Efficacy and Safety of Tranexamic Acid in Pediatric Trauma: A Systematic Review and Meta-Analysis

Abstract

OBJECTIVE

Tranexamic acid (TXA) is a widely used antifibrinolytic agent in surgical and trauma settings in adults. This study aimed to evaluate the efficacy and safety of TXA in pediatric trauma patients across various clinical outcomes.

METHODS

A comprehensive literature search was conducted across 4 databases. We included clinical trials and observational studies that reported the use of TXA in pediatric trauma patients (aged ≤18 years). Data extraction and risk-of-bias assessment were performed by independent reviewers. Meta-analyses were conducted with RStudio software.

RESULTS

A total of 12 studies (2 randomized controlled trials [RCTs] and 10 observational) involving 66,398 pediatric trauma patients were included. Tranexamic acid was not significantly associated with reduction in hospital mortality (OR = 1.06; 95% CI, 0.32–3.45) but was associated with significantly shorter hospital stays (mean difference [MD] = −1.49; 95% CI, −2.43 to −0.56). The need for emergency mechanical ventilation was higher among the TXA group (OR = 4.29; 95% CI, 2.52–7.31), whereas the need for mechanical ventilation at discharge was lower (OR = 0.23; 95% CI, 0.08–0.64). Tranexamic acid use did not significantly alter the risk of thromboembolic events (OR = 0.72; 95% CI, 0.19–2.79) or poor neurological outcomes (OR = 2.51; 95% CI, 0.86–7.35).

CONCLUSION

Tranexamic acid may reduce hospital length of stay in pediatric trauma patients, with inconsistent effects on mortality and adverse events. Its use should be individualized based on injury severity and resource availability. Further high-quality research is needed to confirm these findings and clarify the role of TXA in pediatric trauma care.

Introduction

Trauma, particularly unintentional injuries, is one of the leading causes of morbidity and mortality in children and adolescents worldwide.^1,2 The World Health Organization estimated that approximately 830,000 children (<18 years) die yearly from accidents worldwide.^3–6

Uncontrolled hemorrhage is the most common preventable cause of mortality in trauma patients across both adult and pediatric populations.^7,8 Among various interventions aimed at reducing trauma-related mortality, increasing attention has been directed toward pharmacologic hemostatic agents, most notably tranexamic acid (TXA).

Tranexamic acid is a synthetic antifibrinolytic agent.⁹ Its administration has been shown to confer a survival benefit in trauma-related bleeding in the past 10 years. Its global use was widely adopted following the Clinical Randomization of an Antifibrinolytic in Significant Hemorrhage 2 trial, initially in trauma centers and then expanded to prehospital settings.¹⁰ The Clinical Randomization of an Antifibrinolytic in Significant Hemorrhage 2 trial showed a significant reduction in mortality when TXA was given within 3 hours of injury, and its large, international design influenced global trauma guidelines and led to widespread adoption. Subsequent studies have demonstrated its benefit in severely injured victims experiencing hemorrhagic shock requiring massive transfusions.¹¹ It is considered a safe antifibrinolytic medication for adults, and early use is associated with decreased mortality in cases of major trauma.¹²

Regarding the use of TXA in pediatric trauma, professional societies have adopted varying positions. Although neither the Canadian Pediatric Society nor the American Academy of Pediatrics has issued an official declaration, the Royal College and Child Health in the United Kingdom issued a position statement in 2012 advocating for the use of TXA in pediatric trauma.¹³ This recommendation was based on evidence from pediatric surgical settings (e.g., orthopedic, craniofacial, cardiac surgeries), where TXA showed efficacy and safety.^14,15

A meta-analysis published in 2021 concluded that TXA use in pediatric trauma is associated with a high risk of seizures and showed no significant impact on in-hospital survival.¹⁶ A subsequent meta-analysis by Kornelsen et al¹⁷ also reported no survival benefit, highlighting the ongoing uncertainty. Therefore, we conducted this systematic review and meta-analysis to provide robust evidence regarding the use of TXA in pediatric trauma.

Methods

This review adhered to the recommendations of the Preferred Reporting Items for Systematic Review and Meta-Analyses checklist and was registered in International Prospective Register of Systematic Reviews with the University of York (registration number CRD420251017372).¹⁸

Literature Search.

A comprehensive search was conducted through the following electronic databases: PubMed, Scopus, Cochrane Library, and Web of Science from inception to February 14, 2025, using the following search strategy:

(((“tranexamic acid” OR “TXA” OR “antifibrinolytic agent” OR “Lysteda” OR “Cyklokapron”)) AND ((“pediatric” OR “child” OR “adolescent” OR “newborn” OR “neonate” OR “infant” OR “toddler” OR “young patient” OR “pediatric population” OR “preschool”))) AND ((“trauma” OR “injury” OR “hemorrhage” OR “bleeding” OR “hemorrhagic shock” OR “blunt trauma” OR “penetrating trauma” OR “polytrauma”)).

Then, references were exported, using EndNote 20 software, into a local library. Following this, manual searching was conducted to ensure that we included all relevant publications. The search strategy is provided in Supplemental Table S3.

Selection Criteria and Title and Abstract Screening.

Clinical trials and observational studies that reported the use of TXA in pediatric populations (<18 years) were included. There were no restrictions on the inclusion studies regarding year of publication, place, language, or patient demographics.

Studies with unreliable extracted data, book chapters, conference papers, reviews, theses, posters, editorials, letters, abstract-only articles, and those involving patients older than 18 years were excluded.

The screening step was performed using Rayyan. Any disagreements regarding study inclusion during screening were resolved through discussion, with input from a third reviewer when necessary. Following this, full-text articles were assessed to determine their eligibility for data extraction.

Data Extraction.

Two independent reviewers extracted data from eligible studies, using a standardized data extraction form. Discrepancies were resolved through discussion or consultation with a third reviewer. Extracted information included study characteristics; author name, year of publication, study design, country, and whether the study was single-arm or comparative; population characteristics: total sample size, number of participants in each group (TXA vs control), age, sex distribution, and baseline severity of trauma (including Injury Severity Score [ISS] where available); intervention details: description of the TXA intervention, including route, dose, and timing where reported, and description of the comparator or standard care; and outcomes: clinical endpoints such as mortality, seizure incidence, Glasgow Outcome Scale, need for mechanical ventilation at discharge, and discharge status. Outcomes were extracted as raw event counts, percentages, or means ± SD as reported.

Risk of Bias Assessment.

The quality of the included studies was assessed by 2 reviewers using the Cochrane Risk of Bias-2 tool for randomized controlled trials (RCTs) and risk of bias in non-randomized studies – of interventions for observational studies.^19,20 Under the Risk of Bias-2, domains such as bias in the randomization process, deviation from intended interventions, missing outcomes, measurement of outcome, and selection of reported results were evaluated. Each study was rated as having low, some concerns, or high risk of bias. Using the risk of bias in non-randomized studies – of interventions , studies were analyzed under 7 applicable domains and graded as “serious,” “moderate,” and “low” risk of bias. Discrepancies and disagreements were resolved through a consultation with the senior author (MA).

Statistical Analysis.

RStudio software (2024.04) was used to conduct the meta-analysis using the meta package. Pooled mean differences (MDs) were calculated for continuous outcomes, and ORs were computed for dichotomous outcomes. Pooled proportions were also calculated when applicable. All these measures were conducted with corresponding 95% CI. Statistical significance was defined as p < 0.05. The I² statistic and the Cochran Q test were used to evaluate heterogeneity. A low I² value was 25%, a moderate value was 50%, and a high heterogeneity value was 75% or more. In outcomes with 3 or more studies showing significant heterogeneity (p < 0.05), a leave-one-out sensitivity analysis was used to investigate the effects of each individual study on the total effect size and heterogeneity. Subgroup analyses were performed by study design (RCT vs non-RCT) and setting (combat vs civilian) when applicable. All analyses used a random-effects model; unless heterogeneity was low, a fixed-effect model was considered.

Results

Screening and Study Selection.

A total of 4208 records were obtained on initial search, of which 2929 were finally screened after duplicates were removed. On title and abstract screening, 2892 articles were excluded because they did not meet our inclusion criteria. Of the remaining 37 articles, 22 were excluded at the full-text screening stage: 18 for irrelevance and 7 for assessing adult populations. Ultimately, 12 were finally included in our review (Figure 1).^21–35

Figure 1.

PRISMA flowchart of included studies.

Study Characteristics.

A total of 12 studies (2 RCTs and 10 observational) involving 66,398 pediatric trauma patients were included, spanning from 1981–2025. The studies were conducted across multiple continents, including North America (e.g., the United States and Canada), Asia (e.g., Afghanistan), and Europe (e.g., Finland), reflecting a wide geographic distribution. Sample sizes ranged from 48 to 61,779. Common injuries included traumatic brain injury (TBI), blunt trauma (e.g., head, limbs), polytrauma, and hyphema. Of the included studies, 3 focused on combat trauma, representing injuries sustained in military or war-related settings (e.g., gunshot wounds, shrapnel injuries); 7 focused on civilian trauma, representing injuries sustained in noncombat or community settings (e.g., road traffic accidents, falls, civilian shootings, or stabbings); and 2 focused on mixed-trauma populations.

Follow-up periods varied; most lasted until hospital discharge, though some extended up to 6 months. Tranexamic acid administration protocols varied in dosage (15–30 mg/kg, up to 1 g) and delivery method, including intravenous (IV) bolus and oral administration. A summary of included studies and baseline characteristics are detailed in Tables 1 and 2.

Table 1.

Summary of Included Studies

ID	Study Design/Phase	Country	Sample Size	Type of Trauma	Inclusion Criteria	Exclusion Criteria	Outcomes
Utsumi 202521	Multicenter, retrospective observational study	USA	368	Severe pediatric TBI	1. Pediatric patients (<18 yr) 2. Severe TBI with GCS score 3–8 on arrival	1. Cardiac arrest (SBP = 0 mm Hg) before or on arrival 2. Head AIS score of 6 or 9 (unsurvivable or unspecified head injury) 3. No intracranial hemorrhage (e.g., concussion, cranial nerve injury, scalp contusion)	1. Adverse effects (e.g., contusion, diffuse brain injury) 2. Requirement of surgical intervention 3. Poor neurological outcomes (GOS 1–3) 4. Blood transfusion 5. In-hospital mortality 6. Mechanical ventilation
Iqbal 202332	RCT	Afghanistan (Camp Bastion NATO Role 3 hospital)	260	Major pediatric trauma (blunt and penetrating)	Children with major trauma (ISS 14–49)	1. History of allergic reaction to TXA 2. Preexisting thrombosis or embolism ISS >49 3. Patients requiring immediate surgical exploration	1. Requirement of surgical intervention 2. Blood transfusion 3. In-hospital mortality 4. Length of hospital stay 5. Discharge rates
Gendler 202322	Retrospective cohort study	Canada	911	Pediatric trauma (penetrating and blunt injuries, military and civilian cases)	Pediatric trauma patients (<18 yr old) treated by IDF-MC prehospital teams	Patients pronounced dead at the scene	1. Requirement of surgical intervention 2. In-hospital mortality
Nishijima 202231	Double-blind RCT	United States	31	Hemorrhagic brain and/or torso injuries	1. Children younger than 18 yr with hemorrhagic injuries to the brain and/or torso GCS 3–13 with intracranial hemorrhage (for head trauma) 2. Clinical or radiographic evidence of internal hemorrhage (for torso trauma	1. Inability to administer TXA within 3 hr of injury 2. Known pregnancy, prisoners, or wards of the state 3. Cardiac arrest before randomization 4. GCS 3 with bilateral unresponsive pupils Isolated subarachnoid hemorrhage, epidural hematoma, or diffuse axonal injury 5. Prior TXA administration for current injury	1. Thromboembolism 2. Blood transfusion 3. In-hospital mortality 4. Length of hospital stay 5. Seizures 6. FFP volume 7. Platelet volume 8. PedsQL score 9. CT hemorrhage progression
Gossiome 202228	Multicenter, retrospective study	Finland	934	Road collision and blunt trauma	1. Patients younger than 18 yr 2. Major trauma cases requiring prehospital intervention by an MMT	1. No prehospital intervention by an MMT 2. Missing data on TXA administration	1. Requirement of surgical intervention 2. Poor neurological outcomes (GOS 1–3) 3. In-hospital mortality 4. ICU admission 5. pRBCs
Hamele 202023	Retrospective cohort study	Afghanistan (Camp Bastion NATO Role 3 hospital)	507	Pediatric combat trauma requiring massive transfusions	1. Patients younger than 18 yr 2. Sustained trauma in a combat setting (Iraq and Afghanistan) 3. Required massive transfusion (≥40 mL/kg of blood products within 24 hr of injury)	1. Patients with fatal head injuries 2. Patients with burn injuries	1. In-hospital mortality 2. Total blood products 3. FFP volume 4. Platelet volume 5. Resource use indicators including in-hospital–, ICU-, and ventilation-free days
Thomson 201924	Retrospective chart review	Canada	48	Pediatric trauma (blunt and penetrating)	1. Pediatric patients (≤16 yr old) 2. Trauma patients identified using an MTP order	1. Patients who received TXA at an outside hospital without adequate documentation 2. Those who did not receive MTP were excluded so as to prevent false inclusion of patients who received TXA for nontraumatic indications	1. Thromboembolism 2. Requirement of surgical intervention 3. In-hospital mortality 4. Length of hospital stay 5. Total blood products 6. pRBC:FFP
Maeda 201829	Retrospective cohort study using a nationwide inpatient database	Finland	61,779	Pediatric trauma (various types: head, torso, extremities, multiple injuries)	1. Pediatric patients ≤12 yr old, diagnosed with trauma (6 defined trauma sites: head/neck, thorax, torso, extremities, multiple injuries, unspecified regions) 2. Patients who received a blood transfusion (packed RBCs, FFP, platelets, autologous blood transfusion)	1. Patients with missing data on baseline characteristics 2. Patients who received TXA on the second hospital day or later	1. Thromboembolism 2. In-hospital mortality 3. Seizures 4. Renal replacement therapy
Eckert 201425	Retrospective cohort study	Afghanistan (Camp Bastion NATO Role 3 hospital)	766	Pediatric combat trauma, coalition and civilian trauma casualty	Pediatric trauma patients (≤18 yr) admitted to the combat hospital	NA	1. Requirement of surgical intervention 2. Blood transfusion 3. In-hospital mortality 4. GCS 5. Mechanical ventilation
Albiani 200826	Retrospective observational study	Canada	215	Pediatric traumatic hyphema (ocular trauma)	1. Children diagnosed with traumatic hyphema due to blunt force trauma 2. Patients treated at the Children’s Hospital of Eastern Ontario or Queen’s Department of Ophthalmology Emergency Eye Clinic	Patients with missing treatment data	1. Rebleeding
Uusitalo 198827	Retrospective and prospective comparative study	Finland	340	Non-perforating traumatic hyphema in children	1. Children with non-perforating traumatic hyphema 2. Younger than 16 yr	NA	1. Discharge rates
Uusitalo 198130	Retrospective comparative study	Finland	239	Traumatic hyphema (non-penetrating ocular trauma)	1. Patients with traumatic hyphema admitted to the Department of Ophthalmology, Tampere Central Hospital	Patients with only a hemorrhagic aqueous flare Patients with penetrating ocular trauma	1. Discharge rates

AIS, Abbreviated Injury Scale; CRASH-3, Clinical Randomization of an Antifibrinolytic in Significant Hemorrhage 3 trial; CT, computed tomography; FFP, fresh frozen plasma; GCS, Glasgow Coma Scale; GOS, Glasgow Outcome Scale; ICU, intensive care unit; IDF-MC, Israel Defense Force Medical Corps; ISS, Injury Severity Score; MMT, mobile medical team; MTP, massive transfusion protocol; NA, not available; NATO, North Atlantic Treaty Organization; PedsQL, Pediatric Quality of Life Inventory; pRBC, packed red blood cell; RBCs, red blood cells; RCT, randomized controlled trial; SBP, systolic blood pressure; TBI, traumatic brain injury; TXA, tranexamic acid

Table 2.

Baseline Detail of Included Studies

Study	Age, yr		Female		Type of Injury		Baseline Injury Severity Score		Initial GCS
	I	C	I	C	I	C	I	C	I	C
	Mean ± SD	Mean ± SD	n (%)	n (%)	n (%)	n (%)	Mean ± SD	Mean ± SD	Mean ± SD	Mean ± SD
Gossiome, 202228	15 ± 2.98	11.33 ± 7.43	102 (29.8%)	193 (32.6%)	323 (94.4%) blunt 19 (5.6%) penetrating	577 (97.6%) blunt 14 (2.4%) penetrating	15.5 ± 11.85	6.75 ± 8.15	Events 49 ± 16.2 of GCS <8	Events 35 ± 6.8 of GCS <8
Eckert, 201425	11 ± 4	11 ± 5	14% (9/66)	12% (83/700)	91% (60/66) penetrating trauma	72% (504/700) penetrating trauma	18 ± 13	10 ± 9	8 ± 6	13 ± 5
Utsumi 202521	12.33 ± 5.92	10.00 ± 6.86	31% (42/137)	30% (69/231)	Blunt injury, n (%) = 135 (98) Injury type, —– Falling from height 26 (19) Falling at ground level 3 (2) Falling on stairs 13 (9) Motor vehicle crash 12 (9) Motorcycle 25 (18) Pedestrian 21 (15) Bicycle 29 (21)	Blunt injury, n (%) = 224 (97) Injury type,—– n (%). Falling from height 30 (13) Falling at ground level 15 (6) Falling on stairs 23 (10) Motor vehicle crash 25 (10) Motorcycle 21 (9) Pedestrian 25 (11) Bicycle 57 (25)	24 ± 9.63	17.67 ± 11.85	5.33 ± 2.96	5.33 ± 2.96
Thomson 201924	14.00 ± 2.96	9.67 ± 7.41	6 (20.7%)	8 (42.1%)	76% blunt trauma in TXA group	84% blunt trauma in no-TXA group	32.0 ± 16.30	33.75 ± 17.04	3.75 ± 2.22	6.25 ± 8.15
Maeda 201829	6.75 ± 3.70	6.75 ± 3.70	662 females (34.6%)	662 females (34.6%)	NA	NA	NA	NA	NA	NA
Iqbal 202332	8.29 ± 2.46	8.20 ± 2.62	41 (31.54%)	54 (41.5%)	NA	NA	26.61 ± 11.17	29.00 ± 11.64	NA	NA
Hamele 202026	≈9.5 ± 1.734	8.75 ± ≈1.171	11 females (19%)	121 (27%)	Blunt: 2 (3.4) Blast: 48 (83) Penetrating: 8 (13.6)	Blunt: 49 (10.9) Blast: 255 (57) Penetrating: 144 (32)	18.0 ± 8.89	17.75 ± 11.11	8.0 ± 2.6002	10.0 ± 2.007
Albiani 200826	12.6 ± 0.43	10.7 ± 0.55 (p = 0.017)	26 females (19.2)	19 females (24.1)	NA	N/A	NA	NA	NA	NA
Gendler 202322 AFTER PROPENSITY MATCHED SCORING	13.2 ± 3.38	13.4 ± 3.70 (p = 0.356)	13 (18.6)	15 (20.3)	N/A	N/A	N/A	N/A	N/A	N/A
Nishijima, 2022 PILOT study31	TXA 15 mg/kg: 12.6 ± 4.4	9.2 ± 6.1	TXA 15 mg/kg: 4 (44)	3 (25)	TXA 15 mg/kg: TBI: 4 (44), Torso: 5 (56)	TBI: 7 (58) Torso: 5 (42)	TXA 15 mg/kg: 15.1 ± 5.8	13.8 ± 5.7	Brain injury GCS: 8.5 ± 7.41 Torso injury GCS: 13.5 ± 4.44	NA
	TXA 30 mg/kg: 10.7 ± 4		TXA 30 mg/kg: 8 ± 20		TXA 30 mg/kg: TBI: 5 (50), Torso: 5 (50)		TXA 30 mg/kg: 15.2 ± 5.7			NA
Uusitalo 198827	TXA + bed rest: 10.8 ± 2.6 TXA + ambulation: 9.4 ± 3.1	9.7 ± 3.3	TXA + bed rest: 5 (19.2) TXA + ambulation: 17 (17.9)	37 (16.9)	NA	NA	NA	NA	NA	NA
Uusitalo 198130	NA	NA	40 of 239 patients (∼16.7)		NA	NA	NA	NA	NA	NA

C, control; GCS, Glasgow Coma Scale; GSW, gun-shot wound; I, intervention; N/, not available; RTC, road traffic collision; TBI, traumatic brain injury; TXA, tranexamic acid

Risk of Bias.

The risk of bias assessment included 10 observational studies and 2 RCTs. Among the observational studies, 7 studies had a serious risk of bias,^21–28 while the remaining 3 showed a moderate risk.^29–33 Confounding and participant selection bias were serious in 50% of the studies and affected the most domains. Regarding the RCTs, 1 showed a low risk of bias,³⁴ while the other raised some concerns due to missing outcome data³⁵ (Supplemental Figure S1). Summary of findings for primary outcomes, including effect estimates and GRADE certainty, is shown in Figure 2. GRADE assessment is provided in Table 3.

Figure 2.

Summary of findings for primary outcomes, including effect estimates and GRADE certainty.

Table 3.

GRADE Summary of Findings*

Outcome	Anticipated Absolute Effects (95% CI)		Relative Effect (95% CI)	Participants, n (Studies)	Certainty of the Evidence/GRADE	Comments
	TXA Administration	Placebo and Standard Care or Standard Care Alone
Mortality	91 per 1000 (29–246)	87 per 1000	OR 1.06 (0.32–3.45)	7528 (8)	Very low ⊕⊝⊝⊝	Downgraded owing to inconsistency, serious risk of bias, and imprecise findings (wide 95% CI). We are uncertain about the effect of TXA in increasing mortality rate.
Blood transfusion	695 per 1000 (314–919)	395 per 1000	OR 3.49 (0.70–17.38)	1452 (3)	Very low ⊕⊝⊝⊝	Downgraded owing to inconsistency, imprecise findings (wide 95% CI), and serious risk of bias in the selection of reported results. We are uncertain about the association between TXA and the requirement for blood transfusion.
Surgical intervention	493 per 1000 (247–742)	342 per 1000	OR 1.87 (0.63–5.53)	3189 (6)	Very low ⊕⊝⊝⊝	Downgraded owing to inconsistency, selection bias, reporting bias, and imprecise findings (wide 95% CI). We are uncertain that TXA may increase the need for surgical intervention.
Poor neurological outcomes (GOS 1–3)	187 per 1000 (73–402)	84 per 1000	OR 2.51 (0.86–7.35)	934 (2)	Very low ⊕⊝⊝⊝	Downgraded owing to inconsistency, selection bias, missing data, and imprecise findings (wide 95% CI). We are uncertain about the association between TXA and the poor neurological outcomes.
Thromboembolic events	2 per 1000 (0.5–7)	3 per 1000	OR 0.72 (0.19–2.79)	3859 (3)	Low ⊕⊕⊝⊝	Downgraded owing to serious risk of bias in the selection of reported results and participants. Imprecise findings (wide 95% CI) were reported. Our confidence about the effect of TXA in thromboembolic event was limited.
Seizure	1 per 1000 (0–63)	0.5 per 1000	OR 1.86 (0.03–130.14)	3859 (2)	Low ⊕⊕⊝⊝	Downgraded owing to inconsistent and very serious imprecise findings (very wide 95% CI). Our confidence about the effect of TXA in the seizure events was limited.
Length of hospital stays, days	The TXA group’s hospital stay was 1.49 lower (2.43 lower to 0.56 lower) than non-TXA		MD 1.49 (−2.43 to −0.56)	339 (3)	Moderate ⊕⊕⊕⊝	Downgraded owing to serious risk of bias in the selection of reported results and participants. We are moderately certain that TXA decreases significantly the length of hospital stays. The findings were consistent and had low risk of bias.
Volume of total blood products, mL/kg	The TXA group’s use of total blood products was 15.62 higher (2.15 higher to 29.09 higher) than non-TXA		MD 15.62 (2.15–29.09)	555 (2)	Low ⊕⊕⊝⊝	Downgraded owing to serious risk of individual study limitations, ineffective randomization, and selection bias. Our confidence about the effect of TXA in the volume of total blood products used was limited.
Volume of FFP, mL/kg	The TXA group’s FFP use was 6.47 higher (77.30 lower to 90.23 higher) than non-TXA		MD 6.47 (−77.30 to 90.23)	538 (2)	Very low ⊕⊝⊝⊝	Downgraded owing to inconsistent and very serious imprecise findings (very wide 95% CI). We are uncertain about the effect of TXA in the volume of FFP used.
Volume of platelets, mL/kg	The TXA group’s platelet use was 15.62 higher (2.15 higher to 29.09 higher) than non-TXA		MD 4.43 (−38.73 to 47.60)	555 (2)	Very low ⊕⊝⊝⊝	Downgraded owing to serious risk of bias in the selection of reported results and participants, serious risk of confounding was reported as well. Also, inconsistent and imprecise findings (very wide 95% CI) were reported. We are uncertain about the effect of TXA in volume of platelets used.

FFP, fresh frozen plasma; GOS, Glasgow Outcome Scale; ICU, intensive care unit; TXA, tranexamic acid

* Population: pediatric patients with trauma; intervention: TXA administration; comparator: non-TXA care protocol.

Clinical Outcomes.

Hospital mortality results in slight increase in TXA group compared with non-TXA (OR = 1.06; 95% CI, 0.32–3.45; I² = 88.8%; 8 studies; 92/2733 TXA vs 415/4795 non-TXA; very low certainty), indicating 5 per 1000 more mortality events (from 29 fewer to 246 more), and a leave-one-out analysis confirmed robustness of this null result (Figure 3; Supplemental Figure S4).^{1–4,6,10–12,15} Of the 9 studies reporting on hospital mortality, 7 were observational and revealed a null effect. Two were RCTs; 1 favored the TXA group (OR = 0.19; 95% CI, 0.04–0.87), while the other did not report mortality outcomes in either group, but involved a small sample size (TXA = 19 vs non-TXA = 12) (Supplemental Figure S2). Furthermore, the subgroup analysis by setting (e.g., combat vs civilian) revealed no significant differences (Supplemental Figure S2).

Figure 3.

Forest plots of effectiveness outcomes: (A) mortality, (B) blood transfusion, (C) surgical intervention, and (D) poor neurological outcome.

Two studies reported complete discharge rates in both arms.^28,33 The pooled data revealed significantly shorter hospital stays in the TXA group (MD = −1.49; 95% CI, −2.43 to −0.56; I² = 9.6%; 3 studies; 339 participants; moderate certainty), which was the most consistent and clinically relevant benefit (Figure 5).^24,34,35 In Gossiome et al,³⁰ the intensive care unit (ICU) admission rate among TXA recipients was 51.7% (164/317). Resource use indicators, including in-hospital, ICU, and ventilation-free days, were reported by Hamele et al²³ and showed no significant increases.

Figure 5.

Forest plots of resource use outcomes: (A) hospital length of stay, (B) volume of blood transfusion, (C) volume of FFP, and (D) volume of platelets.

Tranexamic acid administration results in little to no difference in thromboembolic events, compared with non-TXA group (OR = 0.72; 95% CI, 0.19–2.79; I² = 0%; 3 studies; 5/1962 TXA vs 5/1945 non-TXA; very low certainty), indicating 1 per 1000 fewer thromboembolic occurrences in the TXA group (from 0.5 to 7), with consistent findings (Figure 4).^24,32,34 Nishijima et al³⁴ reported lower to no difference in computed tomography hemorrhage progression in pediatric patients administered TXA (OR = 0.52; 95% CI, 0.12–2.25; 1 study; 8/19 TXA vs 7/12 non-TXA), resulting in 162 per 1000 fewer events (from 144 to 759). Furthermore, the TXA group showed little to no significant difference in adverse events, including contusion, diffuse brain injury, and rebleeding (OR = 1.27; 95% CI, 0.89–1.81; I² = 6.6%; 2 studies; 2/411 TXA vs 99/540 non-TXA) (Figure 4).^21,27 These findings suggest 40 per 1000 more adverse events in patients who received TXA than in those who did not receive TXA (from 166 to 289).

Figure 4.

Forest plots of safety outcomes: (A) thromboembolism, (B) seizures, and (C) other adverse events.

Tranexamic acid may result in more to no difference in additional surgical intervention to control post pediatric trauma bleeding when compared with the non-TXA group (OR = 1.87; 95% CI, 0.63–5.53; I² = 95.1%; 6 studies; 332/748 TXA vs 831/2431 non-TXA; very low certainty), resulting in 151 per 1000 more surgical interventions in the TXA group (from 247 to 742) (Figure 3).^{21,22,24,26,30,35} However, combat settings were associated with more need for surgical intervention in the TXA group (OR = 7.36; 95% CI, 2.83–19.19; I² = 69.7%; 2 studies; 96/136 vs 591/1541), indicating 437 per 1000 more interventions among patients receiving TXA in the combat settings (638 to 923) (Supplemental Figure S3). Civilian settings result in more to no difference in the TXA group (OR = 1.21; 95% CI, 0.33–4.40; I² = 96.3%; 3 studies; 216/593 vs 224/871), with 35 per 1000 more patients requiring interventions among those receiving TXA in civilian settings (103 fewer to 604 more) (Supplemental Figure S3).

Utsumi et al²¹ reported that the requirement for emergency mechanical ventilation was higher in the TXA group than the non-TXA group (OR = 4.29; 95% CI, 2.52–7.31; p < 0.0001; 1 study; 116/137 TXA vs 130/231 non-TXA), with 284 per 1000 more instances of emergency mechanical ventilation in the TXA group (from 764 to 904). However, Eckert et al²⁶ found that at the time of discharge or transfer, a significantly lower percentage of patients in the TXA group still required mechanical ventilation (OR = 0.23; 95% CI, 0.08–0.64; p = 0.0049; 1 study; 4/66 TXA vs 154/700 non-TXA), with 159 per 1000 fewer patients receiving TXA requiring mechanical ventilation at time of discharge or transfer (from 22 to 153).

Patients who received TXA had more to no difference in poor neurological outcome rates, defined as a Glasgow Outcome Scale score of 1 to 3 (indicating death, persistent vegetative state, or severe disability) when compared with those who did not receive TXA (OR = 2.51; 95% CI, 0.86–7.35; I² = 85.2%; 2 studies; 74/479 TXA vs 69/823 non-TXA; very low certainty), resulting in 103 per 1000 more patients with poor neurological outcomes (from 73 to 402) (Figure 3C). This outcome was assessed in 2 different types of traumatic injuries; Utsumi et al²¹ included patients who had TBI, while Gossiome et al³⁰ considered patients with blunt trauma, including head trauma. An additional outcome, seizures (as detailed in Figure 4B), showed inconsistent and imprecise findings.^32,34

Nishijima et al³⁴ reported lower to no difference in Pediatric Quality of Life Inventory scores at 1 week (MD = −9.62; 95% CI, −28.55 to 9.31; 1 study; 31 participants), suggesting a 9.62 lower Pediatric Quality of Life Inventory score (from 28.55 to 9.31), which is a measure of health-related quality of life. Descriptive analysis of Glasgow Coma Scale (GCS) scores revealed that 86% (95% CI, 0.86–0.87; 1 study; 104,862 participants) had high GCS (14–15), and 7% (95% CI, 0.04–0.13; 1 study; 104,862 participants) had GCS less than 9 among TXA recipients.²⁹

Blood Transfusion and Blood Products.

The need for blood transfusion results in more to no difference in patients who received TXA (OR = 3.49; 95% CI, 0.70–17.38; I² = 90.3%; 3 studies; 271/352 TXA vs 424/1073 non-TXA; very low certainty) (Figure 3).^21,26,34 One study reported blood transfusion in all participants in both groups and was therefore excluded from the pooled analysis of this outcome, because the lack of variance prevented OR estimation.³⁵ The null effect was consistent among different study designs (Supplemental Figure S3I). However, combat settings demonstrated a significantly higher need for blood transfusion (OR = 14.98; 95% CI, 7.28–30.82; 1 study; 57/66 TXA vs 208/700 non-TXA), resulting in 566 per 1000 more instances of blood transfusion in the TXA group in combat settings (from 755 more to 929 more), with fewer to no difference in civilian settings (OR = 1.93; 95% CI, 0.76–4.93; I² = 37.7%; 2 studies; 214/286 TXA vs 216/373 non-TXA) (Supplemental Figure S3).

In addition, Gossiome et al³⁰ reported that packed red blood cell use was not significantly higher in the TXA group (OR = 7.61; 95% CI, 0.88–65.40; 1 study; 5/333 TXA vs 1/500 non-TXA), indicating 13 per 1000 more instances of packed red blood cell use in the TXA group (from 1 to 116). Further outcomes, including total blood products, platelets, and fresh frozen plasma volumes, did not reveal consistent group differences (Figure 5).^23,24,34 Thomson et al²⁴ showed that the packed red blood cell:fresh frozen plasma ratio was elevated in the TXA group (MD = 0.46; 95% CI, 0.19–0.73; 1 study; 48 participants).²⁴

Discussion

Tranexamic acid is an antifibrinolytic agent widely used in hemorrhagic trauma.³⁶ It acts by blocking lysine binding sites on plasminogen, thereby inhibiting its activation to plasmin.³⁷ Although its benefits are well established in adult trauma, evidence is still sparse in pediatric trauma.³⁸ There is a need to carefully balance the potential benefits and adverse effects of TXA in children, because their relatively lower circulating blood volume predisposes them to a poorer prognosis, with an earlier onset of tachycardia and rapid progression to hypotension.³⁹ Detecting massive blood loss in children may be difficult to assess, owing to the higher physiological reserve and thus less prominent clinical signs.⁴⁰ Therefore, a comprehensive assessment of TXA, considering patient outcomes and safety, is crucial for framing guidelines for the management of pediatric trauma.

Our study provides a comprehensive analysis of the efficacy of TXA in pediatric trauma, synthesizing data from 12 studies and 66,398 children. Regarding clinical outcomes, in-hospital mortality was comparable between the TXA and non-TXA groups. Subgroup analysis by trauma setting (civilian or combat) did not show a significant difference. However, TXA use was associated with a significantly shorter hospital stay, which emerged as the most consistent benefit across studies. Our meta-analysis provides the most recent evidence on the use of TXA, adding to the meta-analysis in 2022 by including 3 additional published studies.^21,22,35

Approximately 51% of children who received TXA required ICU admission. In terms of safety, no significant difference was found in thromboembolic events or computed tomography hemorrhage progression between groups. Although he rate of surgical intervention was nonsignificantly higher in the TXA group, the effect was particularly pronounced in combat settings. As for neurological outcomes, 86% of TXA recipients achieved a GCS score of 14 or higher, while 7% had scores below 9.

Regarding transfusion requirements, overall blood product use was similar between groups. However, a notably higher transfusion rate was observed in the TXA-treated patients within combat settings, suggesting that TXA was likely administered to those with greater injury severity.

Although our analysis revealed no statistically significant effect on mortality, some studies suggested a trend toward increased mortality in TXA recipients. This may reflect either adverse effects or inadequate hemostatic efficacy in the pediatric setting. A 2022 meta-analysis by Kornelsen et al¹⁷ conducted in children similarly reported no survival benefit (OR = 0.61; 95% CI, 0.30–1.22).

Tranexamic acid maintains the presence of the clot formed via the coagulation cascade, delaying its breakdown into fibrin degradation products. Although this reduces bleeding, elevated fibrin levels may cause thrombus formation in the blood vessels of end organs such as the kidney and retina. This results in arterial occlusion and necrosis.⁴¹ Thus, the outcome of hospital mortality reflects the efficacy of TXA in reducing bleeding by maintaining coagulation, coupled with fewer adverse events. The slightly higher mortality rate may indicate more incidences of adverse events, or a lack of enough efficacy in children to sufficiently stop blood loss. On the other hand, the higher mortality rate may imply a delay in the initiation of action of the drug. Zhang et al⁴² evaluated the use of TXA in pediatric surgeries and reported a significant reduction in blood loss only after 12 hours. This delayed effect may limit the usefulness of IV-TXA in children experiencing rapid hemodynamic decompensation, where earlier control of bleeding is critical.⁴³

Another factor that may account for our findings is that TXA is an antifibrinolytic rather than a procoagulant. Although it can prevent the breakdown of existing clots, it does not initiate or enhance the coagulation cascade.⁴⁴ Thus, in children with significant coagulopathy, TXA alone may be insufficient to control bleeding. In pediatric patients, differences in pharmacokinetic parameters, such as faster clearance and variable distribution volumes, may influence how quickly and effectively TXA exerts its antifibrinolytic action in trauma settings.⁴⁴ Our analysis also demonstrated significant heterogeneity that was not resolved by sensitivity analysis or subgrouping based on trauma setting, which may reflect variability in patient characteristics, injury severity, and dosing protocols.

The variability in dosing protocols may also affect outcomes. Children older than 12 years typically receive adult-equivalent doses (1 g IV bolus followed by 1 g over 8 hours), while younger children are dosed by weight (15 mg/kg bolus followed by 2 mg/kg/hr infusion).⁴⁵ Additionally, the severity of trauma described as ISSs may significantly influence response to any medical treatment, and patients with a higher ISS may not completely respond to TXA administration.⁴⁶ Further, the use of concurrent medications, based on the protocol of the hospital, may significantly influence the overall mortality of the child; hospital mortality may not be necessarily due to the inefficacy of TXA, but may result from a delay in administering TXA or other crucial drugs like vasopressors, or inadequate doses used.⁴⁷

We found that TXA significantly reduced hospital stay and was associated with higher discharge rates, consistent with findings from multiple studies.^26,27 Shorter hospitalization may indicate earlier hemodynamic stabilization, which is especially valuable in combat environments, where medical resources are limited and rapid turnover is essential.⁴⁸ Nonetheless, these results should be interpreted cautiously owing to the small number of studies reporting this outcome.

Regarding the safety profile, use of TXA was associated with a slightly lower risk of thromboembolic events, although the difference was not statistically significant. This contrasts with findings from the previous meta-analysis by Kornelsen et al¹⁷ that reported a slightly higher risk of thromboembolic events with TXA use in pediatric trauma. This suggests that recent evidence has demonstrated a significantly reduced risk of thromboembolic events. Studies conducted in children undergoing surgical procedures have also concluded that there is no significant increase in thromboembolic risk.⁴⁹

Additionally, TXA is also associated with a lower risk of computed tomography hemorrhage progression, suggesting a potential benefit in limiting cerebral bleeding, particularly in TBI. However, this finding did not reach statistical significance. Overall risk of other adverse events, such as contusion and brain injury, was similar between TXA and non-TXA groups, indicating a generally favorable safety profile in the pediatric group.⁵⁰

The use of TXA did not significantly alter the overall rate of surgical interventions; subgroup analyses showed a marked increase in surgical procedures among the TXA group in combat settings. This likely reflects confounding by indication, as TXA may have been preferentially administered to more severely injured children in combat environments. These patients are more likely to require surgical interventions owing to the nature and severity of their injuries. Tranexamic acid might have been used earlier and more aggressively in these high-risk patients, which can lead to a bias.⁵¹

Combat zones may operate under different triage and intervention protocols, where surgical care is more readily provided to patients stabilized with TXA. In resource-limited or austere settings, rapid surgical stabilization may be the next step once TXA helps control bleeding. Tranexamic acid may improve early survival, allowing more patients to reach surgery, which can paradoxically increase surgical rates.²⁶

Regarding the neurological outcomes, 2 included studies examined different trauma populations: Utsumi et al²¹ assessed patients with TBI, while Gossiome et al³⁰ focused on blunt trauma, not limited to the brain. As a result, pooled estimates may reflect clinical heterogeneity. Neurological outcomes after TXA are likely more relevant in TBI populations, where cerebral bleeding and intracranial pressure dynamics play a major role. In non-isolated TBI, poor neurological outcomes may be influenced by systemic injury, confounding attribution to TXA.⁵²

Seizures represent an important safety concern with TXA, particularly in the setting of TBI. Although seizures were infrequently reported in the included studies, TXA is known to lower the seizure threshold, especially at higher doses or when blood–brain barrier disruption allows greater central nervous system penetration.⁵³ This adverse effect has been described in pediatric surgical populations and adult trauma cohorts, and its relevance in children with TBI warrants further evaluation. Given that our findings showed that more TXA recipients achieved higher GCS scores, careful interpretation is required, as the potential for proconvulsant effects complicates any suggestion of neuroprotection. Future trials should systematically monitor and report seizure outcomes to better clarify the risk-benefit balance of TXA in pediatric trauma. Even in patients with potential brain trauma, preventing further hemorrhage may avert neurological deterioration.⁵⁴ Tranexamic acid might limit expansion of intracranial hemorrhage, reducing the number of patients progressing to severe neurological depression (GCS <9).⁵⁵

Our findings also suggested a higher need for blood transfusion during TXA use in combat settings. In combat zones, TXA is often given selectively to patients with life-threatening injuries, particularly hemorrhage, under military trauma protocols. This leads to confounding, in which patients receiving TXA are also the ones most likely to require transfusions, regardless of the intervention.⁵⁶ Additionally, in military settings, delays in surgical intervention or evacuation may necessitate greater transfusion volumes, even with TXA.^57,58

This meta-analysis has several notable strengths. First, it includes a large sample size (66,398 pediatric trauma patients) from 12 studies spanning multiple continents, enhancing the generalizability of findings across a diverse population. Including both RCTs and observational studies provides a comprehensive evaluation of TXA effects in pediatric trauma. Additionally, the assessment of multiple clinical outcomes—mortality, thromboembolic events, hospital stay, blood transfusion needs, and neurological outcomes—offers a comprehensive overview of the impact of TXA.

Our meta-analysis builds upon the systematic review by Kornelsen et al¹⁷ that evaluated the effectiveness and safety of TXA in pediatric trauma and found no overall survival benefit (OR = 0.61; 95% CI, 0.30–1.22), although a benefit was observed in combat settings, and no significant increase in thromboembolic events. By including 3 additional studies published after 2021, our analysis encompasses 12 studies and more than 66,000 pediatric patients, providing a more comprehensive evaluation. Consistent with Kornelsen et al,¹⁷ we observed no significant effect of TXA on overall mortality; however, we found a significant reduction in hospital stay among TXA recipients, suggesting earlier hemodynamic stabilization and faster recovery, particularly relevant in combat and resource-limited settings. In terms of safety, while both analyses report no increased risk of thromboembolic events, our updated data indicate a trend toward lower risk, further supporting a favorable safety profile in children. Of not, we also report that 86% of those who received TXA achieved GCS scores of 14 or more, providing additional insight into potential neuroprotective benefits, particularly in TBI populations, which was not extensively examined by Kornelsen et al.¹⁷ The observed higher rates of transfusion and surgical intervention in combat settings likely reflect confounding by indication, where severely injured children are preferentially treated with TXA, rather than a causal effect of the drug. Differences in pediatric pharmacokinetics, including faster clearance and variable weight-based dosing, may influence the efficacy of TXA in rapidly decompensating children, highlighting the importance of timing and dosing strategies. Overall, our findings corroborate the previous evidence on mortality and safety, while adding critical new insights into hospital stay, neurological outcomes, and context-specific effects, underscoring the need for standardized dosing protocols and prospective trials to optimize TXA use in pediatric trauma.

Despite these strengths, this study also has important limitations. Significant heterogeneity was observed in key outcomes, such as ICU admission rates and blood transfusion needs, which may reflect the variability in patients’ characteristics, TXA dosing protocols, and trauma settings, limiting the certainty of pooled estimates. The predominance of observational studies reduces the reliability of the findings. Only 2 RCTs were included, 1 of which had some concerns regarding missing data. Small sample sizes in certain analyses, such as those of thromboembolic events or Pediatric Quality of Life Inventory scores, and the exclusion of studies with null events (due to statistical constraints) may have introduced selection bias.

Additional sources of heterogeneity include variation in trauma type (blunt trauma vs TBI) and differences in TXA timing and dosing, which may further influence our findings.

Conclusion

Our findings suggest that TXA does not significantly alter mortality or thromboembolic risk in pediatric trauma patients but may lead to shorter hospital stays, a clinically meaningful benefit. However, the higher rate of surgical interventions and blood transfusions in combat settings warrants caution. The lack of consistent adverse effects supports the safety profile of TXA. Clinicians should weigh the potential benefits of shorter hospitalization against the uncertain effects on transfusion requirements and neurological outcomes, particularly in specific settings (combat vs civilian). Until more robust evidence becomes available, TXA use should be individualized, taking into account injury severity and resource availability.

Future research should prioritize well-designed, large-scale RCTs to detect clinically meaningful differences in mortality, thromboembolic events, and functional outcomes. Standardized TXA dosing, timing (early administration, ideally within 3 hours of injury), and route (IV vs oral) should be used to minimize variability. Newer studies should explore different formulations of TXA, including liquid solutions. The efficacy of these should be compared with existing IV regimens. Subgroup analyses by injury type and severity (e.g., using the ISS) may help identify pediatric trauma patients who would most benefit from TXA.

ABBREVIATIONS

GCS

Glasgow Coma Scale;

ICU

intensive care unit;

ISS

Injury Severity Score;

IV

intravenous;

MD

mean difference;

RCT

randomized controlled trial;

TBI

traumatic brain injury;

TXA

tranexamic acid