Ninety-Day Outcomes in Primary Hypercoagulable Disease Patients Undergoing Total Joint Arthroplasty Vs Normal: A Matched Case-Control Series

Abstract

Background

Perioperative complications of deep vein thrombosis are well described in the total joint arthroplasty (TJA) literature. Few studies have investigated short-term perioperative outcomes of patients with primary hypercoagulable diseases (PHDs). Optimal perioperative management of PHD patients remains unknown, and they are often referred to tertiary centers for care. We investigated the influence perioperative hematology consultation and anti-coagulation use had on PHD patient outcomes following TJA surgery within the 90-day postoperative period.

Methods

This retrospective cohort study examined perioperative outcomes of PHD patients undergoing TJA. Thirty-eight PHD patients were identified and compared to a 3:1 matched control group in a consecutive series of 6568 cases (2007-2019). Perioperative hematology consultations, use of anticoagulants (AC) or antiplatelet therapy, emergency department (ED) visits, readmissions, and complications within 90 days of surgery were determined.

Results

The PHD cohort exhibited more frequent hematology consultations (odds ratio 5.88, 95% confidence interval: 2.59-16.63) and AC use (odds ratio 7.9, 95% confidence interval: 3.38-23.80) than controls. PHD patients did not show significantly greater rates of deep vein thrombosis, transfusion, infection, ED visits, or need for operative intervention. Similarly, AC vs antiplatelet therapy yielded comparable ED visits and readmissions within 90 days postoperatively (11.0% vs 9.7%, P = .85 and 5.5% vs 5.5%, P = 1, respectively).

Conclusions

These findings suggest that despite increased hematology consultation and AC use, PHD patients do not demonstrate significantly elevated perioperative risks post-TJA, favoring careful preoperative workup and outpatient postoperative follow-up.

Introduction

Optimal perioperative venous thromboembolism (VTE) prophylaxis in total joint arthroplasty (TJA) remains a subject of ongoing debate. Despite the use of physical and chemical prophylaxis, deep vein thrombosis (DVT) and pulmonary embolism (PE) continue to pose significant morbidity and mortality [1]. The incidence of VTE is estimated to be between 0.5 and 1.5% of primary total hip arthroplasty (THA) and total knee arthroplasty (TKA) in the general population, although some studies have shown rates of asymptomatic DVT to be as high as 40.8% [2,3]. While perioperative outcomes are well known, the long-term morbidity of DVT and PE is underappreciated. Conditions such as post-thrombotic syndrome and chronic thromboembolic pulmonary hypertension can pose significant impairment to a patient’s health and physical function [4]. In addition, postoperative complications such as increased blood loss, perioperative transfusion needs, infection, or wound dehiscence can result in a large physical, psychological, and financial burden on patients and the healthcare system [[5], [6], [7], [8]].

Primary hypercoagulable diseases (PHDs) are conditions defined as an inherited or acquired abnormality in hemostasis, increasing the risk of VTE and carry an incidence of 0.2% in the general United States population [[9], [10], [11]]. PHDs include Factor V Leiden, antithrombin III deficiency, activated protein C resistance, antiphospholipid antibody syndrome, and lupus anticoagulant. Patients with these diseases have been shown to be predisposed to adverse THA and TKA outcomes including DVT, PE, and inpatient mortality [9,12,13]. Therefore, special attention must be paid to their VTE prophylaxis, diagnosis, and treatment. Their postoperative protocols often differ from those of standard arthroplasty patients, as they are frequently already on anticoagulants (AC). Often, these patients are referred to tertiary care centers for elevated complexity of care including perioperative hematology consultation. However, no study to our knowledge has shown that these additional referrals improve patient outcomes in the 90-day perioperative period.

We investigated the short-term perioperative outcomes of patients with PHD, specifically evaluating the use of perioperative hematology consultation, postprocedural VTE prophylaxis use, and the relationship with 90-day perioperative outcomes.

Material and methods

A retrospective, institutional review board-approved study was conducted on a consecutive series of 6658 primary THA or TKA from 2007-2019 identified by Current Procedure Terminology (CPT) code from a single tertiary referral center. Utilizing International Classification of Diseases 9/10 codes, a database query for PHDs including antithrombin III deficiency, lupus anticoagulant, prothrombin gene mutation, sickle cell disease, Factor V Leiden, antiphospholipid syndrome, hyperhomocysteinemia, primary hypercoagulable state, protein C deficiency was performed. Patient demographics, comorbidities, perioperative hematology consultation, DVT prophylaxis regimen with AC vs antiplatelet therapy, all-cause emergency department (ED) visits, and readmission in a 90-day period were collected from the electronic medical record with manual chart review.

Thirty-eight patients with PHDs were identified. These patients were then matched in a 3:1 randomized allocation according to age (±12 years), body mass index (BMI) (±7), CPT code (exact match according to primary THA/TKA), gender (exact match), and year of operation (±5 years). A total of 114 patients were identified as a matched control cohort for comparison utilizing these parameters with no statistically significant difference identified (Insert Table 1). In total, 152 patients were included for statistical analysis.

Table 1.

Demographics.

Parameter	PHD	Control	P value
Age (y)	58.9	59.6	.79
Sex, (n)	38	114	1
Male	10	30	1
Female	28	84	1
BMI	31.8	31.2	.69
Length of stay (d)	3.6	2.9	.12
TKA	42%	42%	1
THA	56%	56%	1

Statistical analysis was conducted using SPSS (Version 26.0, Armonk, NY). Univariate analysis included Student’s t-test to compare age, height, and weight among groups. A chi-squared test was used to compare all categorical variables. In cases involving less than 10 samples, a Fisher’s exact test was used instead. An alpha value < 0.05 was used as a threshold for statistical significance in this study Table 2.

Table 2.

Diseases.

Primary hypercoagulable disease	N
Antithrombin III deficiency	2
Lupus anticoagulant	6
Protein C deficiency	2
Anti-phospholipid syndrome	4
Factor V Leiden	12
Prothrombin mutation	8
Primary hypercoagulable state	4

Demographics

Matched variables between cohorts included age, BMI, CPT code, gender, and year of operation. Comparing the control group to the PHD group revealed an average age of 59.6 and 58.9 years, respectively (P = .79), BMI of 31.2 and 31.8 (P = .69), 73.7% of women in both cohorts (P = 1.00), and 42.1% of TKA (P = 1.00), with years of operation ranging from 2007 to 2019 (P = 1.00) for the control group and the PHD group, respectively. These values indicate an appropriately matched cohort series. Comparing the Charlson comorbidity index between groups, there was a significantly different number of patients with reported congestive heart failure in 21% vs 7% (P = .028), diabetes without end organ damage in 5% vs 2% (P = .048), and chronic kidney disease in 21% vs 7% (P = .028) between PHD and control cohorts, respectively. The American Society of Anesthesiologists (ASA) score differed among groups with PHD 3.1 and control 2.7 (P = .008).

Univariate analysis revealed differences among preoperative VTE prophylaxis, postoperative VTE prophylaxis, and perioperative hematology consultation among the PHD group and the control cohort Table 3.

Table 3.

Charlson comorbidity index.

Comorbid condition	PHD (%)	Control (%)	P value
AIDS	0 (0%)	0 (0%)	1
Malignancy	8 (21%)	16 (14%)	.312
CVD	11 (29%)	16 (14%)	.05
COPD	12 (32%)	21 (18%)	.112
CHF	8 (21%)	8 (7%)	.028a
Dementia	2 (5%)	2 (2%)	.260
Diabetes without end organ damage	9 (24%)	11 (10%)	.048a
Diabetes with end organ damage	6 (16%)	10 (9%)	.232
Hemiplegia	0 (0%)	1 (1%)	1
Metastatic solid tumor	0 (0%)	7 (6%)	.193
Mild liver disease	12 (32%)	18 (16%)	.056
Moderate liver disease	3 (8%)	2 (2%)	.1
MI	4 (11%)	6 (5%)	.268
Peptic ulcer	2 (5%)	5 (4%)	1
PVD	7 (18%)	14 (12%)	.415
CKD	8 (21%)	8 (7%)	.028a
Rheumatic disease	3 (8%)	10 (9%)	1

AIDS, acquired immunodeficiency syndrome; CVD, cerebrovascular disease; COPD, chronic obstructive pulmonary disease; CHF, congestive heart failure; MI, myocardial infarction; PVD, peripheral vascular disease; CKD, chronic kidney disease.

^aDenotes significant value of P < .05.

Results

Cumulatively, more PHD patients were on chemical prophylaxis preoperatively, including both anticoagulants and antiplatelet therapy (84.2% vs 27.2%, P < .001) compared to the control group. The PHD group demonstrated a greater preoperative use of ACs (55.3% vs 7.0%, P < .001). There was no statistical difference among preoperative anti-platelet use (28.9% vs 20.2%, P = .26) when compared to the control group for VTE prophylaxis (Table 4). Among the PHD group, a greater proportion of patients were placed on anticoagulants postoperatively AC 7.9 (86.8% vs 40.4%, P < .001, odds ratio [OR] 95% confidence interval [CI]: 3.38-23.80), as compared to our current institutional standard of aspirin.

Table 4.

Postoperative VTE prophylaxis use.

VTE prophylaxis	PHD	Control
Aspirin	5	64
Coumadin	27	42
Rivaroxaban	3	6
Enoxaparin	2	1
Apixaban	1	1

Univariate analysis revealed no difference in postoperative DVT or PE rates, postoperative infection rates, length of stay, postoperative transfusion rate or number of units received, ED visit within 90 days for a potentially related cause or all-cause event, or all-cause need for operative intervention within 90 days. Specifically, comparing PHD vs control, the average length of stay was 3.4 vs 2.9 days (P = .12), rate of DVT or PE within a 90-day period was 0.0% vs 1.8% (P = 1.00), rate of postoperative infection was 0.0% vs 1.8% (P = 1.0), and postoperative transfusion rate was 5.3% vs 6.1% (P = 1.0) including no difference in number of units received (P = 1.0).

ED visits within a 90-day period for pain, infection, or hemarthrosis were 13.2% vs 5.3% (P = .14), and all-cause need for operative intervention within 90 days was 5.3% vs 1.7% (P = .26) for the PHD group and control group, respectively.

Hematology consultation

Further, the PHD cohort showed greater odds of perioperative hematology consultation of 5.88 (OR 95%, CI, 2.59-16.63, 36.8% vs 7.8%, P < .001). Within the PHD cohort, patients without perioperative hematology consultation were more likely to be readmitted within 90 days (13.0%) vs those with hematology consultation (6.6%), though this was not statistically significant (P = .53). Additionally, patients without perioperative hematology consultation were more likely to visit the ED within 90 days (17.4%) vs those with hematology consultation (6.6%), but this was not statistically significant (P = .33).

Anticoagulants and antiplatelet VTE plan

All patients were pooled to report anticoagulants vs antiplatelet VTE postoperative outcomes Table 4. Among patients on AC, 2 patients were readmitted for hemarthrosis, one of whom required a return to the operating room for irrigation and debridement; another patient was readmitted for periprosthetic fracture requiring revision; and another for prosthetic joint infection requiring revision. In the antiplatelet group, one patient was readmitted for cellulitis, one for urinary tract infection, one for PE, and one for pain control. However, there was no significant difference in ED visits (11% vs 9.7%, P = .76) or inpatient readmission (5.5% vs 5.5%, P = 1). We did find significantly more all-cause revision operations among the AC group compared to the anti-platelet group (5% vs 0%, P = .04).

Discussion

The role of PHD in THA and TKA remains an area of active investigation. Conflicting evidence exists regarding postoperative risks associated with THA and TKA in PHD patients. Increased adverse outcomes including VTE, in-hospital resource utilization, prosthetic joint infection, and inpatient mortality have been reported [9,14]. Pan et al. studied primary hypercoagulable patients, comparing postoperative complications with rates in a control population. They identified increased odds of DVT in THA and TKA patients with PHD 8.34 (OR 95% CI, 5.36-12.98) and 4.71 (OR 95% CI, 3.5-6.2, respectively). Conversely, a prospective study by Joseph et al. found that PHDs such as Factor V Leiden, prothrombin, and hyperhomocysteinemia were not significant risk factors for VTE [15]. Instead, patient factors such as age, type of TJA procedure, and postoperative protocols including use of sequential compression device, surgical drain, and therapy protocols have a greater impact on VTE risk. Despite the conflicting evidence in the literature, it is notable that a higher proportion of PHD patients receive AC therapy in the perioperative period compared to antiplatelet therapy, as demonstrated by our study results showing 87% of PHD patients receiving AC compared to only 40% of controls.

Historically, the rate of in-hospital VTE after TJA has declined over time due to the use of VTE prophylactic agents. The American Academy of Orthopaedic Surgeons, the American College of Chest Physicians, and the American Society of Hematology, all advocate for the use of thromboprophylaxis in the perioperative period [16,17]. Yet, the choice of VTE prophylaxis in the standard TJA patient remains up for debate [14,[18], [19], [20]]. Matharu et al. reviewed the safety of aspirin use for VTE prophylaxis following TJA in a systematic review and meta-analysis including over 6000 patients. Their findings concluded that aspirin, while having a slightly higher relative risk of VTE at 1.12 (95% CI, 0.78-1.62), was not statistically different compared to other anticoagulant VTE prophylaxis agents [21]. Additionally, risk of wound hematoma, infection, and major bleeding did not statistically differ. However, the question remains whether this type of information can be applied to patients who are theoretically at higher risk of VTE.

Early identification of higher-risk patients may allow surgeons to make more informed preoperative decisions regarding VTE prophylaxis. Zhou et al. identified genetic variants that predispose patients to a higher risk of VTE. In their meta-analysis, the authors concluded certain molecular variants in gene mutations such as Factor V Leiden and prothrombin gene/G20210A mutation carried an increased risk of VTE in certain ethnic groups [8,22]. Bedair et al. retrospectively reviewed 1944 patients and preoperatively performed hematologic genetic testing to identify hypercoagulable state patients with an increased risk of VTE [22]. The authors identified that patients with a prior personal or family history of VTE had a 37% chance of genetic abnormality putting them at higher risk of hereditary prothrombotic events.

When treating patients with PHD, optimal postoperative AC plans are frequently complicated by pre-existing AC use or bridging therapy requirements. Generally, hematologists are consulted to risk stratify and create perioperative AC plans in patients with an increased risk of VTE such as PHD. Because of this, patients with PHD are often sent to tertiary care centers due to the availability of subspecialty staff, including hematology services.

In our study, PHD patients carried a significantly greater odds of having a hematology consultation in the perioperative period compared to controls 5.88 (OR 95%, CI, 2.59-16.63, 36.8% vs 7.8%, P < .001). However, postoperative outcomes and statistical chance did not significantly differ when comparing DVT, prosthetic joint infection rates, transfusion requirements, ED visits, or inpatient readmissions over the 90-day perioperative period among PHD patients and controls. We did note a trend toward higher rates of inpatient readmission and ED visits within 90 days, specifically within the PHD group, in patients not seen by a hematologist within the perioperative period compared to those that did, with 13.0% vs 6.6% inpatient admission and 17.4% vs 6.6% ED visit, respectively. Postoperatively, we noted a significantly higher number of PHD patients on AC compared to control (86.8% vs 40.4%, P < .001).

Our findings support that with careful preoperative evaluation and a postoperative VTE prophylaxis plan, PHD patients did not demonstrate a large, short-term perioperative outcome difference compared to the control group. This implies that hematology consultation and a VTE perioperative plan can be obtained to reduce the risk of negative postoperative outcomes to a similar level as for non-PHD patients. Practically, this may also suggest that tertiary center transfer of care for PHD patients may not be necessary if appropriate preoperative and postoperative VTE prophylaxis and hematology evaluation are performed. However, our findings may not be generalizable as both groups of patients were cared for in a tertiary care setting with possible uncaptured benefits that prevented subsequent short-term complications.

Limitations

This study is not without limitations. First, our data was collected in a retrospective fashion. Second, the rare nature of PHDs limited the sample size and thus, our ability to identify a more nuanced statistical difference among the cohorts. Patient complications may have been diagnosed and treated at an outside facility without reporting to our institution. However, patient records were independently reviewed including outpatient office visits within the 90-day perioperative period.

The authors agree that there are generalizing assumptions made when considering PHDs, as there are several types that all carry a different theoretical risk of VTE. Additionally, specific genetic polymorphisms and heterozygosity may further complicate this matter. Patients with more severe thrombophilia may have been on stronger AC as well. ASA scores, while statistically different between groups (3.1 vs 2.7, P = .008), are inherently higher secondary to PHD diagnosis compared to the control group. In addition, the authors did consider potential cofounders for other prothrombotic states such as malignancy within 1 year, but deemed this to be too difficult to control with our sample sizes.

At our institution, we routinely start antiplatelet or AC on postoperative day 1 for 4 weeks of VTE prophylaxis. Prescribing practices have changed over time, that is, ASA 325 mg daily vs ASA 81 mg twice daily in the control group, which is unable to be examined statistically in this study but could influence the results. Similarly, preoperative tranexamic acid usage has uniformly been used at our institution since 2015, with rare but variable instances of holding a dose that was not formally analyzed. Currently, our protocol includes 1 g of tranexamic acid pre-incision in nearly all arthroplasty cases.

Finally, our study focused on short-term 90-day perioperative findings. This may underestimate postoperative complications beyond the study period.

Despite these limitations, this is the first single-center study to our knowledge to report on short-term outcomes of VTE prophylaxis use and perioperative hematology consultation in the setting of PHD patients undergoing TJA.

Conclusions

Overall, PHD patients did not demonstrate a large difference in perioperative risk following TJA in the 90-day period. This questions the need for tertiary care center referrals and supports hematology consultation with close outpatient follow-up after appropriate preoperative evaluation and VTE prophylaxis planning. Large, heterogeneous randomized studies may further clarify optimal location of surgery for PHD patients.

Conflicts of interest

A. M. Boubekri is a fellow member of the American Shoulder and Elbow Society. H. Rees is an editorial/governing board member of Journal of Arthroplasty, Arthroplasty Today, Orthopedics, and PLOS One and is a board/committee member of the American Association of Hip and Knee Surgeons and the American Academy of Orthopaedic Surgeons. N. Brown is a paid consultant for Corin USA, DePuy, A Johnson & Johnson Company, and Link Orthopaedics; is an editorial board member of the Journal of Arthroplasty; and is a board/committee member of the American Association of Hip and Knee Surgeons. All other authors declare no potential conflicts of interest.

For full disclosure statements refer to https://doi.org/10.1016/j.artd.2024.101424.

CRediT authorship contribution statement

Amir M. Boubekri: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing. Michael P. Murphy: Data curation, Formal analysis, Methodology, Writing – review & editing. Nicolas Jozefowski: Writing – review & editing, Data curation, Investigation. Nicholas M. Brown: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization. Harold W. Rees: Conceptualization, Methodology, Supervision, Writing – review & editing.