Substantial perioperative blood loss in major surgeries is expected¹⁾. Hemorrhage can be a serious complication that can lead to hypovolemic shock and anemia, which is an important factor in patients’ postoperative recovery²⁾, which can increase the need for use of blood products to replace the loss associated with a substantial immunologic response and disease transmission for patients in addition to the costs of hospitalization^3,4). Intraoperative bleeding can increase the risk by increasing operative time^5,6). Tranexamic acid (TXA) is an artificial plasmin inhibitor capable of breaking fibrin by oversaturating the lysine-binding site of plasminogen molecules, thus, it has a role in protection against fibrinolysis^4,5). TXA has recently been used in performance of arthroplasty and trauma surgeries in orthopedics to reduce morbidity and blood loss, considering the associated thromboembolic events that may occur^4-6).

Several studies have reported that it can reduce total blood loss, intraoperative blood loss, and the rate of blood transfusion during joint replacement surgery without increasing the risk of venous thrombosis^7-12).

Pelvic ring fractures have a unique complex neurovascular structure and numerous plexuses. These fractures and their planned fixation are usually associated with significant blood loss and massive transfusions required for one fourth of any isolated pelvis fractures with all morbidity consequences associated with this transfusion up to death^13,14).

Use of TXA has been reported to reduce intraoperative and postoperative blood loss in acetabular and pelvic fracture surgery^15-18). However, no studies on its application during pelvic fractures (Tile B and C) alone has been reported. The objective of our study is to examine the effect of TXA in patients undergoing surgical fixation of pelvic fracture (Tile B and C) type in decreasing perioperative blood loss.

This study was a randomized controlled trial conducted at Kasr Alainy Hospital, Cairo University over a period of 24 months from January 2021 to January 2023 after approval was obtained from the Ethical Research Committee of Faculty of Medicine, Cairo University (approval No. MS-263-2022); from 124 patients who were enrolled, 24 patients did not meet the inclusion criteria. This study was conducted according to the principles of the Declaration of Helsinki (2013). Informed consent was obtained from all individual participants included in the study.

Patients with pelvic ring fractures Tile B and C caused by high energy trauma aged between 16 to 60 years were included and underwent treatment with open reduction and internal fixation. Other types of pelvic fractures or combined pelvic acetabular cases or poly trauma patients were excluded. In addition, patients with bleeding disorders, high risk patients with atrial fibrillation, valve replacement and stent implantation or patients with a previous history of thromboembolic events and major organ disease such as chronic kidney, liver, brain or heart disease or peripheral vascular diseases (all are considered contraindications for TXA use), or patients requiring same sitting multiple operations were excluded. Preoperative hemoglobin (Hb) was reserved at a value of at least 10 mg/dL while postoperatively Hb levels of <8 mg/dL call for blood transfusion.

The randomization framework was established prior to the scheduled surgeries using a sealed opaque envelope method confirming an equal number of patients in each group over the pathway of the study in case an early finish was necessary, where both the surgical team and the patients were blinded to the injected substance.

1. Surgical Procedures

The 100 patients were randomized in a blind manner according to two groups: group A received an intravenous (IV) dose of TXA 15 mg/kg+40 mL normal saline 30 minutes before incision and another dose 3 hours later, while group B, the control group received an equal volume of normal saline, twice in the same manner. The randomization agenda was applied in a blind manner prior to surgery using closed opaque sealed wrapping with an equal number in each group. All patients underwent open reduction and internal fixation for management of their pelvic fracture including anterior ring fixation using a symphyseal plate using a Pfannenstiel approach and posterior ring fixation using sacroiliac plating through the lateral window using the ilioinguinal approach under spinal epidural anesthesia. A combination of both approaches was utilized for complex fracture patterns.

Drains were clamped for two hours to allow for a tamponade effect on the hematoma and later removed after 48 hours. Postoperative prophylactic antibiotics for the first week and thromboprophylaxis medications were administered over a period of 28 days.

2. Outcome Measures

Evaluation was based on the estimated total blood loss calculated. Calculation of blood loss was based on the differences between the preoperative Hb and the postoperative Hb on the third day.

Blood loss was calculated according to the formula reported by Good et al.¹²⁾. Calculation of first Hbloss was based on the following formula: Hbloss=blood volume×(Hbi–Hbe)×0.001+Hbt (where Hbloss (g) indicates the amount of Hb lost, Hbi (g/L) is the concentration of Hb before surgery, Hbe (g/L) is the concentration of Hb on the third day after surgery, and Hbt (g) indicates the total amount of allogeneic Hb transfused).

Blood loss (mL) associated with the patient’s preoperative Hb value (g/L) was calculated as follows: Blood loss=1,000×Hbloss/Hbi. Other equations for estimating patient blood volume and blood loss have since been designed based on Nadler’s foundational work. The simplified formula used in this study can commonly be found in anesthesia and surgical textbooks¹²⁾.

The amount of blood collected in the drains was measured and documented. Other parameters including duration of hospital stay, operative time, postoperative complications, rate of perioperative blood transfusion, and number of units transfused were also measured.

During hospitalization, patients underwent a daily clinical examination for symptoms of deep venous thrombosis (DVT) in addition to performing venous duplex before discharge. During follow-up in the outpatient department, venous duplex was repeated weakly even in asymptomatic patients until partial weight bearing was allowed for detection of any early thrombotic event.

3. Statistical Analysis

The sample size was calculated using the G*Power program version 3.1.9.7 for Macintosh and Windows and according to a previous study¹⁸⁾, using t-tests and effect size=0.57 (medium), the minimum sample size was calculated as 100 patients. The power of the test is 80%.

Data were collected, coded, revised, and entered into the IBM SPSS Statistics for Windows (ver. 28.0; IBM Corp.). Considering this, the independent t-test would be the most appropriate statistical test for these data. The data were presented as number and percentages for the qualitative data, mean, standard deviations, and ranges for the quantitative data with a parametric distribution and a median with interquartile range for quantitative data with non-parametric distribution.

The chi-square test and independent t-test were used in the comparison between two groups with qualitative data and parametric distribution while Fisher and Mann–Whitney test exact test were used in the comparison between two groups with quantitative data and non-parametric distribution. The confidence interval was set to 95% and the accepted margin of error was set to 5%. Therefore, the P-value was considered significant as follows: P>0.05, non-significant (NS); P<0.05, significant; P<0.01, highly significant.

The mean age of subjects in the trial group was 42.32±14.26 years (range, 15-65 years) compared to 40.16±13.78 years in the control group, while the mean body mass index was 28.29±4.13 kg/m² (range, 21-37 kg/m²) in the trial group compared to 26.94±3.14 kg/m² (range, 20-36 kg/m²) in the control group. No statistically significant difference in demographic data, anthropometric measure, preoperative labs, or fracture types was observed between the studied groups (P>0.05) (NS). Variables in the two groups were closely comparable (Table 1, 2).

Table 1 . Fracture Types and Fixation Techniques in Both Groups.

Tile class	Trial cases			Total	Control cases			Total
	ORIF anterior arch	ORIF posterior arch	ORIF of both arches		ORIF anterior arch	ORIF posterior arch	ORIF of both arches
B1	3	1	2	6	2	2	1	5
B2	0	3	7	10	0	6	5	11
B3	0	7	7	14	0	3	10	13
C1	0	6	10	16	0	8	10	18
C2	0	1	3	4	0	2	1	3
Total	3	18	29	50	2	21	27	50

ORIF: open reduction and internal fixation..

Table 2 . Demographic and Anthropometric Data for Both Groups.

	Control group (n=50)	Trial group (n=50)	Total (n=100)	χ2	t	P-value
Sex				0.679		0.410
Female	21 (42.0)	17 (34.0)
Male	29 (58.0)	33 (66.0)
Age (yr)	40.16±13.78	42.32±14.26			–0.77	0.443
Height (m)	1.70±0.05	1.67±0.08			1.993	0.059
Weight (kg)	77.88±9.82	79.16±12.20			–0.578	0.565
Body mass index (kg/m2)	26.94±3.14	28.29±4.13			–1.834	0.070
Associated comorbidities
Diabetes mellitus	8 (16.0)	8 (16.0)	16 (16.0)
Hypertension	5 (10.0)	8 (16.0)	13 (13.0)
Asthma	0 (0)	1 (2.0)	1 (1.0)
Hypothyroidism	2 (4.0)	0 (0)	2 (2.0)
Medications
Insulin	2 (4.0)	3 (6.0)	5 (5.0)
Oral hypoglycemic	6 (12.0)	5 (10.0)	11 (11.0)
Oral betablocker	5 (10.0)	8 (16.0)	13 (13.0)
Bronchodilator	0 (0)	1 (2.0)	1 (1.0)
L-thyroxin	2 (4.0)	0 (0)	2 (2.0)

Values are presented as number (%) or mean±standard deviation..

The mean preoperative Hct was calculated as 34.49%±3.16% compared to 35.32%±3.41% in the trial and control groups, respectively, and postoperative values were 31.32%±3.76% versus 31.61%±3.40%, and the mean preoperative Hb was 11.67 g/dL versus 11.74 g/dL in the trial group and the control groups, respectively, and the postoperative values were 10.66 g/dL versus 10.47 g/dL. Regarding outcomes, no statistically significant difference in the preoperative and postoperative Hb or hematocrit values was observed between groups (P>0.05) (NS) (Table 3).

Table 3 . Outcome Measurement in Both Groups.

	Control group (n=50)	Trial group (n=50)	t	P-value
Pre. Hb (g/dL)	11.67±1.18	11.74±1.15	–0.293	0.770
Day 3 Post. Hb (g/dL)	10.66±1.09	10.47±1.34	0.793	0.430
Pre. Hct (%)	35.32±3.41	34.49±3.16	1.258	0.211
Post. Hct (%)	31.61±3.40	31.32±3.76	0.393	0.695
TBV (mL)	4,750.27±537.52	4,747.68±618.49	0.022	0.982
TBL (mL)	1,340.87±457.60	1,106.81±627.55	2.131	0.036
Drain (mL)	170.00±83.91	155.00±80.97	0.91	0.365
Stay (day)	8.28±2.23	7.06±2.57	2.536	0.013
Op. time (min)	130.30±40.88	122.54±37.38	0.991	0.324
VTE, DVT, PE	0	0	NA	NA
Hb loss (g/dL)	1.57±0.59	1.33±0.81	1.697	0.093
Hct drop (L/L)	–3.71±3.23	–3.16±2.82	–0.9	0.371

Values are presented as mean±standard deviation..

Pre.: preoperative, Post.: postoperative, Hb: hemoglobin, Hct: hematocrit, TBV: total blood volume, TBL: total blood loss, Op. time: operation time, VTE: venous thromboembolism, DVT: deep venous thrombosis, PE: pulmonary embolism, NA: not applicable..

The mean total blood loss was 1,106.81±627.55 mL in the trial group compared to 1,340.87±457.60 mL in the control group and no statistically significant difference was observed between the two groups (P<0.05). Mean blood collected in the drain during the 48-hour postoperative stay was 155 mL in the trial group compared to 170 mL in the control group with no statistically significant difference (P>0.05) (NS) (Table 3).

The rate of perioperative blood transfusion was 28% in the trial group and 76% in the control group and the mean number of transfused blood units to patients was lower in the trial group. The mean hospitalization duration was 7.06±2.57 days in the trial group compared to 8.28±2.23 in the control group and a statistically significant difference was observed for all three outcomes (P<0.05) (Table 3).

Mean operation time in the trial group was 122.54± 37.38 minutes compared to 130.30±40.88 minutes in the control group, with no statistically significant differences (P>0.05) (NS) (Table 4). There were no cases of symptomatic or asymptomatic deep vein thrombosis or pulmonary embolism during the minimum follow-up period of 12 months (Table 3).

Table 4 . Blood Transfusion Units’ Measurement in Both Groups.

	Control group (n=50)	Trial group (n=50)	χ2	P-value
Blood transfusion (units)			23.02	0.001
0	12 (24.0)	31 (62.0)
1	21 (42.0)	3 (6.0)
2	17 (34.0)	15 (30.0)
3	0 (0)	1 (2.0)
11	5 (10.0)	2 (4.0)
12	1 (2.0)	1 (2.0)
13	2 (4.0)	1 (2.0)
14	1 (2.0)	2 (4.0)

Values are presented as number (%)..

TXA reduces the drop in Hb, overall blood loss, and transfusion frequency in trauma patients without increasing the rate of venous thromboembolism^7,8,10). However, some centers are still not using TXA routinely due to the fear of embolic events, particularly for pelvic trauma patients who have an increased risk of thromboembolism compared to arthroplasty patients due to prolonged pre and postoperative immobilization^7,8). Most data from the literature agree that there is no substantial relationship between IV administration of TXA and deep vein thrombosis or pulmonary embolism in trauma patients^7-16).

In our study, we reported no cases of symptomatic thromboembolic events or myocardial infarction that may be attributed to the small sample size, which could also be explained by the benefits of TXA through decreasing perioperative bleeding, reducing operation time, encouraging rapid recovery which in turn could decrease hospitalization, postoperative infection, DVT, and pulmonary embolism. Prevention of unnecessary blood transfusion can influence the wellbeing of patients supporting a more rapid return to the activities of daily living and reducing health costs.

However, the practice of IV TXA administration during pelvic fracture surgery remains questionable^17,18). For hip fractures, the results are promising, However, there is less support for TXA usage in fractures of the pelvic ring^10,19,20). Knowledge regarding its safety and benefits’ may be helpful in improving the prognosis of pelvic surgeries as a major objective in this complex surgery to reduce blood loss and transfusion, long operative time and hospital stay for satisfactory recovery^21-23) (Table 5).

Table 5 . Comparison of the Current Study and Published Trials in Literature.

	Year	Inclusion criteria	Dose	No. of patients	Op. time (min)	Hct drop (L/L)	TBL (mL)	Transfusion rate (%)	Units transfused	DVT/PE
Current (TXA)	2023	Patients scheduled for pelvic ring ORIF	15 mg/kg (twice)	50	122	3.2	1,106	28	0.72	0/0
Current (control)	2023		0	50	130	3.7	1,340	76	1.1	0/0
Gümüştaşet al.18)	2022	Patients scheduled for pelvic or acetabular fractures ORIF	1,000 mg	73	120	4.1	1,137	21	0.9	0/0
Sadeghpour et al.21) (TXA)	2022	Patients scheduled for acetabular fractures ORIF	15 mg/kg	26	125	-	492	7	1	0/0
Sadeghpour et al.21) (control)	2022		0	25	156	-	1,084	79	1.52	1/0
Spitler et al.14) (TXA) Spitler et al.14) (control)	2019 2019	Patients scheduled for pelvic, acetabular and/or femur fractures ORIF	15 mg/kg (twice) 0	47	306	6.7	952	55	1.51	2/0
				46	284	9.6	1,325	43	1.17	1/0
Lack et al.20) (TXA)	2017	Patients scheduled for acetabular fractures ORIF	10 mg/kg (twice)	42	251	-	-	50	2.65	1/1
Lack et al.20) (control)	2017		0	46	212	-	-	33	2.36	0/0

TXA: tranexamic acid, ORIF: open reduction and internal fixation, Op. time: operation time, Hct: hematocrit, TBL: total blood loss, DVT: deep venous thrombosis, PE: pulmonary embolism..

To the best of our knowledge, our study is the first to examine the effect of TXA on the previously mentioned blood loss parameters in a specific type of pelvic fracture that was nearly fixed using the same approach and implants by the same pelvic trauma team to decrease the confounding variables that may affect the result and conclusion.

After searching data in the literature, seven studies (four randomized controlled trials^14,20-22), one prospective¹⁸⁾ and two retrospective cohort studies^13,19)) as well as a review article¹⁷⁾ and a systematic review²³⁾ discussing the same theme but either in acetabular fracture fixation, or acetabular and pelvic fracture, or pelvis, acetabular and femur fixation. No study involving a specific type of pelvic fracture fixation was reported. The limitation of these studies was either lack of a control group, or including different races in the same study, or including different types of fractures and fixation using multiple approaches representing multiple variables reflecting hazy conclusions and effect (Table 5)^13,22).

Comparing these studies with one another or with our study was difficult due to multiple factors. The TXA dose protocol was dissimilar between these trials. Transfusion threshold was diverse among different studies^13-22). In one study the estimated blood loss was not clearly determined¹⁶⁾ while the other study did not include a control group¹⁸⁾.

Regarding the perioperative parameters for blood loss, a randomized controlled trial on acetabular fracture fixation reported no statistical significance in differences between groups, although the rate of transfusion was higher for TXA and there was more estimated blood loss²⁰⁾.

By contrast, a randomized controlled trial study reported that there was no significant reduction in blood loss and blood transfusion using the TXA and no statistically significant difference in Hb and Hct drop in the postoperative period was observed between the groups²³⁾. Spitler et al.¹⁴⁾, Wadhwa et al.¹⁹⁾, and Sadeghpour et al.²¹⁾ reported significantly higher blood loss in the control group; however, statistically significantly differences in the total transfusion rates were observed between the TXA and control groups. Some studies have reported a substantial reduction in transfusion rates in patients who underwent surgery for management of pelvic and/or acetabular fracture while using TXA^15-17,23) and this data agrees with the current study.

We attempted to address the previously described limiting factors by including a control group, and all cases received care in the same hospital of specific type in management of pelvic ring fractures using similar approaches; operations were performed by one team.

The results of our study may appear marginally different, where there was a statistically significant decrease in blood transfusion units among subjects in the trial group compared with the control group. In addition, there was a statistically significant decrease in the total blood loss among those in the trial group compared with the control group. A decrease in drainage was observed among subjects in the trial group compared with the control group with no statistically significant differences (P>0.05). The results of our study also showed a statistically significant decrease in the hospital stay among subjects in the trial group compared with the control group, consistent with previous studies^22,23).

The limitation of this study is the lack of accurate measurement of actual total blood loss either hidden or observable in addition to the small sample size of a single center approach which may not be sufficient to determine definite risk and the embolic complications.

In conclusion, IV administration of TXA during surgery for treatment of pelvic fractures Tile B and C²⁴⁾ was effective in reducing total blood loss, improving Hb drops, and decreasing the rates of transfusion with no apparent increase in occurrence of venous and pulmonary thromboembolism as well as the potential for reducing the hospital stay and the operative time. Our study results look promising and pelvic surgeons are encouraged to take further steps in applying the TXA protocol in their management plan. However, further research using a more tailored method to address multiple variables will be required. Multicenter randomized control studies for use of TXA in selective types of pelvic surgeries should be conducted for development of new guidelines in the calling literature.

No funding to declare.

No potential conflict of interest relevant to this article was reported.