A break-even analysis of tranexamic acid for prevention of periprosthetic joint infection following total hip and knee arthroplasty

Abstract

Purpose

Despite the commonplace use of tranexamic acid in total joint arthroplasty, much of the current data regarding its cost-effectiveness examines savings directly related to its hemostatic properties, without considering its protective effect against periprosthetic joint infections. Using break-even economic modeling, we calculated the cost-effectiveness of routine tranexamic acid administration for infection prevention in total joint arthroplasty.

Materials and methods

The cost of intraoperative intravenous tranexamic acid, the cost of revision arthroplasty for periprosthetic joint infections, and the baseline rates of periprosthetic joint infections in patients who did not receive intraoperative tranexamic acid were obtained from the literature and institutional purchasing records. Break-even economic modeling incorporating these variables was performed to determine the absolute risk reduction in infection rate to make routine intraoperative tranexamic acid use economically justified. The number needed to treat was calculated from the absolute risk reduction.

Results

Routine use of intraoperative tranexamic acid is economically justified if it prevents at least 1 infection out of 3125 total joint arthroplasties (absolute risk reduction = 0.032%). Cost-effectiveness was maintained with varying costs of tranexamic acid, infection rates, and periprosthetic joint infection costs.

Conclusion

The routine use of intraoperative tranexamic acid is a highly cost-effective practice for infection prevention in primary and revision total joint arthroplasty. The use of tranexamic acid is warranted across a wide range of costs of tranexamic acid, initial infection rates, and costs of periprosthetic joint infection treatment.

1. Introduction

Periprosthetic joint infection (PJI) is a devastating complication that can occur after total joint arthroplasty (TJA). Following total knee (TKA) and hip arthroplasty (THA) with PJI, one-year mortality is 4% and five-year mortality is greater than 20%.^1,2 In 2009, the national economic burden of PJI was $566 million, and was predicted to increase nearly threefold to $1.6 billion by 2020.³ Because of the morbidity, mortality, and financial costs associated with PJIs, new interventions are needed to reduce the incidence of this complication.

Tranexamic acid (TXA) is an antifibrinolytic agent that acts as a reversible competitive inhibitor of the lysine receptor found on plasminogen.⁴ Reduced conversion of plasminogen to plasmin prevents the degradation of fibrin, ultimately stabilizing the fibrin matrix at the conclusion of the clotting cascade.⁵ While TXA has been utilized among other surgical subspecialties for several decades to prevent excessive blood loss, only recently has it become routinely utilized in TJA as a blood conserving strategy.^6,7 In fact, given its safety (even in high risk patients), TXA has now become the standard of care for patients undergoing elective TJA.^8,9

Interestingly, as the utilization of TXA in primary and revision arthroplasty has grown, some of the indirect benefits of its use have become apparent. Most notably, emerging evidence has found that intraoperative use of TXA may reduce the rate of PJI after primary and revision TJA.^8,10 The decreased rate of infection in patients who receive tranexamic acid may be related to lower rates of allogeneic blood transfusion.¹¹ However, further studies examining the effect of tranexamic acid on infection risk are required to better understand TXA's mechanism of infection prevention.

Despite the widespread adoption of TXA in TJA, the cost effectiveness of this intervention – especially when considering it as an infection control measure – is unknown. Several previous studies have found that TXA substantially reduced total hospitalization costs for individual patients. For example, Gillette et al. (2013) found that the administration of TXA before primary THA and TKA reduced hospitalization costs from $15,978 to $15,099 per patient.¹² In similar analyses conducted by Chimento et al. (2013), average cost savings were $1500 per patient with the use of topical TXA for primary TKA.¹³ However, previous economic analyses did not account for the benefits of PJI prevention, a costly arthroplasty complication. As a result, prior studies likely underestimated the true cost savings associated with TXA use in lower limb arthroplasty. Given the commonplace use of TXA in TJA, it is imperative for providers to have an accurate understanding of its economic implications – especially as payment models shift towards placing more financial risk on health systems.⁹ In this study, we utilized break-even economic modeling to determine whether the routine use of intraoperative TXA is an economically viable therapy for preventing PJI after TJA.

2. Methods

We developed a break-even economic model, as described previously by Hatch et al. (2017), to assess the viability of tranexamic acid as an infection prevention measure following TJA.¹⁴ Break-even modeling establishes the effectiveness of an intervention necessary to justify routine clinical use. The break-even analysis is dependent on a defined cost of the preventive strategy, infection rate, and cost of infection treatment. The use of the intervention is economically justified if it is capable of reducing the baseline infection rate below the break-even value (Fig. 1). The difference between the initial and final infection rate was used to calculate the absolute risk ratio (ARR), which was then used to calculate the number needed to treat (NNT). Furthermore, the model has the ability to calculate the economic viability of the intervention over a range of costs and infection rates by holding certain variables constant, while changing others.

Fig. 1.

Break-even analysis equation.

The institutional cost of tranexamic acid was calculated using a standard 2-g (2g) dose of intravenous tranexamic acid. At our institution, 1g is given prior to incision and 1g is given during closure. With a cost of $4.40 per 1g dose, a total cost of $8.80 was used for total cost of tranexamic acid therapy. We also conducted sensitivity analysis considering different costs of tranexamic acid in order to account for variability in cost across institutions. Mean blood loss does not differ among the oral, intravenous, and topical groups, but the cost of treatment varies amongst the three treatment modalities, with the oral route typically incurring the lowest cost.¹⁵ Assessing the cost-effectiveness of intravenous TXA – the most costly of the three administration routes – prevented overestimation of the cost-effectiveness of TXA^16(p).

The initial infection rate used was 3.4%. This represents the rate of PJI in patients who did not receive TXA during primary TJA, as reported by Yazdi et al. (2019).¹⁰ However, given the potential variability in infection rates, we also performed a sensitivity analysis to assess the impact across a wide range of initial infection rates (1–10%). This range also included the baseline rate of PJI in patients undergoing revision TJA for aseptic failure.¹⁷

The cost associated with treating a PJI after TJA used in our analysis was $27,870. This cost was adjusted for inflation (USD) and derived from costs reported by Kamath et al. (2015).¹⁸ These costs represented the mean inpatient hospitalization costs for revision TKA to treat PJI. This cost was chosen because it is likely an underestimation of the actual cost to treat a PJI and is the most recently reported estimate. Specifically, the reported cost by Kamath et al. (2015) did not include both stages of a two-stage exchange (which is widely considered the gold standard to treat a PJI), the cost of subsequent antibiotics, and the rate of re-revision procedures for failed clearance of infection.¹⁹ Furthermore, while the costs to treat a PJI after THA and TKA were reported by Kamath et al. (2015), we chose to use the cost reported for revision TKA for our analysis as it was the more conservative of the two estimates. This way we minimized the risk of overestimating the cost-effectiveness of our intervention. In order to address variation in cost of treatment of PJI, we also considered a wide range of hypothetical costs, from $10,000 to $400,000. Costs as high as $400,000 have been reported previously by Parisi et al. (2017), after accounting for direct costs, costs of failed treatments, and indirect costs such as lost wages.²⁰ Since there was no protected health information involved for this study, approval from the Institutional Review Board was not required.

3. Results

At our institutional cost of $8.80 and assuming a cost of PJI treatment of $27,870, tranexamic acid would be considered cost-effective if the rate of infection decreased from 3.40% to 3.368% (ARR = 0.032%) (Table 1). Alternatively, if tranexamic acid prevented 1 infection out of 3125 TJAs, its use would be economically justified. When keeping the initial infection rate constant, even the highest modeled cost of TXA, $200, led to an ARR under 1% to break even.

Table 1.

Cost-effectiveness of tranexamic acid for infection prevention in total joint arthroplasty^a

Cost of Tranexamic Acid, $ (USD)	Initial Infection Rate, %	Break-Even Infection Rate, %	ARR, %
0.50	3.40	3.398	0.002
4.00	3.40	3.386	0.014
8.80	3.40	3.368	0.032
50.00	3.40	3.221	0.179
200.00	3.40	2.682	0.718

Bolded values denote actual cost.

ARR = absolute risk reduction; USD = United States Dollar.

^aPresumes a baseline infection rate (cohort not receiving Tranexamic Acid) of 3.4% for primary TJA, and a treatment cost of $27,870 for revision TJA due to PJI.

Because the rate of PJI in patients who do not receive intraoperative TXA may vary, we also tested the effects of different initial infection rates. When cost of tranexamic acid and cost of infection treatment were held constant with varying infection rates, ARR was unchanged at 0.032% (Table 2).

Table 2.

Maintaining constant the cost of tranexamic acid and the cost of treating infection, while varying initial infection rate^a.

Initial Infection Rate, %	Break-Even Infection Rate, %	ARR, %
1.00	0.968	0.032
2.00	1.968	0.032
3.40	3.368	0.032
5.73	5.698	0.032
10.00	9.968	0.032

ARR = absolute risk reduction; USD = United States Dollar.

^aPresumes cost of 2g of intravenous TXA is $8.80, with an infection treatment cost of $27,870 for TJA.

Since the cost of treating a PJI varies, we analyzed how varying the cost of infection treatment affected the ARR, when holding the cost of tranexamic acid and the initial infection rate constant. Higher treatment costs were associated with greater cost-effectiveness of tranexamic acid (Table 3). Even at the lowest cost of infection treatment tested, $10,000, tranexamic acid would be effective if it prevented as little as 1 in 1136 infections.

Table 3.

Maintaining constant the cost of tranexamic acid and initial infection rate, while varying the cost of treating infection^a.

Cost of Treating Infection, $ (USD)	Break-Even Infection Rate, %	ARR, %
10000	3.312	0.0880
15000	3.341	0.0587
25000	3.365	0.0352
50000	3.382	0.0176
150000	3.394	0.0059
400000	3.398	0.0022

ARR = absolute risk reduction; USD = United States Dollar.

^aPresumes cost of 2g IV TXA is $8.80, with an initial infection rate of 3.40% for primary TJA.

4. Discussion

Despite the commonplace use of TXA as a hemostatic agent in TJA, the existing literature demonstrates only modest cost savings associated with its routine administration. Given the link between TXA administration and a decreased rate of PJI, we hypothesized that the use of TXA in TJA is highly cost-effective when considering it as an infection prevention intervention.

In this study we utilized break-even modeling to evaluate the economic viability of TXA to prevent PJI after total knee and hip arthroplasty. We demonstrate that the routine use of TXA is financially justified across a range of costs of TXA, baseline infection rates, and costs of PJI treatment.

Tranexamic acid has been studied extensively with respect to its properties reducing blood loss, and even its side effects at higher doses, including myocardial infarction, thromboembolic events, and seizures.^21,22 However, recent studies have also reported lower odds of PJI with the use of tranexamic acid. In multivariable logistic regression models of a population of patients undergoing revision TJA for aseptic loosening, Klement et al. (2020) found that patients who received tranexamic acid had an odds ratio of 0.47 (95% CI, 0.23 to 0.90) for subsequent PJI.⁸ Similarly, Yazdi et al. (2020) reached similar conclusions in their study of tranexamic acid in the context of primary total joint arthroplasty.¹⁰ In the non-tranexamic acid group 90 (3.4%) patients had subsequent PJI, whereas only 60 (1.6%) patients in the tranexamic acid group later developed PJI (P < 0.001). Recently, Hong et al. (2020) analyzed outcomes of nearly one million arthroplasty surgeries and found that the adjusted odds ratio of PJI within 90 days of surgery with TXA use was 0.49 (95% CI, 0.69 to 0.91).²³ While all three studies utilized multivariate models, it is important to note that the three studies were observational cohort studies and may still be limited by residual confounding. Nonetheless, the authors of these studies hypothesized that the decreased incidence of PJI associated with the use of TXA was related to lower rates of allogenic blood transfusions.

Apart from direct infection of blood transfusion recipients, blood transfusions may also increase susceptibility to infection through additional mechanisms.²⁴ The immunomodulatory properties of allogeneic blood transfusions have been known since 1973, when Opelz et al. (1973) found that blood transfusions improved the survival of renal allografts.²⁵ Transfusion-related immunomodulation (TRIM) – as the modulatory process came to be known – is likely related to the immunosuppressive effects of donor white blood cells, which has been demonstrated in both animal models and humans.^26,27 Although, when considered as a whole, the results of randomized controlled trials investigating TRIM have been contradictory, and the magnitude of the effect of TRIM is still debated.²⁸

While previous studies have investigated the cost-effectiveness of TXA for TJA in several different settings, none have accounted for differences in rates of PJI. After adjusting for the cost of TXA, Tuttle et al. (2014) found that its use led to savings of $87.73 per patient, just based on the cost of transfusions alone.²² The administration of TXA also had other notable benefits, including increased rates of discharge to home (instead of a skilled nursing facility) and higher post-operative hemoglobin. For preventing hemorrhage, TXA is even more cost-effective when compared to alternatives like iron supplementation (2.5 times more expensive) and erythropoietin (17 times more expensive).²⁹ Gillette et al. (2013) and Chimento et al. (2013) calculated that total initial cost savings with use of tranexamic acid were approximately $900 to $1,500^12,13 However, neither study investigated the cost savings due to reduced incidence of subsequent PJI. In our analysis, we demonstrate that TXA would need to decrease the baseline infection rate by 0.032% in order to be considered cost-effective. In other words, it would need to prevent 1 infection out of 3125 TJA. Given the absolute-risk reduction of 1.8% reported by Yazdi et al. (2020) in primary procedures, and 2.43% in patients undergoing revision for aseptic failure – the 0.032% ARR demonstrated in our study is certainly attainable.^8,10 Given results of our analysis as well as the recent AAOS/AAHKS clinical guidelines supporting its use in lower limb arthroplasty, providers should feel confident with the safety, effectiveness, and cost-savings associated with routine TXA use in both primary and revision TJA.⁹

One of the major strengths of this study was our ability to test a wide range of values for each variable of interest: cost of tranexamic acid, initial infection rate, and cost of PJI treatment. Rather than choosing a single value for each variable, our analyses better reflect real-life variation of each factor, by institution and population. Even at the extremes of each factor, tranexamic acid remained a cost-effective therapy for preventing PJIs. For example, even at our highest estimated cost for TXA ($200), the ARR necessary for cost-effectiveness (0.718%) was still below the reported ARRs for primary (1.8%) and revision TJA (2.43%).^8,10

There are several additional limitations to this study. First, while we were able to consider a variety of different values for each variable in our break-even model, the gold standard for determining the cost-effectiveness of tranexamic acid, along with its different routes of administration, is a multi-center randomized controlled trial. Second, the break-even modeling only investigated the cost-effectiveness of TXA with respect to its effect on infection. However, the primary indication for the use of TXA is intra-operative hemorrhage. Incorporating the break-even cost of multiple benefits of TXA, including costs of transfusion, into a single model would improve the accuracy of the true, overall break-even cost of TXA use. Third, future calculations of cost-effectiveness should also investigate whether TXA affects costs of other long-term complications, including periprosthetic fracture, venous thromboembolism, and aseptic loosening. Although, TXA's effect on other long-term complications may be limited. For example, in a matched outcome study of patients with a history of venous thromboembolism, there was no statistically significant difference in the risk of venous thromboembolism between patients who received TXA (2.3%) when compared to those who did not (1.8%, P = 0.6).³⁰ Even still, investigations of other long-term complications may identify additional cost savings associated with the application of TXA.

In conclusion, tranexamic acid is a highly economically justified therapy for infection prevention in primary and revision lower limb arthroplasty. Tranexamic acid is substantiated across a range of values, including different costs of tranexamic acid, infection rates, and costs of treatment. These analyses further emphasize the importance of tranexamic acid, not only as a modifier of hemodynamics, but also as an important agent for infection prophylaxis. Economic analyses, like the one presented herein, can further enhance institutional decision-making for providing effective care.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and material

All data analyzed in this study are available through pubmed.gov.

Funding

There was no separate funding for this study.

Authors’ contributions

All authors contributed to each aspect of the study design, analysis, and writing of the manuscript.

Declaration of competing interest

Dr. Kavolus declares stock or stock options with Conformis, Histogenics, Neogenomics, Nuvasive, and Vericel. No other author has anything to declare.