Safety and Efficacy of Tranexamic Acid in Hip Hemiarthroplasty for Fracture Neck Femur: a Systematic Review and Meta-analysis

Abstract

Purpose

Although numerous systematic reviews and meta-analyses have established the efficacy of tranexamic acid (TXA) in hip fracture surgeries, the included studies in those reviews have included all types of surgical interventions ranging from fixation to arthroplasty. Hip hemiarthroplasty is usually indicated in the elderly patients with femoral neck fracture and these patients have associated severe comorbidities and cognitive impairment. These subsets of patients with femoral neck fracture needs appropriate perioperative care and judicious use of antifibrinolytics. There is no meta-analysis evaluating the safety and efficacy of intravenous TXA in these patients.

Methods

Searches of PubMed, Embase and Cochrane Central Register of Controlled Trials databases revealed 102 studies on TXA in hip fracture surgeries. After screening, eight studies were found to be suitable for review. The primary objective of this meta-analysis was to compare blood transfusion rate between TXA vs. control in hip hemiarthroplasty. The secondary objectives were total blood loss, postoperative haemoglobin, surgical duration, length of hospital stay and side effects (VTE, readmission and 30 days mortality).

Results

There were one RCT, one prospective cohort study and six retrospective studies. All studies recruited the elderly patients. Intravenous (IV) TXA administration resulted in significant reduction in requirement of blood transfusion (12.7% vs. 31.9%; OR 0.28; 95% CI 0.17–0.46; p < 00,001; I² = 73%). The TXA group had significantly decreased total blood loss (MD − 100.31; 95% CI − 153.79, − 46.83; p < 0.0002). The postoperative Hb in the TXA group was significantly higher than the control group (MD 0.53; 95% CI 0.35, 0.71; p < 0.00001). There was no significant difference in the incidences of VTE (0.97% vs. 0.73%, OR 1.27; p = 0.81; I² = 64%) and readmission rate (9.2% vs. 9.64%; OR 0.79; p = 0.54), but 30-d mortality rate was significantly lower in the TXA group (3.41% vs. 6.04%; OR 0.66; p = 0.03).

Conclusions

Intravenous TXA is efficacious in the reduction of blood loss and transfusion need in hip hemiarthroplasty surgery for hip fracture, without increased risk of VTE. The blood conservation protocol led to decreased 30 days mortality in these fragile elderly patients.

Level of evidence

III.

Introduction

The rising trend of hip fracture among the elderly is alarming. Associate comorbidities, osteoporosis and perioperative blood loss contribute to significant morbidity and mortality in these patients. The reported incidence of mortality is as high as 10% at 30 days and 30% at 1 year [1, 2]. Bleeding during initial traumatic insult and subsequent surgical intervention may induce anaemia in these elderly patients. Anaemia in hip fracture has been associated with delayed functional recovery and increased long-term mortality [3–6]. Allogeneic blood transfusion also carries the risks of numerous complications such as postoperative infection, transfusion reactions, lung injury and prolonged hospital stay. Overall, it leads to increased health care cost [3–7].

An effective blood conservation strategy in this scenario can lead to early functional recovery, shorter hospital stay and reduced healthcare cost. Controlled hypotensive anaesthesia, antifibrinolytic agents, perioperative cell saver and meticulous surgical haemostasis are the standard measures to minimize blood loss in hip fracture surgery [7–9].

Tranexamic acid (TXA) is an inexpensive lysine derivative antifibrinolytic agent, used to minimise bleeding by inhibiting the conversion of plasminogen to plasmin. Studies suggest that perioperative use of TXA minimizes blood loss and decreases the need of allogeneic blood transfusion without increasing the risk of venous thromboembolism [7–9]. A recent meta-analysis by Farrow et al. reported moderate-quality evidence to support the role of TXA in the reduction of blood transfusion in hip fracture surgery [7]. However, they found low-quality evidence suggesting no increased risk of thrombotic events with its use. Zufferey et al. had similar observations [10]. They reported that TXA reduces erythrocyte transfusion but may promote a hypercoagulable state. They suggested further studies to evaluate the safety of this drug for its off-level use. All the previous systematic reviews and meta-analyses have established the efficacy of TXA in hip fracture surgeries [7, 9, 11, 12]. However, the included studies in those reviews considered all types of surgical interventions. Cannulated hip screw, dynamic hip screw and proximal femoral nail fixations are usually performed with small incision, and thus associated with minimal bleeding. Open or closed reduction and internal fixation is usually performed for healthy patients with undisplaced or minimally displaced fracture and better bone stuck. Total hip arthroplasty is also indicated in these patients who are independently mobile and without any severe comorbidities or cognitive impairment. Contrary to it, hip hemiarthroplasty is indicated in frail patients with severe comorbidities or cognitive impairment. The hemiarthroplasty patients are relatively older (> 60 years) than the fixation and total hip arthroplasty patients. These subsets of patients with femoral neck fracture needs appropriate perioperative care and judicious use of parenteral antifibrinolytics. Consequently, we conducted this systematic review and meta-analysis to evaluate the safety and efficacy of intravenous TXA in fracture neck femur in hip hemiarthroplasty patients.

Materials and Methods

Ethical Concurrence

This meta-analysis was prepared in accordance to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines [13], and the protocol was registered with PROSPERO (Regd no:CRD42022314116) (Table 1).

Table 1.

PRISMA flowchart showing methods of study recruitment

Literature Search

The electronic databases of PubMed/ Medline, Embase and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched to retrieve studies on tranexamic acid in hip hemiarthroplasty for fracture neck of femur. Two authors (SKT, PV) searched these databases using key words “tranexamic acid”, “antifibrinolytic”, “hemiarthroplasty”, “hip fracture”, “fracture neck femur”, “femoral neck fracture”, “trochanteric fracture” on 28th February 2022. The search was limited to human being and English literature. The title and abstract of the retrieved articles were screened for inclusion in this review. The full articles were extracted for studies that were found to be suitable for inclusion. The references of these articles were also manually searched to find studies suitable for inclusion in this review. Any disagreements in screening and study inclusion were resolved by a discussion between the authors. The suggestion of the senior author (RKS) was also taken to resolve the issues in case of confusion.

For inclusion of a study in this meta-analysis, it had to be either a randomized-controlled trial or a comparative cohort study evaluating the effectiveness of intravenous TXA in hip hemiarthroplasty for fracture neck femur and it must have reported the blood loss and/or blood transfusion. The studies that reported hip hemiarthroplasty following pathological fracture or failed fixation were excluded. Case reports and expert opinions were excluded from the review.

Data Extraction and Objectives

The relevant data (author, year of publication, study design, intervention and outcome) from the studies were extracted and it was tabulated in a digital form. The primary objective of this review was comparison of transfusion rate between the patients treated with intravenous TXA vs. placebo/control in hip hemiarthroplasty. The secondary objectives were comparisons of total blood loss, haemoglobin drop, surgical duration, length of hospital stay and side effects (VTE, readmission rate and 30-day mortality).

Methodological Quality and Risk of Bias Assessment

The methodological quality and risk of bias of the studies were assessed using Cochrane Collaboration Risk of Bias Tool (RoB Tool, Table 2) for randomized-controlled studies and Newcastle–Ottawa scale (NOS) for observational studies [14] (Table 3). Two authors (SKT and PV) evaluated the quality of the articles and the risk of bias. The disagreements in the risk of bias assessment were resolved by a discussion between the authors or else, the opinion of senior author (RKS) was sought.

Table 2.

Quality assessment of the randomized controlled trials (the Cochrane Collaboration Risk of Bias Tool)

Quality assessment for RCT	Random sequence generation (selection bias)	Allocation concealment (selection bias)	Blinding of participants and personnel (performance bias)	Blinding of outcome assessment (detection bias)	Incomplete outcome data (attrition bias)	Selective reporting (reporting bias)	Others
Narkbunnam 2021 [23]	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk

Table 3.

Quality assessment for non-randomized trials using Newcastle–Ottawa scale

There are seven components in Cochrane Collaboration Risk of Bias Tool [14], i.e., random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment data (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias) and other bias. All these parameters are grades as "high risk", "low risk" or "unclear risk “of bias. Similarly, the NOS allocate points in three domains (nine items) such as, the selection of the study groups, the comparability of the groups, and the ascertainment of the outcome of interest. One point is allocated to each item for low risk and no point for high risk. Studies with maximum of nine points are considered as low risk of bias, whereas seven or eight points are graded as moderate risk, and six or less as high risk of bias.

Statistical Analysis

The data were pooled and analysed by the software Review Manager (RevMan) V.5.1 [15]. The continuous data were expressed as the mean difference (MD) or standard mean difference (SMD) with 95% confidence interval (CI). The mean and standard deviations were calculated from the quantiles in few studies [16]. The odds ratio was estimated for the dichotomous data. A p value of < 0.05 was considered as statistically significant. The study heterogeneity was assessed using Cochrane’s Q (χ² p < 0.10) and quantified by I². I² of > 50% and p value of < 0.10 were taken as the threshold of significant heterogeneity [17]. Outcome parameters showing significant heterogeneity were evaluated using the random-effects model [18].

Results

Search Results

After a thorough search, we could retrieve 101 studies, of which 8 studies were found to be suitable for review [19–26, Table 1]. There were one RCTs, one prospective cohort study and six retrospective studies [19–26].

Study Characteristics and Methodological Quality Assessment

All studies recruited elderly patients with femur neck fracture. The intravenous TXA was started in the preoperative period either at the time of induction or during skin incision. Six studies used single intravenous injection and three studies used two doses (one study had both one dose and two doses group) of TXA. All total 3876 patients were evaluated in this review, of which 1339 patients received intravenous TXA, and remaining 2537 patients (one control group in the study of Narkbunnam 2021) received either no medication or placebo. The demographic details of the participants and dose schedule of TXA are mentioned in Table 4.

Table 4.

Demographic profiles of the patients

Study	Type	TXA dose schedule	No. of patients	Age	Sex (M:F)	BMI	ASA score > 3	VTE Prophylaxis	Transfusion protocol
Lee 2015 [19]	Retrospective cohort	1 dose TXA at induction	84	85.95 ± 7.60	32:52		57(67.85%)	Tinzaparin 4500 units postoperatively	Case-by-case basis, to maintain a postoperative Hb above 80 g/L
		Control	187	84.66 ± 7.69	53:134		137(73.26%)
Ashkenazi 2020 [20]	Retrospective cohort	2 dose TXA (1.5 g IV during induction and 1.5 g IV during wound closure)	504	84.2 ± 6.9	167:337		295 (58.5%)	LMWH (0.5 mg/Kg once daily) on postoperative day (POD) 1 and it was continued until full mobilization. Pt. on anticoagulants and antiplatelet on postop are continued	An Hb level < 7 g/dL in a healthy patient was the threshold for blood transfusions, while a cut-off level of 9 g/dL was used in patients at high cardiovascular risk
		Placebo	1218	83.2 ± 7.5	434:784		761 (63.5%)
Xie 2019 [21]	Retrospective cohort	1 dose TXA (15 mg/kg, 10 min prior to incision)	289	84.41 ± 7.38	91:198	24.12 ± 2.49	110(38.06%)
		Placebo	320	85.21 ± 8.27	106:214	22.11 ± 3.85	127(39.68%)
Kanthasamy 2021 [22]	Retrospective cohort	1 dose TXA 1 g TXA on induction IV	116	> 60
		Placebo	124	> 60
Narkbunnam 2021 [23]	RCT	1 dose TXA 750 mg IV TXA before skin incision	30	78.8 ± 6.5	7:23	23.1 ± 4.6		Intermittent pneumatic compressive device was applied for VTE prophylaxis. All patients were mobilized under the care of a physiotherapist on the first operative day	Hb level was < 9 g/dL, or if compromised clinical criteria, including light-headedness, orthostatic hypotension, and/or tachycardia, were detected
		2 dose TXA, 750 mg IV TXA before skin incision, and 750 mg TXA at 3 h after surgery	30	78.4 ± 10.3	5:25	23.1 ± 4.4
		Placebo	30	81.8 ± 8.1	9:21	22.8 ± 3.8
van Rijckevorsel 2022 [24]	Prospective cohort study	1 dose TXA 1 g IV at the start of the surgery	235	82.7 ± 7.1	76:159	NS	165 (70%)	Prophylactic doses (2500 IE or 5000 IE) low molecular weight heparin during six weeks	Dutch '4–5-6' rule
		Placebo	629	82.2 ± 7.2	207:422	NS	328 (53%)
Papadimitriou 2018 [25]	Retrospective study	2 dose TXA (1 g, 20 min pre-op and 1 g during wound closure)	31	No statistical Difference between both group	No statistical Difference between both group	No statistical Difference between both group
		Control	24
Akman 2016 [26]	Retrospective cohort	1 dose TXA (1 g IV)	20
		Control	5

The risk of bias, as evaluated using Cochrane Collaboration Risk of Bias Tool, revealed low risk of bias [23]. The non-randomized observational studies had low risk of bias in three studies and moderate risk of bias in four studies. A risk of bias assessment is presented in Table 2. In the study of Ashkenazi et al., the patients receiving TXA were older and had a significantly longer surgical duration compared to control group; hence, the important risk factors were not comparable in the study [20]. The comparability of the two groups was not performed in the study of Kanthasamy et al. [22]. Similarly, the TXA and control group’s patients in the study of van Rijckevorsel et al. had significant differences with regards to ASA grade, antithrombotic drug use and preoperative pulmonary embolism incidence [24].

Outcomes

1. Primary Outcome:

   Transfusion Rate (Fig. 1)

   The number of patients requiring blood transfusion in the postoperative period was mentioned in all eight studies. Narkbunnam et al. (one dose and two doses) had two treatment subgroups and hence, these subgroups of patients were separately included in the analysis [23]. Pooled analysis showed transfusion rate of 12.7% in TXA group compared to 31.9% in the control group. TXA administration resulted in significant reduction in requirement of blood transfusion (OR 0.28; 95% CI 0.17–0.46; p < 00,001; I² = 73%). Sensitivity analysis was performed to minimize heterogeneity. Exclusion of study by Van Rijckevorsel et al. [24] minimized the heterogeneity to 0% without altering the outcome (12.23% vs. 36.64%, OR 0.26, 95% CI 0.21–0.32, p < 0.00001; I² = 0%). The study by Van Rijckevorsel et al. introduced heterogeneity because of its transfusion protocol and dissimilar baseline characteristics of the patients [24].

2. Secondary Outcomes:

3. Blood Loss (Fig. 2)

   The blood loss was analysed using two parameters; total blood loss, and postoperative haemoglobin. Pooled data of two studies revealed significant reduction in total blood loss in the TXA group compared to control group(MD − 100.31; 95% CI − 153.79, − 46.83; p < 0.0002) [21, 23]. Postoperative Hb in the TXA group was significantly higher than the control group (MD 0.53; 95% CI 0.35, 0.71; p < 0.00001).

4. Venous Thromboembolism (Fig. 3)

   Six studies reported on VTE in the form of DVT or PE. The incidence of VTE in the TXA was 0.97% compared to 0.73% in the control group. There was no significant increased risk of VTE in patients receiving TXA in hip fracture surgery with an odds ratio of 1.27 (95% CI 0.18, 8.86; p = 0.81; I² = 64%).

5. Surgical Duration and Length of Hospital Stay and (Fig. 4)

   The pooled analysis of relevant studies to evaluate the impact of TXA on surgical duration and length of hospital stay did not reveal a statistically significant difference between TXA and control groups. The patients receiving TXA stayed 1.16 days less than the control group, but it was not significant (MD − 1.16; 95% CI − 3.46, 1.14; p = 0.32).

   • iv.
     Readmission and 30 days Mortality Rate (Fig. 5)
     Two studies reported the readmission rate [20, 24]. There was no significant difference in the readmission rate between the groups (9.2% vs. 9.63%; OR 0.79; p = 0.54). However, the 30-day mortality rate was significantly lower in the TXA group compared to the control group (3.41% vs. 6.04%; OR 0.66; p = 0.03).

Fig. 1.

Forest plot showing transfusion rate

Fig. 2.

Forest plot showing blood loss

Fig. 3.

Forest plot showing VTE incidence

Fig. 4.

Forest plot showing length of hospital stay and surgical duration

Fig. 5.

Forest plot showing readmission rate and 30-d mortality rate

Discussion

This meta-analysis acknowledged that TXA is efficacious in reduction of blood loss and transfusion rate in hip fracture patients treated with hip hemiarthroplasty without significantly increasing the risk of VTE. However, this review found a lower 30-d mortality rate in patients receiving TXA compared to the control group, and thus indicating the beneficial effect of blood conservation strategy in these fragile elderly patients.

The efficacy of TXA in this meta-analysis was revealed by a lesser transfusion rate, lesser total blood loss and higher postoperative Hb. The transfusion requirement in the control group was 31.9% compared to 12.7% in the TXA group. There was an unacceptable heterogeneity on this interpretation and hence a sensitivity analysis was performed. It was observed that after exclusion of study by Van Rijckevorsel et al., the heterogeneity could be eliminated completely without altering the outcome [24]. Van Rijckevorsel et al. followed Dutch 4-5-6 transfusion protocol (based on clinical assessment) where ASA class I patients were transfused at Hb threshold of 4 mmol/L (6.45 g/dl) and ASA class IV patients were transfused at 6 mmol/L (9.67 g/dl); remaining patients were transfused at a threshold of 5 mmol/L (8.06 g/dl). The study introduced selection bias as the preoperative baseline characteristics of the patients were (ASA grade) not matched between the TXA and control group in the study. There were significantly more ASA class III and IV patients in the TXA group (70%) compared to control group (53%) (p < 0.001).

TXA has already been proven to be efficacious in hip fracture surgeries and a number of meta-analyses have shown no increase incidence of thromboembolic complications [7, 9, 11, 12]. However, these meta-analyses had several limitations. Both intracapsular and extracapsular fractures were evaluated combined. There were different modalities of surgical treatment varying form minimally invasive intramedullary nails and dynamic hip screws to hemi- or total hip arthroplasties. Accordingly, the fractures and treatments were not homogenous and introduced bias into the interpretation. The patients who were treated with fixations and total hip arthroplasty were also relatively younger than hemiarthroplasty. We believe that it is very early to acknowledge the safety of TXA. Most of the studies evaluated the safety of TXA in relatively healthy patients and not in the higher risk patients. With the available evidences, it is difficult to ascertain the safety of TXA in patients with higher risk conditions. Zufferey et al. and Emara et al. reported an increased VTE incidence with TXA use in hip fracture surgeries and concluded that this might be the result of a TXA-induced hypercoagulable state [10, 27]. The previous meta-analyses on hip fracture surgery had similar conclusion. Baskaran et al. reported that the nonsignificant difference in VTE risk may be attributable to heterogeneous VTE detection methods [11]. While few studies evaluated VTE using clinical assessment and routine ultrasound screening, others used radiological investigation in case of symptoms. Variable diagnostic modalities and prophylaxis treatments for VTE were observed in our review as well. Despite than, the present review did not find a difference in VTE risk between the TXA group (0.97%) and the control group (0.73%).

The subset of hip fracture patients operated with hemiarthroplasty are usually older than 80 and have several comorbidities. They are usually on numerous medications including anticoagulants and antiplatelet. These patients are regarded as higher risk category. Although, patients of severe comorbid conditions and who are on blood thinner medications were excluded in the studies, the impact of generalised disability, systemic diseases and advanced age on vascular events cannot be overlooked in these geriatric fractured patients. This meta-analysis clarified the safety of TXA as we could not find any difference in VTE rate between the groups. The impacts of VTE and other acute vascular events in terms of ICU stay, medical management cost, morbidity and mortality need to be evaluated against the benefits of TXA. The mortality data related to these acute vascular events can be better revealed by the in-hospital mortality. However, because of lack of data, we could not evaluate this parameter. The 30-d mortality rate, as revealed in this review might show the impact of postoperative anaemia and transfusion. The lesser incidence of 30-d mortality rate in the TXA group probably indicates that blood conservation strategy is essential for reduction of 30-d postoperative mortality.

Sadeghi et al. reported that TXA administration resulted in improved functional recovery, shorter length of stay, and lower cost [28]. However, the pooled analysis of the studies in our review did not find a significant difference in length of hospital stay and surgical duration. The cost parameters were not evaluated in the studies.

There are several limitations to the evidence generated from this meta-analysis. First, the doses of TXA administration were variable across the studies. However, all studies used intravenous TXA and administered the drug before operation implying the benefits of TXA before the onset of fibrinolysis. Second, the sample size was small in numerous trials and most of the studies were retrospective. Third, the VTE evaluation methods, VTE prophylaxis modalities and transfusion protocol were different across the studies explaining some of the differences in outcomes across studies. However, it is notable that most of the studies have transfused in symptomatic anaemic patients irrespective of their Hb level; consequently, this is the strength of this meta-analysis where indication of transfusion was based on symptoms.

In conclusion, TXA minimizes blood loss and blood transfusion requirements in hemiarthroplasty surgeries of the elderly patients following hip fracture without significantly increasing the risks of venous thromboembolism. An important observation worth further research is decreased 30-d mortality in patients receiving TXA during their surgery.

Author Contributions

SKT, MJ, DN and PV designed the study. SKT, PV, NPM and AKNK collected data. SKT and DN evaluated the outcome and statistics. SKT, MJ, SS and RKS prepared the manuscript and provided intellectual content.

Funding

None.

Data Availability

The supporting data can be provided on request.

Declarations

Conflict of Interest

The authors of this manuscript declare that they have no conflicts of interest to disclose.

Ethical Approval and Consent to Participate

This systematic review has been registered in PROSPERO ((Regd no: CRD42022314116). No ethical approval is needed for such studies in our institute. Consent to participate is not applicable for this study.

Consent for Publication

All authors have read the manuscript and provided consent for publication.