Original Article

Global Spine Journal

2024, Vol. 14(3) 804–811
Effect of Intraoperative Tranexamic Acid on © The Author(s) 2022
Article reuse guidelines:
Perioperative Major Hemorrhage Requiring sagepub.com/journals-permissions
DOI: 10.1177/21925682221123317
journals.sagepub.com/home/gsj
Transfusion in Patients Undergoing Elective
Spine Surgery: A Propensity Score-Matched
Analysis Using a National Inpatient Database

Akira Honda, MD, PhD1 , Yoichi Iizuka, MD, PhD1, Nobuaki Michihata, MD, PhD2,
Kazuaki Uda, PT, MPH3, Tokue Mieda, MD, PhD1, Eiji Takasawa, MD, PhD1 ,
Sho Ishiwata, MD, PhD1 , Yohei Kakuta, MD, PhD1, Yusuke Tomomatsu, MD1,
Shunsuke Ito, MD1, Kazuhiro Inomata, MD1, Hiroki Matsui, PT, MPH4,
Kiyohide Fushimi, MD, PhD5, Hideo Yasunaga, MD, PhD4, and Hirotaka Chikuda, MD, PhD1

Abstract
Study Design: Retrospective cohort study.
Objectives: This study aimed to examine whether the use of intravenous TXA in elective spine surgery is associated with
reduced perioperative massive hemorrhage requiring transfusion.
Methods: We extracted all patients who underwent decompression with or without fusion surgery for the cervical, thoracic,
and lumbar spine between April 2012 and March 2019. The primary outcome was the occurrence of massive hemorrhage
requiring transfusion, defined as at least 560 mL of blood transfusion within 2 days of spine surgery or the requirement of
additional blood transfusion from 3-7 days postoperatively. Secondary outcomes were the occurrence of thrombotic
complications (pulmonary embolism, acute coronary syndrome, and stroke) and postoperative hematoma requiring additional
surgery.
Results: We identified 83,821 eligible patients, with 9747 (12%) patients in the TXA group. Overall, massive hemorrhage
requiring transfusion occurred in 781 (.9%) patients. Propensity score matching yielded 8394 pairs. In the matched cohort, the
TXA group had a lower proportion of massive hemorrhage requiring transfusion than the control group (.7% vs 1.1%; P = .002).
There was no significant difference in the occurrence of thrombotic complications and postoperative hematoma requiring
additional surgery between both groups. The multivariable regression analysis also showed that the use of TXA was associated
with significantly lower proportions of massive hemorrhage requiring transfusion (odds ratio, .62; 95% confidence interval, .43-
.90; P = .012).

1
Department of Orthopaedic Surgery, Gunma University Graduate School of Medicine, Gunma, Japan
2
Department of Health Services Research, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
3
Department of Health Services Research, Faculty of Medicine, University of Tsukuba, Ibaraki, Japan
4
Department of Clinical Epidemiology and Health Economics, School of Public Health, The University of Tokyo, Tokyo, Japan
5
Department of Health Policy and Informatics, Tokyo Medical and Dental University Graduate School of Medicine, Tokyo, Japan

Corresponding Author:
Akira Honda, Department of Orthopaedic Surgery, Gunma University Graduate School of Medicine, 3-39-22 Showa, Maebashi, Gunma 371-8511, Japan.
Email: ahonda0711@gunma-u.ac.jp

Creative Commons Non Commercial No Derivs CC BY-NC-ND: This article is distributed under the terms of the Creative Commons
Attribution-NonCommercial-NoDerivs 4.0 License (https://creativecommons.org/licenses/by-nc-nd/4.0/) which permits non-commercial
use, reproduction and distribution of the work as published without adaptation or alteration, without further permission provided the
original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
Honda et al.
Conclusions: In this analysis using real-world data, TXA use in elective spinal surgery was associated with reduced peri-
operative massive hemorrhage requiring transfusion without increasing thrombotic complications.
Level of evidence: Prognostic Level III
Keywords
elective spine surgery, tranexamic acid, major hemorrhage requiring transfusion, thrombotic complications, nationwide
inpatient database
Introduction
Reducing intraoperative blood loss is one of the most im-
portant research topics in spine surgery. It often requires blood
transfusions due to perioperative blood loss.1-3 Previous
studies have shown that blood transfusion is associated with
poor outcomes, including in-hospital mortality, morbidity, and
postoperative complications.1,4-6 To reduce perioperative
blood loss and blood transfusions, the intraoperative admin-
istration of antifibrinolytic agents, including tranexamic acid
(TXA), has been widely used.
TXA, a synthetic derivative of the amino acid lysine, is an
antifibrinolytic agent that binds to plasminogen and blocks the
interaction of plasmin (ogen) with fibrin.7 Accordingly, TXA
prevents the dissolution of blood clots and reduces bleeding.
TXA is being increasingly used during surgery because of its
hemostatic effect. According to a recent Canadian study, TXA
is used with varying frequencies among non-cardiac surgical
procedures, with a rate of 18% in spinal fusion surgery.8
Additionally, a recent randomized controlled trial involving
patients undergoing non-cardiac surgery showed reduced
incidence of the composite bleeding outcome with tranexamic
acid.9
Although several small randomized controlled trials
(RCTs) have been conducted to examine the efficacy of TXA
in various spinal procedures, the reported efficacy of TXA in
reducing perioperative blood loss is inconsistent.10-16 Despite
meta-analyses suggesting the usefulness of TXA in spine
surgery,17-19 it remains unclear whether the available evidence
on its efficacy from relatively small trials can be directly
extrapolated to real-world clinical settings. Moreover, the
available evidence is inconclusive regarding the lingering
concern of a potential increase in thrombotic complications,
mainly because of the small sample size of prior trials. This
study aimed to evaluate the effectiveness and safety of in-
travenous TXA administration in elective spine surgery using
a Japanese national inpatient database.
Materials and Methods
This was a retrospective cohort study using the Japanese
Diagnosis Procedure Combination database, a national ad-
ministrative claims database, and discharge data.20 All aca-
demic hospitals were obliged to participate in the database,
805
and over 1000 community hospitals voluntarily contributed to
the database. Overall, approximately 50% of all acute-care
inpatients in Japan were provided. The database contains
encrypted unique identifiers; age and sex; body weight and
height; admission and discharge dates; smoking history; in-
stitution type; diagnoses coded according to the International
Classification of Diseases (ICD) 10th revision; surgical and
nonsurgical procedures coded according to Japanese original
codes; duration of anesthesia; drugs prescribed; and discharge
status. A previous study has shown that the validity of di-
agnoses and procedure records in the database was high as the
sensitivity and specificity of the primary diagnoses were
78.9% and 93.2%, respectively.21 The database clearly dif-
ferentiates between comorbidities that were already present at
admission and complications that occurred after admission;
many studies using the database have been reported else-
where.22-24 Because all data were de-identified, the require-
ment for patient informed consent was waived.
From April 2012 to March 2019, we extracted all patients
who underwent elective spine surgery with or without
1000 mg of TXA during hospitalization (Figure 1). To ensure
the generalizability of spine surgery, we included decom-
pression with or without fusion surgery for the cervical,
thoracic, and lumbar spine, and we excluded minimal invasive
surgery and long fusion such as corrective fusion surgery.
Patients who received intravenous TXA on the day of surgery
were defined as the TXA group. Because we could not pre-
cisely identify whether the one-shot topical use of TXA was
administered intravenously or topically, we excluded patients
who were administered less than 1000 mg from the TXA
group. We further excluded patients who (i) underwent spine
surgery with a main diagnosis of hematoma (ICD-10 code:
S141, S241, S341, and T093), infectious disease (tuberculosis
A180, B902, osteomyelitis/spondylitis/discitis M46), tumor
(C412, C795, D166, and M42), and traumatic disease
(M8418, M8448, S12-3, S22-3, S32-3, T02.1, and T03.1); (ii)
had thromboembolic/coagulation disorder, renal failure, and
hemodialysis on admission; (iii) required blood transfusion
and TXA use preoperatively; (iv) had anticoagulation and
antiplatelet drug use postoperatively; (v) required the 5%
maximum blood transfusion volume outliers; and (vi) un-
derwent the 5% minimum and maximum anesthesia time
outliers. The hospital volume of spine surgery was defined as
the number of spine surgeries performed annually at each
806
Figure 1. Flowchart of patient inclusion/exclusion criteria.
hospital and was categorized into quartiles (very low, low,
medium, and high). The hospital volume of TXA use was
defined as the number of spine surgeries performed annually at
each hospital in which TXA was administered. This volume
was categorized into quartiles. To ensure the hospitals that
perform a certain number of spine surgeries and hospitals that
have a certain number of TXA use cases are only chosen, we
also excluded patients who underwent spine surgery at the
hospital which was categorized as having a “very low”
hospital volume of spine surgery and/or a “very low” hospital
volume of TXA use.
Covariates and Outcomes
The covariates included age; sex; body mass index (BMI) (kg/
m2); smoking history (non-smoker, current/past smoker, and
missing); academic hospitals; emergency admission; ambu-
lance use; pre-existing comorbidities, including diabetes
mellitus (E10-E14), hypertension (I10-I15), and chronic lung
disease (J40-J47); history of cerebrovascular disease (I60-
I69), cardiac disease (I20-I25, I30-I52), hepatic cirrhosis
(K74), and at least one comorbidity; Charlson comorbidity
index (CCI) score;25 Barthel index; hospital volume of
elective spine surgery; and hospital volume of TXA use. The
following perioperative factors were also included as
Global Spine Journal 14(3)
covariates: surgical site (cervical, thoracic, and lumbar spine),
presence of ossification of the posterior longitudinal ligament,
cell-saver use during surgery, instrumentation surgery, dura-
tion of anesthesia, and total amount of TXA used. Age was
classified into 3 groups: 19-49 years, 50-64 years,
and ≥65 years. BMI was categorized as underweight
(<18.5 kg/m2), normal weight (18.5-24.9 kg/m2), overweight
(25.0-29.9 kg/m2), and obese (≥30.0 kg/m2) according to the
World Health Organization definition. CCI scores were cat-
egorized into 3 groups (≤1, 2, and ≥3).
The primary outcome was the proportion of massive hem-
orrhage requiring transfusion, defined as at least 560 mL of blood
transfusion within 2 days of spine surgery or additional blood
transfusion requirements from 3-7 days postoperatively.1,26
Because patients who underwent preoperative autologous do-
nations were more likely to be transfused their own blood during
perioperative management, we did not include autologous blood
transfusion for calculating the amount of transfusion. Secondary
outcomes were serious postoperative thrombotic complications
(pulmonary embolism, acute coronary syndrome, and stroke) and
postoperative hematoma requiring additional surgery. Compli-
cations after admission were identified with the following di-
agnoses recorded after admission: surgical site infection (T814)
along with the text data “surgical site infection” or “deep in-
fection” or “postoperative infection”; postoperative hematoma
Honda et al. 807

Table 1. Baseline characteristics before and after propensity score matching.

Unmatched Cohort Matched Cohort

Total Control TXA Control TXA

Variables (n = 83,821) (n = 74,074) (n = 9747) ASD (n = 8394) (n = 8394) ASD

Age (y), mean (SD) 68 (11) 68 (12) 68 (12) 2 68 (13) 68 (13) 1

Male 51499 (61) 45697 (62) 5802 (60) 5 5052 (60) 5013 (60) 1
BMI (kg/m2), mean (SD) 24 (5.1) 24 (5.1) 24 (4.2) 12 24 (3.9) 24 (4.3) 1
Perioperative information
TXA dose (mg), mean (SD) 1354 (800) – – 1354 (800) – – – 1267 (700) –
Anesthesia time (min), mean (SD) 212 (69) 213 (69) 207 (70) 11 208 (.4) 206 (.7) 4
Surgical site
Cervical spine 27556 (33) 24932 (34) 2624 (27) 13 2398 (29) 2348 (28) 1
Thoracic spine 1872 (2.2) 1693 (2.3) 179 (1.8) 2 152 (1.8) 154 (1.8) 0
Lumbar spine 54393 (65) 47449 (64) 6944 (71) 14 5844 (70) 5892 (70) 1
OPLL 5863 (7.0) 5294 (7.1) 569 (5.8) 5 499 (5.9) 513 (6.1) 0
Cell-saver use 4031 (4.8) 3536 (4.8) 495 (5.1) 2 446 (5.3) 416 (5.0) 2
Instrumentation surgery 24094 (29) 20643 (28) 3451 (35) 16 2900 (35) 2873 (34) 1
Smoking history
Nonsmoker 47426 (57) 41863 (57) 5563 (57) 1 4769 (57) 4814 (57) 1
Past/current smoker 29121 (35) 25576 (35) 3545 (36) 5 3095 (37) 3077 (37) 0
Missing data 7274 (8.7) 6635 (9.0) 639 (6.6) 10 530 (6.3) 503 (6.0) 1
Academic hospital 51287 (61) 45886 (62) 5401 (55) 12 4728 (56) 4812 (57) 2
Emergency admission 3562 (4.3) 3121 (4.2) 441 (4.5) 20 322 (3.8) 328 (3.9) 0
Ambulance use 825 (1.0) 727 (1.0) 98 (1.0) 0 61 (.7) 71 (.9) 1
Comorbid conditions
Diabetes mellitus 16266 (19) 14427 (20) 1839 (19) 1 1579 (19) 1566 (19) 0
Hypertension 20112 (24) 16910 (23) 3202 (33) 23 2655 (32) 2646 (32) 0
Chronic lung disease 2208 (2.6) 1918 (2.6) 290 (3.0) 2 243 (2.9) 253 (3.0) 1
Cerebrovascular disease 1281 (1.5) 1079 (1.5) 202 (2.1) 5 175 (2.1) 170 (2.0) 0
Cardiac disease 6492 (7.8) 5571 (7.5) 921 (9.4) 6 768 (9.2) 774 (9.2) 0
Hepatic disease 9357 (11) 7538 (10) 1819 (19) 25 1592 (19) 1557 (19) 1
At least one comorbidity 39398 (47) 33819 (46) 5579 (57) 24 4772 (57) 4727 (56) 1
Charlson comorbidity index
≤1 75487 (90) 67320 (91) 8167 (84) 21 6987 (83) 7033 (84) 2
2 7172 (8.6) 5805 (7.8) 1367 (14) 20 1241 (15) 1186 (14) 2
≥3 1162 (1.4) 949 (1.3) 213 (2.2) 7 166 (2.0) 175 (2.1) 1
Barthel index, mean (SD) 95 (16) 95 (16) 95 (16) 1 95 (16) 95 (16) 0
Hospital volume
Low (23-43) 27149 (32) 24564 (33) 2585 (27) 15 2545 (30) 2498 (30) 1
Medium (44-91) 28352 (34) 25804 (35) 2548 (26) 19 2021 (24) 2135 (25) 3
High (>92) 28320 (34) 23706 (32) 4614 (47) 32 3828 (46) 3761 (45) 2
TXA use volume
Low (22-41) 26626 (32) 23533 (32) 3093 (32) 0 2507 (30) 2555 (30) 1
Medium (42-85) 28563 (34) 25450 (34) 3113 (32) 7 2321 (28) 2383 (28) 2
High (>86) 28632 (34) 25091 (34) 3541 (36) 7 3566 (42) 3456 (41) 3
Data are presented as numbers (%) if not mentioned. ASDs are presented as %.
Abbreviations: TXA, tranexamic acid; ASD, absolute standardized difference; SD, standard deviation; BMI, body mass index; and OPLL, ossification of the
posterior longitudinal ligament.

(S141, S241, S341, T093, and T810) along with the text data coronary syndrome (I21-I24); heart failure (I50); stroke (I60–
“hematoma” or “epidural hematoma”; pulmonary embolism I64); urinary tract infection (N30, N34, N36-N37, and N39); and
(I26); respiratory complications (pneumonia [J12-J18 and J69], renal failure (N17–N19). We identified patients with at least one
respiratory failure [J96], and respiratory disorders [J95]); acute complication during hospitalization.
808 Global Spine Journal 14(3)

Table 2. Outcomes before and after propensity score matching.

Unmatched Cohort Matched Cohort

Control TXA Control TXA

Outcomes (n = 74,074) (n = 9747) (n = 8394) (n = 8394) ARR (95% CI) P-value

Major hemorrhage requiring transfusion 717 (1.0) 64 (.7) 93 (1.1) 56 (.7) –.44 (–.72 to –.16) .002
Blood transfusion
Overall 1667 (2.3) 165 (1.7) 200 (2.4) 139 (1.7) –.73 (–1.15 to –.30) .001
during operation 1015 (1.4) 117 (1.2) 110 (1.3) 99 (1.2) –.13 (–.47 to –.20) .444
within 3 days 632 (.9) 51 (.5) 76 (.9) 44 (.5) –.38 (–.64 to –.13) .003
within 7 days 748 (1.0) 65 (.7) 96 (1.1) 57 (.7) –.47 (–.75 to –.18) .002
Thromboembolic complications 9 (.01) 2 (.02) 2 (.02) 2 (.02) .00 N/A
Pulmonary embolism 2 (<.01) 0 (.0) 1 (.01) 0 (.0) -.01 (-.04 to .01) .317
Acute coronary syndrome 7 (.01) 2 (.02) 1 (.01) 2 (.02) .01 (–.03 to .05) .564
Stroke 0 (0) 0 (0) 0 (0) 0 (.0) .00 N/A
Complication after admission
Postoperative hematoma with operation 127 (.2) 27 (.3) 19 (.2) 17 (.2) .00 N/A
In-hospital death 20 (.03) 2 (.02) 3 (.04) 2 (.02) -.01 (–.06 to .04) .655
Surgical site infection with operation 185 (.2) 35 (.4) 26 (.3) 29 (.3) .00 N/A
At least one complication after admission 1924 (2.6) 226 (2.3) 201 (2.4) 198 (2.4) .00 N/A
Data are presented as numbers (%) if not mentioned.
Abbreviations: TXA, tranexamic acid; ARR, absolute risk reduction; CI, confidence interval; and N/A, not applicable.

Statistical Analysis randomized controlled trial. In this study, IV was defined as

We performed a propensity score matching analysis to balance
the measured potential confounders between groups.27 We
estimated the propensity scores for TXA using a multivariable
logistic regression model. All covariates were included as
predictive variables in the model. We used the one-to-one
nearest neighbor method with a caliper width of .2 of the
standard deviation of the estimated propensity scores. Each
patient in the TXA group was matched to one patient in the
control group without replacement. We calculated the absolute
standardized differences for each covariate before and after
matching and confirmed the balanced distribution of co-
variates between the TXA and control groups. To adjust for
clustering within hospitals, generalized estimating equations
were fitted with multivariable logistic regression models after
propensity score matching. Age, sex, BMI category, surgical
site, instrumentation surgery, cell-saver use, duration of an-
esthesia, CCI category, and hospital volume for spinal surgery
were included in the model.
We also performed the following sensitivity analyses. First,
to assess the robustness of our results in propensity score
matching, we conducted propensity score-stabilized inverse
probability of treatment weighting (IPTW).28 This method can
maintain the sample size of the original data, yield more
precise interval estimations of the variance of the main effect,
and control type I error, as compared with the non-stabilized
IPTW. Second, to control for unmeasured confounders and
confirm our propensity score analyses, we conducted an in-
strumental variable (IV) analysis.29,30 An IV analysis can
adjust for measured and unmeasured confounders, similar to a
the proportion of TXA use in each hospital, that is, the
hospitals’ preference regarding TXA. These preference-based
variables are obviously associated with the treatment of in-
terest but have no direct association with the outcome. We
used a two-stage residual inclusion method for IV analysis.
Third, to improve the generalizability of our results, we in-
cluded patients who were treated with anti-coagulant or anti-
platelet therapy after surgery.
Continuous variables are presented as mean (standard
deviation) and categorical variables are described as numbers
(%). We compared the crude outcomes between the TXA and
control groups using the t-test for continuous outcomes and χ 2
test for binary outcomes. A two-sided P-value <.05 was
considered statistically significant for all tests. All analyses
were performed using Stata/MP version 15 (StataCorp, Col-
lege Station, TX).
Results
We included 83,821 eligible patients who underwent elective
spinal surgery (mean age, 68 years; male, 61%) (Table 1).
Overall, TXA was administered to 9747 patients (12%), and
the mean TXA dosage was 1354 mg. Of these 9747 patients,
TXA was most frequently used for lumbar spine surgery
(13%), followed by thoracic spine surgery (9.6%) and cervical
spine surgery (9.5%). Propensity score matching yielded 8394
pairs. In the unmatched groups, the proportions of emergency
admissions, hypertension, hepatic disease, CCI ≥2, and
hospital volume of spinal surgery were higher in the TXA
Honda et al.
Table 3. Multivariable logistic regression analysis after propensity score matching.
Outcomes Odds Ratio
Major hemorrhage requiring transfusion .62
Postoperative thromboembolic complication 1.07
Postoperative hematoma with operation .98
group than in the control group. They were well-balanced after
propensity score matching.
Table 2 shows the clinical outcomes with and without TXA
use before and after propensity score matching. Overall,
massive hemorrhage requiring transfusion was occurred in
781 (.9%) patients. Postoperative thrombotic complications
and postoperative hematoma requiring additional surgery
occurred in 11 (.01%) and 154 (.2%) patients, respectively. In
the matched cohort, the TXA group had a lower occurrence of
massive hemorrhage requiring transfusion than the control
group (.7% vs 1.1%; P = .002). There was no significant
difference in the occurrence of thromboembolic or other
complications after admission between groups.
The multivariable logistic regression analysis also showed
that massive hemorrhage requiring transfusion was signifi-
cantly less frequent in the TXA group than in the control group
(odds ratio [OR], .62; 95% confidence interval [CI], .43-.90;
P = .012) (Table 3). The use of TXA was not significantly
associated with an increase in the proportion of postoperative
thrombotic complications (OR, 1.07; 95% CI, .16-7.40; P =
.94) or postoperative hematoma requiring additional surgery
(OR, .98; 95% CI, .48-1.99; P .95). The results of sensitivity
analyses using IV and IPTW were similar to those of the main
analyses (Supplemental Tables 1 and 2). Analyses including
patients treated with postoperative anti-coagulant or anti-
platelet therapy revealed similar findings as those of the
main analyses (Supplemental Table 3).
Discussion
In the present analysis of nationwide real-world data, we
found that the intraoperative use of TXA was associated with
reduced perioperative massive hemorrhage requiring trans-
fusion during elective spine surgery without increasing
thrombotic complications. There was no significant difference
in the occurrence of postoperative hematoma that required
additional surgery between groups.
Previous RCTs have shown that TXA reduced intra-
operative estimated blood loss in spine surgery,11,13-15
whereas several other studies on TXA in spine surgery
failed to reveal significant blood loss intraoperatively.10,12,16
The limited settings and small sample sizes of the RCTs may
be the reasons for this inconsistency. Furthermore, most RCTs
have been conducted at a single institution, which may have
led to different results. The present study revealed that the use
of intravenous TXA was associated with reduced massive
hemorrhage requiring transfusion and perioperative blood
809
95% Confidence Interval P-value
(.43 to .90) .012
(.16 to 7.40) .94
(.48 to 1.99) .95
transfusion, whereas intraoperative blood transfusion was not
different between patients who used TXA and those who did
not. These results are similar to those of previous meta-
analyses.17-19
Although previous RCTs10-16 and meta-analyses17-19 have
studied the efficacy of TXA use in spine surgery, some
limitations exist in those RCTs regarding the safety profile in
real-world clinical settings. Postoperative serious complica-
tions in spinal surgery, including pulmonary embolism, acute
coronary syndrome, and stroke, are rare, ranging from .0%-
.9%.23,31 Although a previous meta-analysis suggested that
intravenous TXA was safe for patients who underwent any
medical discipline, no population-based study has evaluated
the risk of TXA use during elective spine surgery.32 The
present study is the largest and has enough sample size to
estimate rare outcomes with an effect size of .2 and 90%
power. Our results indicate that serious thrombotic compli-
cations were not associated with the use of TXA in elective
spine surgery.
Another concern is whether the use of TXA increases or
decreases the formation of postoperative epidural hematoma.
Several studies have reported that the incidence of symp-
tomatic postoperative epidural hematoma in spinal surgery
ranges from .0%-1.0%.33-35 Studies have also reported that
obesity (body mass index ≥35 or higher), multilevel proce-
dures, an American Society of Anesthesiology classification >
III, revision surgery, dural repair intraoperatively, and peri-
operative transfusion were independent risk factors for
postoperative hematoma.33-35 However, it is difficult to pre-
cisely evaluate the risk of postoperative hematoma because of
its low incidence and complex heterogeneity. The present
study showed that the use of TXA was not associated with any
increase or decrease in postoperative epidural hematoma that
required additional surgery.
This study has several limitations. First, we could not
obtain precise information on the operation. The American
Society of Anesthesiology class and intraoperative compli-
cations, including dural tear and uncontrollable bleeding, may
affect perioperative bleeding and postoperative hematoma
formation and could be unmeasured confounders. However,
we used IV analysis as a confirmatory analysis and found
similar results to those of the main analysis. Thus, we believe
that these unmeasured confounders may have little effect on
perioperative blood transfusions and serious postoperative
complications. Second, there was no information regarding
the timing, dose, and administration route of intraoperative
TXA use intraoperatively. Because the common criteria for the
810
timing and dosage of intravenous TXA is a loading dose of
20 mg/kg (maximum 1000 mg) plus a maintenance dose of 1
or 2 mg/kg/h in Japan,36 we excluded patients who were
administered less than 1000 mg of TXA from the TXA group.
It is expected that these patients were administered one-shot
TXA at the end of surgery. We were also unable to identify
whether TXA was used postoperatively, such as for the
treatment of cerebral hemorrhage, or as part of the stan-
dardized postoperative deep venous thrombosis prophylaxis
protocol. However, because no patients experienced stroke,
including cerebral hemorrhage, postoperatively, there were no
cases of TXA use, at least for cerebral hemorrhage postop-
eratively. Furthermore, according to the sensitivity analysis,
there was little impact of anti-coagulant or anti-platelet
therapy on the effect of TXA. Third, the database lacked
information on hemoglobin and hematocrit values. However,
we used blood transfusions as a surrogate variable for these
predictors. Despite these limitations, the present study was a
large-population study to reveal the effectiveness and safety of
intravenous TXA use in elective spine surgery.
In conclusion, our study showed that the use of intravenous
TXA in elective spinal surgery was associated with reduced
perioperative massive hemorrhage requiring transfusion
without any increase in postoperative adverse events, in-
cluding thrombotic complications. This study underlines the
effectiveness and safety of the intraoperative use of TXA in
spine surgery in real-world settings.
Appendix
Abbreviations
CCI Charlson comorbidity index
ICD International Classification of Diseases
TXA Tranexamic acid
IV Instrumental variable
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for
the research, authorship, and/or publication of this article: This work
was supported by grants from the Ministry of Health, Labour and
Welfare, Japan (21AA2007 and 20AA2005) and the Ministry of
Education, Culture, Sports, Science and Technology, Japan
(20H03907).
Research Ethics and Patient Consent
This study was approved by the Institutional Review Board of The
University of Tokyo [approval number: 3501-(3) (December 25th,
2017)]. Because all data were de-identified, the requirement for
patient informed consent was waived.
Global Spine Journal 14(3)
ORCID iDs
Akira Honda, MD  https://orcid.org/0000-0001-8462-6024
Eiji Takasawa, MD, PhD  https://orcid.org/0000-0002-7995-6093
Sho Ishiwata, MD  https://orcid.org/0000-0002-6730-977X
Supplemental Material
Supplemental material for this article is available online.
References
1. Carson JL, Triulzi DJ, Ness PM. Indications for and adverse
effects of red-cell transfusion. N Engl J Med. 2017;377:
1261-1272. doi:10.1056/NEJMra1612789.
2. Nuttall GA, Horlocker TT, Santrach PJ, Oliver WC, Dekutoski
MB, Bryant S. Predictors of blood transfusions in spinal in-
strumentation and fusion surgery. Spine. 2000;25:596-601. doi:
10.1097/00007632-200003010-00010.
3. Purvis TE, Goodwin CR, De la Garza-Ramos RD, et al. Effect of
liberal blood transfusion on clinical outcomes and cost in spine
surgery patients. Spine J. 2017;17:1255-1263. doi:10.1016/j.
spinee.2017.04.028.
4. Guzzetta NA, Allen NN, Wilson EC, Foster GS, Ehrlich AC,
Miller BE. Excessive postoperative bleeding and outcomes in
neonates undergoing cardiopulmonary bypass. Anesth Analg.
2015;120:405-410. doi:10.1213/ANE.0000000000000531.
5. Demaret P, Tucci M, Karam O, Trottier H, Ducruet T, Lacroix J.
Clinical outcomes associated with RBC transfusions in critically
ill children: A 1-year prospective study. Pediatr Crit Care Med.
2015;16:505-514. doi:10.1097/PCC.0000000000000423.
6. Goobie SM, Haas T. Perioperative bleeding management in
pediatric patients. Curr Opin Anaesthesiol. 2016;29:352-358.
doi:10.1097/ACO.0000000000000308.
7. McCormack PL. Tranexamic acid: A review of its use in the
treatment of hyperfibrinolysis. Drugs. 2012;72:585-617. doi:10.
2165/11209070-000000000-00000.
8. Houston BL, Fergusson DA, Falk J, et al. Prophylactic tra-
nexamic acid use in non-cardiac surgeries at high risk for
transfusion. Transfus Med. 2021;31:236-242. doi:10.1111/tme.
12780.
9. Devereaux PJ, Marcucci M, Painter TW, et al. Tranexamic acid
in patients undergoing noncardiac surgery. N Engl J Med. 2022;
386:1986-1997. Online ahead of print doi:10.1056/
nejmoa2201171.
10. Tsutsumimoto T, Shimogata M, Ohta H, Yui M, Yoda I, Misawa
H. Tranexamic acid reduces perioperative blood loss in cervical
laminoplasty: a prospective randomized study. Spine. 2011;36:
1913-1918. doi:10.1097/BRS.0b013e3181fb3a42.
11. Wong J, Beheiry HE, Rampersaud YR, et al. Tranexamic acid
reduces perioperative blood loss in adult patients having spinal
fusion surgery. Int Anesthes Res Society. 2008;107:1479-1486.
12. Wang Q, Liu J, Fan R, et al. Tranexamic acid reduces post-
operative blood loss of degenerative lumbar instability with
stenosis in posterior approach lumbar surgery: A randomized
controlled trial. Eur Spine J. 2013;22:2035-2038. doi:10.1007/
s00586-013-2836-z.
Honda et al.
13. Colomina MJ, Koo M, Basora M, Pizones J, Mora L, Bagò J.
Intraoperative tranexamic acid use in major spine surgery in
adults: A multicentre, randomized, placebo-controlled trial†. Br
J Anaesth. 2017;118:380-390. doi:10.1093/bja/aew434.
14. Shi H, Ou Y, Jiang D, Quan Z, Zhao Z, Zhu Y. Tranexamic acid
reduces perioperative blood loss of posterior lumbar surgery for
stenosis or spondylolisthesis: A randomized trial. Medicine.
2017;96:1e5718. doi:10.1097/MD.0000000000005718.
15. Goobie SM, Zurakowski D, Glotzbecker MP, et al. Tranexamic
acid is efficacious at decreasing the rate of blood loss in ado-
lescent scoliosis surgery: A randomized placebo-controlled trial.
J Bone Joint Surg Am. 2018;100:2024-2032. doi:10.2106/JBJS.
18.00314.
16. Peters A, Verma K, Slobodyanyuk K, et al. Antifibrinolytics
reduce blood loss in adult spinal deformity surgery: A pro-
spective, randomized controlled trial. Spine. 2015;40:
E443-E449. doi:10.1097/BRS.0000000000000799.
17. Willner D, Spennati V, Stohl S, Tosti G, Aloisio S, Bilotta F.
Spine surgery and blood loss: systematic review of clinical
evidence. Anesth Analg. 2016;123:1307-1315. DOI: 10.1213/
ANE.0000000000001485.
18. Hui S, Peng Y, Tao L, et al. Tranexamic acid given into wound
reduces postoperative drainage, blood loss, and hospital stay in
spinal surgeries: a meta-analysis. J Orthop Surg Res. 2021;16:
401. doi:10.1186/s13018-021-02548-6.
19. Cheriyan T, Maier SP 2nd, Bianco K, et al. Efficacy of tra-
nexamic acid on surgical bleeding in spine surgery: A meta-
analysis. Spine J. 2015;15:752-761.
20. Yasunaga H. Real world data in Japan: Chapter. The Diagnosis
Procedure Combination database. Ann Clin Epidemiol. 2019;1:
76-79.
21. Yamana H, Moriwaki M, Horiguchi H, Kodan M, Fushimi K,
Yasunaga H. Validity of diagnoses, procedures, and laboratory
data in Japanese administrative data. J Epidemiol. 2017;27:
476-482. doi:10.1016/j.je.2016.09.009.
22. Honda A, Michihata N, Iizuka Y, et al. Clinical features and
early post-operative complications of isolated C2 odontoid
fractures: a retrospective analysis using a national inpatient
database in Japan. Eur Spine J. 2021;30:3631-3638. doi:10.
1007/s00586-021-06862-9. Online ahead of print.
23. Chikuda H, Yasunaga H, Horiguchi H, et al. Mortality and mor-
bidity in dialysis-dependent patients undergoing spinal surgery:
analysis of a national administrative database in Japan. J Bone Joint
Surg Am. 2012;94:433-438. doi:10.2106/JBJS.K.00183.
24. Chikuda H, Ohya J, Horiguchi H, et al. Ischemic stroke after
cervical spine injury: Analysis of 11, 005 patients using the
Japanese Diagnosis Procedure Combination database. Spine J.
2014;14:2275-2280. doi:10.1016/j.spinee.2014.01.024.
811
25. Quan H, Li B, Couris CM, et al. Updating and validating the
charlson comorbidity index and score for risk adjustment in
hospital discharge abstracts using data from 6 countries. Am J
Epidemiol. 2011;173:676-682. doi:10.1093/aje/kwq433.
26. Elgafy H, Bransford RJ, McGuire RA, Dettori JR, Fischer D. Blood
loss in major spine surgery: Are there effective measures to decrease
massive hemorrhage in major spine fusion surgery? Spine. 2010;
35(suppl l):S47-S56. doi:10.1097/BRS.0b013e3181d833f6.
27. Rosenbaum PR, Rubin DB. Constructing a control group using
multivariate matched sampling methods that incorporate the
propensity score. Am Stat. 1985;39:33-38. doi:10.1080/
00031305.1985.10479383.
28. Xu S, Ross C, Raebel MA, Shetterly S, Blanchette C, Smith D.
Use of stabilized inverse propensity scores as weights to directly
estimate relative risk and its confidence intervals. Value Health.
2010;13:273-277. doi:10.1111/j.1524-4733.2009.00671.x.
29. Baiocchi M, Cheng J, Small DS. Instrumental variable methods
for causal inference. Stat Med. 2014;33:2297-2340. DOI: 10.
1002/sim.6128.
30. Aso S, Yasunaga H. Introduction to instrumental variable
analysis. Ann Clin Epi. 2020;2:69-74. doi:10.37737/ace.2.3_69.
31. Colomina MJ, Bagó J, Pérez-Bracchiglione J, et al. Throm-
boprophylaxis in elective spinal surgery: A protocol for sys-
tematic review. Medicine (Baltim). 2020;99:e20127. doi:10.
1097/MD.0000000000020127.
32. Taeuber I, Weibel S, Herrmann E, et al. Association of intra-
venous tranexamic acid with thromboembolic events and
mortality: A systematic review, meta-analysis, and meta-re-
gression. JAMA Surg. 2021;156:e210884. doi:10.1001/
jamasurg.2021.0884.
33. Glotzbecker MP, Bono CM, Wood KB, Harris MB. Post-
operative spinal epidural hematoma: A systematic review.
Spine. 2010;35:E413-E420. doi: 10.1097/BRS.
0b013e3181d9bb77.
34. Abola MV, Du JY, Lin CC, Schreiber-Stainthorp W, Passias PG.
Symptomatic epidural hematoma after elective cervical spine
surgery: Incidence, timing, risk factors, and associated com-
plications. Oper Neurosurg (Hagerstown). 2021;21:452-460.
doi:10.1093/ons/opab344.
35. Knusel K, Du JY, Ren B, Kim CY, Ahn UM, Ahn NU.
Symptomatic epidural hematoma after elective posterior lumbar
decompression: Incidence, timing, risk factors, and associated
complications. HSS J. 2020;16(suppl 2):230-237. doi:10.1007/
s11420-019-09690-2.
36. Farrokhi MR, Kazemi AP, Eftekharian HR, Akbari K. Efficacy
of prophylactic Low dose of tranexamic acid in spinal fixation
surgery: A randomized clinical trial. J Neurosurg Anesthesiol.
2011;23:290-296. doi:10.1097/ANA.0b013e31822914a1.