Sage Journals: Discover world-class research

Abstract

Background:

Total joint arthroplasty (TJA) patients are mostly of advanced age and with comorbidities such as increased body mass index (BMI) and impaired glucose tolerance. These factors and type of surgery may affect the fibrinolytic system.

Aim:

To investigate the effect of age, sex, BMI, type of surgery, and tranexamic acid (TXA) treatment on the fibrinolytic system in TJA patients.

Methods:

Ninety-nine patients undergoing TJA (32 total hip arthroplasty [THA] and 67 total knee arthroplasty [TKA]) were included in this study. Blood samples were drawn at preoperative clinic appointments and on postoperative day 1. Antigenic levels of d-dimer, plasminogen activator inhibitor 1 (PAI-1), and tissue plasminogen activator (tPA) were measured using a commercially available enzyme-linked immunosorbent assay kit. Antiplasmin activity was measured using functional method. Age, gender, hemoglobin (Hb) levels, and BMI were collected from the records.

Results:

Preoperative d-dimer and tPA levels were positively correlated with age, whereas preoperative antiplasmin was negatively correlated with age. Body mass index was only associated with preoperative tPA levels. There was no significant difference in postoperative levels of d-dimer, PAI-1, tPA, or antiplasmin between patients treated with TXA or without TXA. Percentage change in d-dimer and tPA showed significantly lower values in patients treated with TXA compared to the nontreated group. Type of surgery did not affect the fibrinolytic markers.

Conclusion:

These results confirm that advanced age and elevated BMI positively contribute to fibrinolytic dysregulation in TJA patients, whereas TXA seems to decrease the fibrinolytic activity.

Keywords

hemostasis bleeding hypercoagulability tranexamic THA TKA age BMI

Introduction

The average patient population undergoing major lower extremity total joint arthroplasty (TJA) is of increased age and body mass index (BMI) with multiple medical comorbidities, requiring the use of multiple medications, all of which, including type of surgery, have the potential to alter the perioperative fibrinolytic status in these patients. Total hip arthroplasty (THA) is obviously performed without a tourniquet, whereas a total knee arthroplasty (TKA) can be formed with or without tourniquet application. A tourniquet provides reduced quantity of blood in the operative field for the surgeon^1,2 and decreases intraoperative blood loss.³ However, tourniquet application has been shown to cause severe pressure and tissue ischemia, which can result in metabolic and cellular changes to the underlying tissues in addition to vascular injury. These changes may contribute to the increased thrombotic and fibrinolytic activity in peripheral blood that has been found when a tourniquet is used.^2,4,5 Increased fibrinolytic activity in TKA and THA patients has also been associated with increased bleeding amount.^6,7

One of the main components of the fibrinolytic system is plasminogen activator inhibitor 1 (PAI-1), which is synthesized from endothelium, liver, and adipose tissue.⁸ The PAI-1 inhibits tissue plasminogen activator (tPA), thereby inhibiting its activator effect on inactive plasminogen to active plasmin, which degrades fibrin. Activated plasmin is also inactivated by α2-antiplasmin. Low level of antiplasmin or high level of plasmin–antiplasmin (PAP) complex indicates recent fibrinolytic activity.⁹ High PAI-1 levels have been found to be correlated with high BMI¹⁰ and high glucose levels.¹¹

The tranexamic acid (TXA) is an antifibrinolytic agent, which is a synthetic derivative of the amino acid lysine. The TXA shows antifibrinolytic effect by the reversible blockade of lysine-binding sites on plasminogen molecules and thereby inhibiting the interaction of plasminogen with fibrin. The plasminogen molecule is activated by plasminogen activators after binding to fibrin. The TXA interferes with this binding and results in inhibition of plasminogen activation.¹² Antifibrinolytic treatments prevent the breakdown of a developed thrombus and decrease the bleeding amount. The TXA treatment is used for this purpose in orthopedic surgeries and has been shown to reduce blood loss and the need for blood transfusions in TKA.¹³

The aim of this study is to monitor the effect of age, gender, BMI, type of arthroplasty surgery, and TXA treatment on the fibrinolytic system parameters in TJA patients.

Methods

Upon approval of the institutional review board, 99 consecutive patients undergoing TJA (hip or knee) were recruited at the Loyola University Medical Center. Patients with a history of atrial fibrillation with stroke, thromboembolic event, cardiac or other vascular stents, or color blindness were excluded, as they were unable to receive TXA preoperatively. Patients of only 1 of the surgeons were eligible for TXA because, at the time of the study, the other 2 surgeons did not administer TXA as part of their protocol. One gram of intravenous TXA was given in the preoperative holding area to 25 patients. Seventy-four patients did not receive TXA.

Preoperative day 1 and postoperative day 1 blood samples of the 99 arthroplasty patients who underwent TKA (67) or THA (32) were obtained.

Blood samples were drawn in tubes containing 3.2% (0.109 mol/L) sodium citrate. Samples were centrifuged for preparation of platelet-poor plasma at 3000 rpm, at room temperature, for 20 minutes. Plasma samples were aliquoted and stored at −80°C until analysis.

All of the TKA surgeries were performed with the use of a tourniquet for at least a portion of the procedure, with 5 of the surgeries using the tourniquet only for cementing and the remainder of the surgeries using the tourniquet from incision until the final components was in place. Postoperatively, patients who had undergone a TKA were placed into a knee immobilizer in full extension but were allowed to range the knee as much as possible while in bed. They were allowed to ambulate immediately with the knee immobilizer on. Patients who underwent THA were not immobilized in any way and were allowed to ambulate immediately following surgery. All patients received warfarin on preoperative day 1, which was continued until postoperative day 30, with a goal international normalized ratio (INR) of 2.0 to 2.5 for anticoagulation.

The antigen levels of PAI-1, d-dimer, and tPA were measured by commercially available enzyme-linked immunosorbent assay (ELISA) kits Asserachrom PAI-1 (Diagnostica Stago, France), Asserachrom D-Di (Diagnostica Stago), and Asserachrom tPA (Diagnostica Stago), respectively. Antiplasmin activity was measured by a commercially available chromogenic substrate method (STA-Stachrom antiplasmin; Diagnostica Stago).

Parameters such as age, sex, BMI, glucose, hemoglobin A1c (HbA1c), and hemoglobin (Hb) g/dL values were collected from the electronic medical record.

Statistical Considerations

The data were analyzed using Stata version 12.0 (Statacorp, College Station, Texas). Percentage changes (PCs) in fibrinolytic marker levels were calculated as (postoperative value − preoperative value) × 100/preoperative value. Associations between age and BMI with d-dimer, PAI-1, tPA, and antiplasmin were investigated using linear regression and Spearman correlation coefficients. Differences in d-dimer, PAI-1, tPA, and antiplasmin between sex (male/female), TXA treatment (yes/no), and surgery type (hip/knee) were evaluated using Mann-Whitney U tests. A mixed analysis of variance (ANOVA) was used to determine whether age, sex, BMI, TXA treatment, or surgery type modified the change in d-dimer, PAI-1, tPA, or antiplasmin presurgery/postsurgery. A 2-tailed P value <.05 was considered significant.

Results

Baseline Characteristics

The study comprised 69 females and 30 males with a mean age of 65.5 ± 10.1 years. Sixty-seven patients underwent TKA, and 32 patients underwent THA. Patient characteristics of the TKA, THA, and combined TJA groups are shown in Table 1.

Table 1.

Characteristics of Patients Undergoing TJA.^a

	All TJA (N = 99)	THA (n = 32)	TKA (n = 67)	P
Age, years	65.5 ± 10.1	65.9 ± 11.6	65.3 ± 9.4	.423
Sex, %	30 (30.3)	12 (37.5)	18 (26.9)	.282
BMI, kg/m²	33.6 ± 6.8	31.6 ± 6.0	34.6 ± 7.1	.044
Tranexamic acid use, %	25 (25.3)	10 (31.3)	15 (22.4)	.343

Abbreviations: BMI, body mass index; THA, total hip arthroplasty; TKA, total knee arthroplasty; TJA, total joint arthroplasty; SD, standard deviation.

^a P values compare THA and TKA groups. TJA group includes both THA and TKA. Values presented as mean ± SD or n (%).

Effect of Age

Preoperative d-dimer levels were positively correlated with age (r = 0.28, P = .006; Figure 1). A 1-year increase in age was associated with an 18 ng/mL increase in preoperative d-dimer levels. When patients were categorized into 10-year intervals (ie, 30-39 [n = 1], 40-49 [n = 4], 50-59 [n = 27], 60-69 [n = 27], 70-79 [n = 27], 80-89 [n = 9]), d-dimer levels showed an increasing trend from age 40 to 49 years through 80 to 89 years (Figure 2). There was no association between PAI-1 and age (r = 0.03, P = .777). A positive correlation between age and preoperative tPA levels (r = 0.252, P = .02) and a negative correlation between age and antiplasmin levels (r = −0.27, P = .009) were noted. In mixed ANOVA, age did not modify the change observed in d-dimer, PAI-1, or antiplasmin postsurgery.

Figure 1.

Correlation plot between d-dimer levels. Age is positively associated with pre–d-dimer (ng/mL) levels (r = 0.28, P = .006).

Figure 2.

Box plots depicting the d-dimer levels in various age groups over the 30 to 90 years age span.

Effect of Sex

No significant differences were observed between female and male sex in relation to d-dimer, PAI-1, tPA, and antiplasmin. The change observed in d-dimer, PAI-1, and antiplasmin postsurgery was not modified by sex. Preoperative d-dimer, PAI-1, tPA, and antiplasmin values were found to be such as in females and males: d-dimer values (603.3 [339.2-1419] ng/mL vs 833 [493.2-1235] ng/mL, respectively, P = .508), PAI-1 values (47 [35.5-72.05] ng/mL vs 53.7 [45.3-69.3] ng/mL, respectively, P = .679), tPA values (8.8 [7.15-10.1] ng/mL vs 8.25 [6.35-10.55] ng/mL, respectively, P = .572), and antiplasmin values (75.85% [66.22%-88.02%] vs 70.53% [57.86%-85.79%], respectively, P = .108).

Effect of BMI

A positive correlation was observed between BMI and preoperative tPA levels (r = 0.26, P = .015; Figure 3), but no association was noted between BMI and d-dimer (r = 0.129, P = .212), PAI-1 (r = 0.156, P = .14), or antiplasmin (r = 0.13, P = .208). The BMI did not modify the change observed in d-dimer, PAI-1, or antiplasmin postsurgery.

Figure 3.

Correlation plot between BMI and tPA. The BMI is positively correlated with the circulating tPA antigen levels (r = 0.26, P = .015). BMI indicates body mass index; tPA, tissue plasminogen activator.

Effect of TXA

There were no significant differences in the postoperative levels of d-dimer, PAI-1, tPA, and antiplasmin between patients treated with TXA and those who were not with TXA (Table 2). Patients treated with TXA had a lower PC in d-dimer compared to patients not treated with TXA (14.3% [−8.9% to 99.1%] vs 68.3% [−1.4% to 305.2%], P = .037; Figure 4), that is, TXA treatment was associated with a lesser increase (14.3%) in d-dimer following surgery compared to non-TXA treatment (68.3% increase). Patients treated with TXA had a greater percentage decrease in tPA compared to patients not treated with TXA (−14.2% [−36.2% to −1.0%] vs 0.93% [−20.7% to 20.2%], P = .0073; Figure 5). There was no difference in the PC of PAI-1 or antiplasmin between TXA and non-TXA groups (Table 3).

Table 2.

Fibrinolytic Parameters in Patients Day 1 Postsurgery Treated With and Without TXA.^a

	TXA Treated	No TXA	P
d-dimer, ng/mL	1061 (605.3-1617)	1375 (906-1920)	.42
PAI-1, ng/mL	61.3 (37.9-74.5)	58.2 (41.8-88.0)	.96
tPA, ng/mL	7.4 (5.3-9.1)	7.7 (6.7-10)	.15
Antiplasmin, %	69.0 (55.8-82.0)	69.5 (56.2-81.9)	.88

Abbreviations: IQR, interquartile range; PAI-1, plasminogen activator inhibitor 1; tPA, tissue plasminogen activator; TXA, tranexamic acid.

^aData presented as median (IQR).

Figure 4.

Box plots demonstrating the percentage changes in d-dimer levels in patients treated (TXA+) and untreated (TXA−) with TXA. TXA indicates tranexamic acid.

Figure 5.

Box plots demonstrating the percentage changes in tPA levels in patients treated (TXA+) and untreated (TXA−) with TXA. tPA indicates tissue plasminogen activator; TXA, tranexamic acid.

Table 3.

Percentage Change in Fibrinolytic Parameters in Patients Treated With and Without TXA.^a

	TXA Treated	No TXA	P
d-dimer (% change)	14.3 (−8.9 to 99.1)	68.3 (−1.4 to 305.2)	.037
PAI-1 (% change)	8.5 (−16.2 to 46.9)	19.7 (−16.5 to 60.2)	.605
tPA (% change)	−14.2 (−36.2 to −1.0)	0.93 (−20.7 to 20.2)	.0073
Antiplasmin (% change)	−17.7 (−30.0 to 15.3)	−8.2 (−21.6 to 13.6)	.343

Abbreviations: IQR, interquartile range; PAI-1, plasminogen activator inhibitor 1; tPA, tissue plasminogen activator; TXA, tranexamic acid.

^aData presented as median (IQR).

Type of Arthroplasty

Postoperative levels and PC of d-dimer, PAI-1, tPA, and AP were not different between TKA and THA patients. Postoperative d-dimer values were similar among patients who underwent TKA or THA (median [interquartile range, IQR]: 1338 ng/mL [926.7-1594 ng/mL] vs 1364 ng/mL [602.6-1883 ng/mL], P = .484). Percentage change was also similar among patients who underwent TKA or THA (median [IQR]: 51.4% [−3.9 to 296.8] vs 17.7% [−4.9 to 186.3], P = .345). There was no significant difference in postoperative PAI-1 levels between patients who underwent TKA or THA (58 ng/mL [41.8-75.95 ng/mL] vs 62 ng/mL [36.3-99.8 ng/mL], P = .451). Percentage change in PAI-1 was similar in patients who underwent TKA or THA (7.5% [−16.25% to 44.9%] vs 24.4% [−16.6% to 87.8%], P = .317). Postoperative tPA values were not statistically different between patients who underwent TKA or THA (7.95 ng/mL [6.75-10 ng/mL] vs 6.96 ng/mL [5.53-9.7 ng/mL], P = .13). The PCs in tPA were similar between patients who underwent TKA or THA (−2.14% [−17.68 to 10.02] vs −3.7% [−33.56 to 23.47], P = .074). Postoperative antiplasmin values were similar between patients who underwent TKA or THA (69.26% [56.61%-82.08%] vs 68.65% [55.32%-79.95%], P = .907). Percentage changes in antiplasmin were not statistically different between patients who underwent TKA or THA (−10.03% [−26.71% to 3.105%] vs −8.41% [−27.28% to 22.14%], P = .327).

Effect of Blood Glucose Levels and HbA1c

As it has already been reported that there is a correlation between PAI-1 and glucose,¹¹ a potential correlation with preoperative glucose level and PAI-1 (r = 0.194; P = .184) and d-dimer (r = −0.237; P = .081) was monitored. However, no significant correlation was found. It is possible that no correlation was found since more emphasis is now placed on obtaining better perioperative glucose control.

Preoperative PAI-1 levels were not shown to be significantly different between patients with normal and high HbA1c level (median [IQR]: 47.04 ng/mL [34.76-72.06 ng/mL] vs 53.79 ng/mL [39.71-66.64 ng/mL], respectively, P = .375).

There was also no difference observed between patients taking Celebrex, a selective COX-2 inhibitor, or not in regard to PAI-1, tPA, d-dimer, or antiplasmin (data not shown).

Discussion

The population for this study was mostly comprised of patients of advanced age having high BMI with multiple medical comorbidities. Advanced age was found to correlate with elevated levels of d-dimer, increased tPA, and decreased antiplasmin. These findings support that the increased fibrinolytic activity in arthroplasty patients at preoperative period has age-related dependence. In physiological conditions, PAI-1 would be expected to accompany these changes. However, no correlation was observed between age and PAI-1 levels. The patients were divided into 10-year age intervals for reanalysis of the relationship of age and d-dimer. This dichotomization yielded obvious relationships between age and d-dimer. Our age-related d-dimer results were consistent with previous studies.^14,15 The Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial reviewed 2321 patients with type 2 diabetes and stable coronary artery disease. They found that d-dimer levels were increased proportionately with age. In contrast to our study, the BARI 2D study group found that PAI-1 antigen and activity were decreased and tPA did not change. Their interpretation for these results was that increased fibrinolytic activity was associated with increased age.¹⁵ Feinberg et al⁹ studied PAP levels in patients with atrial fibrillation. They found PAP levels were independently associated with older age and were not associated with warfarin therapy or INR levels. Their hypothesis was that increased fibrinolytic activity might be the result of increased thrombin generation. In the current study, increased tPA and d-dimer levels support increased fibrinolytic activity preoperatively. Low antiplasmin levels were thought to be the result of consumption during fibrinolysis.

In the current study, there is no gender-related difference in regard to fibrinolytic markers, which argues against the literature in which antiplasmin and d-dimer were found to be higher in female controls compared to males.¹⁶

The main sources of plasma PAI-1 are primarily liver and adipose tissue.^8,10 In this study, PAI-1 levels were not correlated with increased BMI. There was also no correlation between d-dimer and antiplasmin with BMI. Only tPA values were found to be correlated with BMI. These results led us to postulate that patients with advanced age and BMI have increased fibrinolytic activity in comparison to patients of younger age and lower BMI.

Previous studies have shown that glucose is one of the biologic substrates that increase PAI-1 values.¹¹ Results of the current study do not support the triggering effect of glucose on the levels of PAI-1. As there has been an emphasis on better perioperative glucose control and optimization, these random glucose measurements may not be a reliable marker of overall blood glucose regulation. Because of this, HbA1c levels were also used to monitor the effect of glucose on PAI-1 levels. In contrast to the current literature,¹¹ no correlation was found between blood glucose and PAI-1.

Total joint arthroplasty can lead to large amounts of blood loss. It has been shown that a TKA can lead to a decrease in preoperative hemoglobin values of 3.88 g/dL by postoperative day 3.¹² According to a meta-analysis by Zhang et al,¹³ TXA significantly reduces total blood loss by a mean of 486.7 mL in TKA and decreased the proportion of the patients needing a blood transfusion by 37%. The TXA also decreased the number of the packed red blood cell transfusion per patient. Routes of administration of TXA are different in the literature such as intra-articular,¹⁷ intravenous, and oral,¹⁸ with no literature showing one route to be superior to another.

As there is no reliable method to detect exact blood loss during surgery, we preferred to analyze the postoperative hemoglobin values in comparison to preoperative values to analyze blood loss. Blood loss during surgery was not measured directly. Patients’ postoperative hemoglobin changes were not found to be different between the TXA-treated and the nontreated group. Percentage change in hemoglobin values were also not different between the TXA-treated and the nontreated group.

The PCs in tPA and d-dimer were significantly lower in the TXA-treated group in comparison to the nontreated group. These findings support the antifibrinolytic effect of TXA in arthroplasty patients. The lower PC in d-dimer in the TXA-treated group is an expected result when the mechanism of action of TXA is considered. However, TXA’s effect on tPA is unexpected as its effect is on the plasminogen. The TXA works by inhibiting the binding of plasminogen to fibrin and has no effect on plasminogen activators. Postoperative tPA levels in TXA-treated patients were lower than preoperative values. Despite this change in TXA-treated patients, PAI-1 values were not changed in this group. In the current study, the results in regard to d-dimer are consistent with literature.¹⁹

Blanié et al¹⁹ found increased d-dimer levels 18 hours postoperatively in patients who had undergone TKA or THA when compared to preoperative values. However, the increased d-dimer levels were not observed at 6 hours postoperatively in patients who had received TXA during surgery.

Both TKA and THA have been shown to cause an increase in the fibrinolytic markers PAP and d-dimer.^20,21 Their increased levels have been observed especially after tourniquet deflation.^4,5 As there is no tourniquet use in THA and all TKA surgeries were done with the tourniquet inflated for at least a portion of the surgery, we monitored for a difference between the TKA and THA group. However, there were no significant changes noted between the 2 groups.

Many arthroplasty patients take nonsteroidal anti-inflammatory drugs (NSAID) for their analgesic and anti-inflammatory effects. The most commonly prescribed NSAID in our patient population was celecoxib. An attempt was made to monitor the effect of celecoxib on the fibrinolytic profile. There were no significant effects of celecoxib noted on the fibrinolytic markers.

Limitations of this Study

A major limitation to this study is that the only patients who received TXA were treated by 1 surgeon and that the administration of TXA was not randomized. It can be assumed that there are no significant differences between surgeons who could alter the fibrinolytic markers, but without randomization of the groups, one cannot be sure. Another limitation is that the levels of PAI-1 and tPA were evaluated with an antigenic method. This can cause insufficient evaluation of these markers during fibrinolysis. Measurement of activity of these 2 markers could give more information and allow for better evaluation of the data.

Conclusion

Our results confirm that age and BMI are essential components for imbalanced/increased fibrinolytic activity in TJA patients preoperatively. Surgeons must be aware of this, and care must be tailored for patients with older age and high BMI in order to decrease perioperative bleeding.

The TXA appears to decrease the fibrinolytic activity in arthroplasty patients and can be considered in order to diminish blood loss during TJA surgeries and potentially help decrease bleeding-related complications and blood transfusion rates. However, if this increased fibrinolytic activity is due to increased thrombin generation as a result of increased age in this patient population, the use of TXA could potentially be counterproductive. Despite the proven safety of TXA in the literature during major lower extremity orthopedic surgery, thrombotic risk with TXA usage should be evaluated in well-designed studies.

Footnotes

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) received no financial support for the research, authorship, and/or publication of this article.

References

Harsten

Bandholm

Kehlet

Toksvig-Larsen

. Tourniquet versus no tourniquet on knee-extension strength early after fast-track total knee arthroplasty; a randomized controlled trial. Knee. 2015;22(2):126–130. pii: S0968-0160(14)00313-5. doi:10.1016/j.knee.2014.12.010.

Kam

Kavanagh

Yoong

. The arterial tourniquet: pathophysiological consequences and anaesthetic implications. Anaesthesia. 2001;56(6):534–545.

Alcelik

Pollock

Sukeik

Bettany-Saltikov

Armstrong

Fismer

. A comparison of outcomes with and without a tourniquet in total knee arthroplasty: a systematic review and meta-analysis of randomized controlled trials. J Arthroplasty. 2012;27(3):331–340. doi:10.1016/j.arth.2011.04.046.

Sharrock

Sculco

Ranawat

Maynard

Harpel

. Changes in circulatory indices of thrombosis and fibrinolysis during Total Knee Arthroplasty performed under tourniquet. J Arthroplasty. 1995;10(4):523–528.

Reikerås

Clementsen

. Time course of thrombosis and fibrinolysis in total knee arthroplasty with tourniquet application. Local versus systemic activations. J Thromb Thrombolysis. 2009;28(4):425–428.

Matsuda

Nozawa

Katsube

Maezawa

Kurosawa

. Activation of fibrinolysis by reinfusion of unwashed salvaged blood after total knee arthroplasty. Transfus Apher Sci. 2010;42(1):33–37.

Yoshida

Wada

Hasegawa

. Increased fibrinolysis increases bleeding in orthopedic patients receiving prophylactic fondaparinux. Int J Hematol. 2012;95(2):160–166.

Ohkura

Shirakura

Nakatani

. Associations between plasma PAI-1 concentrations and its expressions in various organs in obese model mice. Thromb Res. 2012;130(6):e301–e34.

Feinberg

Macy

Cornell

. Plasmin-alpha2-antiplasmin complex in patients with atrial fibrillation. Stroke Prevention in Atrial Fibrillation Investigators. Thromb Haemost. 1999;82(1):100–103.

10.

Ceo

Petruceli

Bosco

. Negative correlation between D-dimer and plasminogen activator inhibitor-1 levels is absent in obese women. Blood Coagul Fibrinolysis. 2012;23(5):402–405.

11.

Mao

Kawao

Tamura

. Plasminogen activator inhibitor-1 is involved in impaired bone repair associated with diabetes in female mice. PLoS One. 2014;9(3):e92686.

12.

Sabatini

Atzori

Revello

Scotti

Debiasi

Massè

. Intravenous use of tranexamic acid reduces postoperative blood loss in total knee arthroplasty. Arch Orthop Trauma Surg. 2014;134(11):1609–1614.

13.

Zhang

Chen

Que

. The effect of tranexamic acid on blood loss and use of blood products in total knee arthroplasty: a meta-analysis. Knee Surg Sports Traumatol Arthrosc. 2012;20(9):1742–1752.

14.

Kelly

Rudd

Lewis

Hunt

. Plasma D-dimers in the diagnosis of venous thromboembolism. Arch Intern Med. 2002;162(7):747–756.

15.

McBane

II Hardison

Sobel

; BARI 2D Study Group. Comparison of plasminogen activator inhibitor-1, tissue type plasminogen activator antigen, fibrinogen, and D-dimer levels in various age decades in patients with type 2 diabetes mellitus and stable coronary artery disease (from the BARI 2D trial). Am J Cardiol. 2010;105(1):17–24.

16.

Agren

Jörneskog

Elgue

. Increased incorporation of antiplasmin into the fibrin network in patients with type 1 diabetes. Diabetes Care. 2014;37(7):2007–2014.

17.

Jang

Kao

Bohm

. Intra-articular injection of tranexamic acid to reduce blood loss after total knee arthroplasty. J Orthop Surg (Hong Kong). 2014;22(2):146–149.

18.

Irwin

Khan

Jameson

Tate

Copeland

Reed

. Oral versus intravenous tranexamic acid in enhanced-recovery primary total hip and knee replacement: results of 3000 procedures. Bone Joint J. 2013;95-B(11):1556–1561.

19.

Blanié

Bellamy

Rhayem