1 Introduction

Venous thromboembolism (VTE), defined as either pulmonary embolism (PE) or deep vein thrombosis (DVT), is a known complication in the early post-operative period following primary total knee arthroplasty (TKA). Although overall incidence of VTE in primary TKA is low, rates of DVT in patients without chemoprophylaxis are high, ranging from 26 to 84%.^1–3 The morbidity, mortality, and stress of VTE to the health care system are significant and well established, including increased length of stay, readmissions, reoperations, and higher health care costs.⁴

The selection of a VTE prophylaxis is a balance between safety and efficacy. Although more aggressive anticoagulation has proven efficacy in VTE prevention, it is also associated with post-operative complications such as bleeding, wound complications, and need for reoperation.^5,6 While all patients undergoing primary TKA are recommended to use VTE prophylaxis, no specific drug regimen or universal standard of care currently exists.⁷

Aspirin has become a common agent used in the prevention of VTE in patients undergoing primary TKA. Its use is well studied and efficacy is also well demonstrated.^8,9 However, optimal dose and frequency of Aspirin is still debated, as no consensus exists. Recent literature has supported the use of low dose Aspirin in patients undergoing primary TKA, finding it non-inferior to higher dose Aspirin with the theoretical benefit of also providing fewer adverse events associated with higher dose Aspirin.¹⁰ To our knowledge, there is only one study that has shown that equal effect of low dose Aspirin compared to high dose Aspirin in preventing VTE after TKA.¹¹

The purpose of our study was to determine the impact of an anticoagulation protocol change in a large academic medical center on VTE rates in patients undergoing primary TKA. We hypothesized that there would be no difference in rates of VTE occurrence in the post-operative period between a cohort receiving low dose Aspirin versus a cohort receiving high dose Aspirin.

2 Methods

A retrospective chart review of all primary TKA cases performed at a single academic institution was conducted. An institutional TKA database was queried by (CPT) Codes to identify all primary TKAs. The study period was October 2011–January 2019. The CPT code utilized were 27447 and 27446 (TKA). All patients entered into our clinical system with a history of thromboembolic event were identified by International Classification of Diseases Tenth Revision (ICD-10) codes I82.40 or I82.90. Chart review was performed to confirm that the diagnosed thromboembolic event took place after the date of surgery and within 90-days post-operatively. A total of 12,866 patients were included in the study.

Patients were divided into two cohorts; those who received primary TKA between to October 2011 to August 2017 (Cohort 1) and those who received primary TKA from September 2017 to January 2019 (Cohort 2). These dates were selected as they correspond with our institution's protocol change from high dose (325 mg) Aspirin twice per day (BID), to low dose (81 mg) Aspirin BID.

Patients were excluded if they had a prior history of a thromboembolic event, active cancer, BMI >40, current smokers, or treated with chemoprophylaxis agents other than Aspirin. Patients already on chemoprophylaxis treatment were continued on their medication and therefore were excluded.

Patients in Cohort 1 (October 2011–August 2017) were instructed to take 325 mg enteric coated Aspirin the evening before surgery, and to continue taking it twice daily for 28 days post-operatively. Patients in Cohort 2 (September 2017–January 2019) were instructed to take 81 mg enteric coated Aspirin the evening before surgery, and to continue taking it twice daily for 28 days post-operatively. Patients in both cohorts were discharged with sequential pneumatic compression devices (SPCDs) and instructed to wear them for 20 hours per day for 28 days. The post-operative protocols were identical except for the lowered dose of enteric coated Aspirin.

Patients in both cohorts received the same peri-operative care. There were no differences in standard anesthesia, physical therapy, or inpatient pain medication received by patients in the two cohorts. Institution-wide VTE monitoring was utilized; there was no regular screening or surveillance for VTE. Patients with clinical signs or symptoms of DVT or PE received either duplex ultrasonography or spiral computer tomography (CT) PE protocol scan, respectively. Post-operative follow-up care varied per surgeon preference, however this generally entailed an initial two week follow up visit as well as another visit at six weeks.

Categorical data were reported as counts (%). Statistical analysis between the cohorts was performed using chi-squared test for categorical variables. All statistical analyses were performed using SPSS v25 (IBM Corporation, Armonk, New York). Results were deemed to be significant at p < 0.05 throughout.

3 Results

A total of 12,866 patients underwent primary TKA between the dates of October 1, 2011 and January 31, 2019 who were deemed low risk and therefore given the standard institution-wide VTE prevention protocol with Aspirin. In cohort 1 (October 2011–August 2017), 9413 primary TKAs were performed, for which patients were prescribed high dose (325 mg BID) Aspirin. In cohort 2 (September 2017–January 2019), 3453 primary TKAs were performed, for which patients were prescribed low dose (81 mg BID) Aspirin.

There were 122 cases of VTE diagnosed in Cohort 1, for a rate of 1.29% of cases. Of those 122 cases, 114 cases were due to PE, five cases were due to DVT, and three cases were due to both PE and DVT. There were six cases of VTE diagnosed in Cohort 2, for a rate of 0.17% of cases. Of those six cases, four were due to PE, one was due to DVT, and one was due to both PE and DVT.

The incidence of symptomatic VTE in Cohort 2 was found to be statistically significantly lower than in Cohort 1 (0.17%–1.29%, p < 0.001). Regarding DVT and PE separately, the incidence of symptomatic DVT in Cohort 2 was found to be statistically similar to Cohort 1 (0.03%–0.05%, p = 0.57). The decreased incidence of PE in Cohort 2 was also found to be statistically significant (0.12%–1.21%, p < 0.001). Results are summarized in Table 1.

4 Discussion

Venous thromboembolism is a known complication after TKA, and given its potentially severe consequences, prophylaxis for VTE is a well-established standard of care.^7,8 Venous thromboembolism prophylaxis choice, like all perioperative management, is a balance between risks and benefits. More aggressive anticoagulation has proven efficacy in limiting VTE but also comes at the cost of increased incidence of post-operative complications, including gastrointestinal (GI) bleeds, surgical site bleeding, wound complications, and reoperation.^5,6 Our retrospective study analyzed two doses of Aspirin for VTE prophylaxis in a total of 12,866 patients. We found that a dose of 81 mg of Aspirin BID for 28 days post-operatively had lower rates of VTE than a dose of 325 mg taken on the same schedule. The results of our study add to the growing body of evidence that low dose Aspirin is sufficient for VTE prophylaxis. This study was conducted in order to review our institution's protocol change from 325 mg 81to mg of Aspirin BID for standard risk patients.

The efficacy and mechanism by which Aspirin works to prevent thromboembolic events is well-understood and studied. Aspirin irreversibly inhibits thromboxane A2 formation via the cyclooxygenase 1 (COX-1) pathway, which prevents platelet aggregation. It has been shown in patients undergoing carotid endarterectomy that rates of myocardial infarction, stroke, and death were lower in patients given low versus high doses of Aspirin.¹² It is hypothesized that higher doses of Aspirin may in fact have a paradoxical effect on platelet aggregation via a separate mechanism than the thromboxane A2 pathway. A proposed mechanism for this effect is as follows: secretion of prostacyclin (PGI-2) also limits platelet aggregation and is synthesized by endothelial cells. In addition to the previously mentioned thromboxane A2 inhibition, higher doses of Aspirin also inhibit cyclooxygenase 2 (COX-2) mediated PGI-2 production, which may increase platelet aggregation.¹³ Continued research is needed to solidify an ideal dose and frequency of Aspirin for VTE prevention.

As with all retrospective studies, ours has its limitations. Firstly, it relied on electronic medical records (EMR) to accurately identify and assign patients into cohorts. Errors in documentation or coding could lead to patients that were missed. Secondly, our study did not look at adverse effects of Aspirin, in particular GI bleeding. Although EMR has made it more realistic to chart review and identify complications, many of our patients receive non-orthopedic care at other institutions; we felt that chart reviewing for adverse effects of Aspirin would miss too many occurrences and therefore be inaccurate. Thirdly, a potential limitation of our study was our protocol for diagnosing VTE, more specifically not screening all patients. Diagnostic tests were only ordered if patients had signs or symptoms of VTE. While this likely missed some instances of non-symptomatic VTE, the purpose of our study was to determine the effect of a complication of TKA that affected patient care; therefore, it is reasonable to assume that any undiagnosed asymptomatic VTE events do not undermine the intent nor the conclusions from the study. This is consistent with the AAOS guidelines, which strongly recommends against routine screening for post-operative VTE in all patients.⁷

Another potential limitation of our study was the 8-year time course over which the data was collected. Although there were no other institution-wide major protocol changes, there were undoubtedly advancements made in peri-operative patient care from 2011 to 2019. There were no major changes in type of anesthesia used, medication order sets, post-operative physical therapy regimens, or sedating pain medications. However, it is reasonable to assume that growing experience for anesthesiologists, physical therapists, and nurses helped patients ambulate earlier, which may have played a factor in lower rates of VTE. For example, with the growing concern and awareness for the opioid epidemic in our country and improved opioid-sparing pain control measures, patients may be less sedated peri-operatively and therefore able to ambulate sooner.

Our study looked at patients of multiple surgeons, which is both a limitation and a strength. It is a limitation in that there were inherently variations in surgical approaches, types of implants used, and attending preferences. This variability is accounted for given the large study population and multiple surgeons.

To our knowledge, our study is the largest study identifying post-operative VTE rates for patients undergoing primary TKA and treated with varying dosages of Aspirin. It strengthens the growing body of evidence that low dose Aspirin is sufficient for VTE prevention in standard risk patients. Recently, Parvizi et al. performed a prospective study looking at 4651 patients with low versus high dose Aspirin and Faour et al. published a retrospective review of 5666 patients, both demonstrating that low dose Aspirin is non-inferior to higher doses.^10,11 A post-hoc analysis was conducted and showed that our results were powered to detect a difference in PE rates and total VTE rates (P = 1.0 for both PE and VTE rates).

5 Conclusions

This retrospective study of over 12,000 patients demonstrates that low dose Aspirin given twice per day for 28-days post-operatively is sufficient for prevention of VTE in low-risk patients undergoing primary TKA, as compared to high dose Aspirin. It supports our institution's decision to change our protocol from 325 mg of Aspirin BID to 81 mg BID. Further studies are needed to optimize dosages and timing of Aspirin as well as other, non-chemoprophylaxis, forms of VTE prevention measures for improving patient care following TKA.