours (n=2; 95% CI, 0.02–5.11; I²=97%). Within that 8–72 hour threshold, the mean pain score in the TXA group was lower than in the control group; however, the absolute differences in mean pain scores and wide confidence intervals reflect uncertainty in the effect estimate.

Fig. 3.

Forest plot of pain in the tranexamic acid group. TE: treatment effect, SE(TE): standard error of treatment effect.

Fig. 4.

Forest plot of pain in the control group. TE: treatment effect, SE(TE): standard error of treatment effect.

Functional assessment

The forest plot in Fig. 5 summarizes the MD in post-operative ASES scores between patients who received TXA and those in the control group. The pooled MD using a fixed-effects model was –2.03 (95% CI, –7.62 to 3.56; P=0.48), indicating no statistically significant difference. Even though the chi-square test and I² value suggested no considerable heterogeneity, the results from the studies show wide confidence intervals that cross the line of no effect (zero).

Fig. 5.

Forest plot of functionality (American Shoulder and Elbow Surgeons score). TXA: tranexamic acid, SD: standard deviation, IV: inverse variance.

Total operation time

Seven studies evaluated the total operation time [19-21,23,25,26,30]. Four of those studies reported the use of the beach chair position [19-21,25], two studies positioned the patient in lateral decubitus [23,26], and one study did not describe the surgical position [30]. Three studies used the double-row technique (transosseous equivalent or suture bridge) [19,21,23], one reported a single-row or transosseous equivalent [26], another reported a single-row technique [20], and two studies did not report the surgical technique [25,30]. Most studies showed no significant difference in operation time between TXA and control groups (P>0.05). Only Bildik et al. [20] reported that the operation time was significantly shorter in the TXA group (TXA, 55.73±8.62 minutes; control, 67.26±7.34 minutes; P=0.001). No evidence suggests that TXA shortens surgery time; operation durations in the TXA and control groups were generally comparable (Table 3).

Table 3.

Total operation time

Mean arterial pressure

Five studies included a comparison of MAP [19,20,23,25,26]. Three of those studies administered the TXA intravenously [24-26], and the other two used the intra-articular approach [19,20]. Changes in this value showed no difference between the treatment and control groups, indicating no clear benefit from the use of TXA (Table 4).

Table 4.

Mean arterial pressure

Amount of fluid

Four studies measured the total irrigation fluid used during the arthroscopic procedure [19,21,25,29]. Two studies reported the use of significantly less fluid in patients who received TXA [21,29]. Wang et al. [29] compared four groups, and the patients who received TXA intravenously and in the irrigation fluid showed irrigation with significantly less fluid than those in the other groups. The effect of TXA in reducing the amount of fluid used during surgery remains uncertain. However, the results of Wang et al. [29] showed a significant reduction in fluid usage with the administration of TXA both intravenously and intra-articularly (Table 5).

Table 5.

Amount of fluid

Hospital length of stay

Three studies evaluated the hospital length of stay [23,26,30], with TXA administered intravenously in all those experimental groups. No significant differences were observed between the treatment and control groups. The use of TXA did not have any effect on the duration of the hospital stay following the procedure (Table 6).

Table 6.

Hospital length of stay

Tear size

Four studies evaluated the size of the treated rotator cuff tear [19-21,23]. Among them, only Ersin et al. [21] categorized the tears as small, medium, or large. None of the four studies reported a significant difference between groups.

Complications

No thromboembolic complications associated with the use of TXA were reported in any of the included studies. Five studies did not provide data on complications [21,25,27-29]. One study reported complications unrelated to TXA use [24], including six cases of adhesive capsulitis and three cases of rotator cuff re-tear. Another study reported pain and shoulder swelling [30], with no significant difference between groups.

DISCUSSION

Studies of the use of TXA in arthroscopic rotator cuff repair have yielded inconclusive results across multiple evaluated outcomes. Visual clarity was the primary endpoint in most studies; however, evidence for the efficacy of TXA remains inconsistent, irrespective of intravenous or intra-articular administration. The secondary outcomes of patient-reported measures, operation duration, MAP, and hospital length of stay also failed to demonstrate significant benefits associated with TXA use. Even the potential effect of TXA on reducing intraoperative irrigation fluid volume remains uncertain. Most of the included studies compared TXA with placebo (saline solution), but one trial evaluated TXA against epinephrine [19].

Several meta-analyses have assessed the effect of TXA on visual clarity during arthroscopic procedures, and the results have been mixed. Malik et al. [31] and Hurley et al. [32] reported that TXA improves intraoperative visualization, whereas Zhao et al. [12] and Jain et al. [15] described only a potential benefit. In contrast, Sun et al. [10] and Alyousef et al. [13] observed no significant improvement, which is consistent with our findings. Closer examination of those studies highlights important limitations. In the meta-analysis by Malik et al. [31], data from certain studies (e.g., [20,23]) were incorrectly extracted, raising concerns about the validity of their conclusions. Hurley et al. [32] did not conduct a pooled analysis or provide a forest plot but instead offered a narrative synthesis that suggested improved visualization with TXA. Conversely, Sun et al. [10], Alyousef et al. [13], and Alzobi et al. [14] reported no benefit, findings that align with our results.

Our analysis focused on arthroscopic rotator cuff repair found no definitive benefit of TXA (intravenous or intra-articular) in improving visual clarity. We included eight studies for critical appraisal of this outcome, three more than the largest previously reported analysis. A significant challenge in synthesizing the available evidence is the considerable heterogeneity in the methods used to assess visual clarity. The reported scales ranged from 3- to 10-point systems, limiting comparability and hindering data pooling. One study used time-based assessments, such as calculating the proportion of operative time with optimal visualization [27], and another evaluated clarity across surgical stages [26]. Given that level of variability, the construction of forest plots and subsequent meta-analysis were not recommended. One study (Bildik et al. [20]) reported that intra-articular TXA administration produced better surgical field visualization than intravenous administration; the route of administration for TXA might influence its effects, and future studies comparing administration approaches should clarify that difference. The lack of standardized assessment methods might have either under- or over-estimated the true effect of TXA on surgical visualization. The implementation of a standardized surgical visual clarity evaluation could help to elucidate whether the use of TXA offers true clinical benefit for surgeons and patients.

The one RCT that compared TXA, epinephrine, and their combination in shoulder arthroscopy for various pathologies reported that TXA was less effective than epinephrine in improving visualization [33]. Those findings are consistent with the meta-analysis by Zhao et al. [12], which demonstrated that TXA reduced the operation time compared with saline but not compared with epinephrine, suggesting that TXA might not be a superior option for enhancing surgical field clarity.

From a biological mechanism point of view, TXA's limited visualization improvement could be due to non-hemostatic factors that require complementary techniques. TXA functions by inhibiting plasminogen activation to plasmin, reducing fibrin degradation and blood loss [34]. However, visual clarity depends on multiple factors, including irrigation fluid composition and pressure management [35] and surgical manipulation. One study reported that, although TXA reduces bleeding, epinephrine's vasoconstriction properties can more effectively manage intraoperative bleeding by reducing blood flow, improving visual clarity [19].

In our study, postoperative pain improved in both the TXA and control (saline) groups, with no statistically or clinically meaningful differences between interventions. Although patients receiving TXA demonstrated marginally lower VAS scores at 8 and 24 hours (0.5–0.7 points), that reduction is well below the established minimal clinically important difference (MCID) for postoperative pain after arthroscopic rotator cuff repair (>1.5 points). These findings indicate that, despite minor numerical differences, TXA does not provide a clinically relevant advantage in terms of patient-reported pain relief [36,37].

Previous studies have yielded conflicting results for this outcome. Whereas some meta-analyses suggest that TXA could shorten operation time and reduce postoperative shoulder pain, even within the first 24 hours [13,14,31,32], others have found no meaningful effect [9,10]. For instance, Alzobi et al. [14], reported only a modest reduction in pain, the clinical significance of which remained questionable. More importantly, substantial heterogeneity in study design, timing, and methods of pain assessment weakened the reliability of those pooled effects. MacKenzie et al. [24] evaluated pain at different postoperative time points, and Takahashi et al. [27] assessed pain across multiple contexts (at rest, activity, at night). In contrast, the meta-analysis by Hurley et al. [32] considered only resting pain, neglecting potentially relevant dimensions of patient experience. Moreover, the meta-analysis by Hurley et al. [32] incorporated data from studies with inconsistent protocols, including VAS scores obtained on day 0 and from patients who did not receive the allocated intervention (e.g., 24), further compromising interpretability. Similarly, Jiang et al. [22] compared TXA alone and in combination with dexamethasone against a control group and found no clear benefit of either treatment. Bayram et al. [19], who compared TXA with epinephrine, did not evaluate postoperative pain, limiting the scope of their findings. These methodological shortcomings underline the fragility of the current evidence base and highlight the need for more rigorously designed trials to clarify the analgesic role of TXA in arthroscopic rotator cuff repair.

By focusing exclusively on arthroscopic rotator cuff repair and using consistent pain assessment time points, our study avoids many of the methodological inconsistencies seen in prior research, offering a more targeted evaluation of TXA’s effect on postoperative pain. Even when statistical significance can be observed with the use of TXA in arthroscopic rotator cuff repair, the difference is likely clinically meaningless because the MCID was not reached.

Regarding functional outcomes, the pooled effect size revealed no significant difference in postoperative ASES scores between the TXA and control groups. The wide confidence intervals and not significant p-values reflect substantial uncertainty in the effect estimate, likely influenced by the small sample size and limited patient follow-up. Importantly, this endpoint has not been evaluated in prior meta-analyses due to the paucity of trials reporting functional scores, which emphasizes the limited evidence base that precludes any definitive conclusion about the functional impact of TXA in shoulder arthroscopy.

Most studies reported no significant differences in operation time, suggesting that TXA has no substantial effect on surgical duration except under specific conditions. However, Bildik et al. [20], observed a significant reduction of approximately 12 minutes in operation time with intra-articular administration of TXA. In contrast with intravenous administration, intra-articular TXA could exert localized hemostatic effects, reducing intraoperative bleeding and enhancing visualization, which could contribute to a more efficient surgical process. Nonetheless, differences related to the route of administration warrant further investigation because the surgical technique, surgeon experience, and institutional protocols could act as confounding factors. A more detailed sub-analysis comparing administration routes was not feasible in this study due to the heterogeneity of visualization assessment and the small number of studies that evaluated the intra-articular approach.

Two previous meta-analyses [13,32] reported a significant reduction in operation time with TXA use in arthroscopic rotator cuff repair. Both those analyses included the same five trials, which likely explains their identical findings. In contrast, our review incorporated eight studies—the same five plus three additional trials. Among them, only Bildik et al. [20] demonstrated a significant reduction in operation time with TXA. The remaining studies, including Bayram et al. [19], which compared TXA with epinephrine, found no meaningful difference in surgical duration between groups. Furthermore, a broader meta-analysis including various shoulder procedures (arthroscopic, open, and arthroplasty) reported reduced blood loss with TXA use [10], which could indirectly shorten operation time. However, that effect appears less consistent when analyses are limited to arthroscopic rotator cuff repair.

One of the major concerns during shoulder surgery performed in the beach chair position is the risk of cerebral hypoperfusion. Current recommendations advise maintaining systolic blood pressure above 90 mmHg and limiting reductions in systolic pressure and MAP to less than 20% from baseline to minimize that risk [38]. As noted in this study, there is no universally preferred patient position for shoulder arthroscopy; the choice is generally based on the surgeon’s preferences. The use of TXA has not been shown to affect MAP; all the studies included here reported values comparable to those in the control groups.

Several concerns are associated with the volume of fluid used in shoulder arthroscopy. These include hypothermia [39], fluid retention in the surrounding tissues (particularly when automated pump systems are used), a variable local inflammatory response [40], decreased hemoglobin levels, and alterations in serum sodium [41]. Although no standardized or recommended fluid volume has been established for shoulder arthroscopy, lower volumes are generally preferred to minimize those risks. Evidence indicating that TXA can significantly reduce fluid requirements remains inconclusive.

The use of TXA in shoulder surgery appears to be safe with respect to thromboembolic events [8]; no adverse events—general or thromboembolic—were reported across any of the studies we analyzed. This is consistent with the benefits of TXA established in other orthopedic procedures [42]. Specifically, in total shoulder arthroplasty, intravenous TXA has been associated with reduced perioperative blood loss and drainage, as well as decreased early postoperative pain [8]. In arthroscopic shoulder surgery, its use has shown promise in reducing hemarthrosis and early postoperative complications compared with controls [33].

Several limitations of this review must be acknowledged. Heterogeneity in outcome assessment, particularly for visual clarity, disallowed meaningful data pooling and meta-analysis in some cases. Functional outcomes were underreported and, where they were available, analysis was limited by small sample sizes and wide confidence intervals. The overall power of the included studies might have been insufficient to detect modest effects, particularly for secondary outcomes. Comparisons with epinephrine—a commonly used agent—were limited to a single study, restricting conclusions on the relative efficacy of TXA. Because epinephrine showed better visualization than TXA, it is necessary to conduct additional RCTs to determine the true superiority between them. Additional variability in TXA administration (e.g., dosage, route, and timing), surgical techniques, and perioperative protocols across studies might have introduced bias and reduced the generalizability of the findings. A limited comparison of functional scores was observed; only two studies included this evaluation, and they both had short follow-up times. Therefore, a long-term evaluation of large-scale studies with sufficient power is needed to determine whether TXA influences shoulder functionality. To overcome those limitations, we focused exclusively on arthroscopic rotator cuff repair, providing a more targeted and clinically relevant analysis than previous reviews that considered a broader range of shoulder procedures. We also included a greater number of studies than earlier meta-analyses, allowing a more comprehensive evaluation of TXA effects. Finally, we critically addressed methodological inconsistencies in prior research, such as inconsistent pain assessment and data extraction errors.

CONCLUSIONS

Current evidence is insufficient to support the routine use of TXA in arthroscopic rotator cuff repair, so it should be considered only in specific clinical scenarios. Across key outcomes—visual clarity, postoperative pain, operation time, functional recovery, MAP, and fluid use—TXA showed no significant advantage over the control. Although intra-articular TXA might offer localized effects, evidence about its effects on surgical efficiency and recovery remains inconclusive. Given its favorable safety profile, TXA might be a valuable option for specific high-risk patient populations (e.g., patients on anticoagulants or revision surgery) in which the risk of bleeding is a serious concern. Future studies should focus on determining the optimal dose, administration route, and efficacy of TXA in specific patient subgroups.

Supplementary materials

Article information Continued