Risk of Postoperative Complications After Total Hip and Total Knee Arthroplasty in Behcet Syndrome Patients

Abstract

Background:

Behcet syndrome (BS), a multisystem autoimmune disorder, has unclear effects on outcomes after total hip arthroplasty (THA) and total knee arthroplasty (TKA). This study assessed the relative risk of perioperative adverse events in patients with BS.

Methods:

This retrospective cohort study used the PearlDiver M157Ortho data set, a large national administrative database. Total hip arthroplasty and TKA patients with BS were identified and matched 1:4 to those without BS based on patient age, sex, Elixhauser Comorbidity Index scores, and procedure performed (THA or TKA). The incidence of 90-day adverse events was determined and compared by multivariate analysis. 5-year survival to revision surgeries was assessed and compared with the log-rank test.

Results:

After matching, 282 THA/TKA patients with BS were identified and compared with 1127 without BS. On multivariate analysis, patients with BS were at independently greater risk of aggregated any (odds ratio [OR] 2.16, P < 0.0001), serious (OR 1.78, P = 0.0051), and minor (OR 2.39, P < 0.0001) adverse events compared with those without BS. No significant difference was observed in 5-year survival to revision surgery (P = 0.3).

Conclusions:

Patients with BS undergoing THA or TKA experienced markedly greater 90-day postoperative adverse events. The findings underscore the need for optimized perioperative management for patients with BS undergoing arthroplasty.

Total hip arthroplasty (THA) and total knee arthroplasty (TKA) are frequently performed orthopaedic procedures to alleviate pain and restore function in patients with end-stage osteoarthritis and other joint disorders.¹ Although these surgeries are generally associated with high success rates and patient satisfaction, they are not without risks, particularly in patients with underlying systemic conditions.²

Behcet syndrome (BS) is a multisystem autoimmune disorder that is associated with immune dysregulation, cardiopulmonary issues, and vascular dysfunction.^3,4 Specifically, the immune dysregulation in Behcet is best characterized by small and large vessel vasculitis, leading to widespread tissue damage.⁵ Notably, joint involvement is seen in up to 50% of patients with Behcet, leading to recurrent migratory arthritis, synovitis, and joint degradation—most commonly of the knee.⁵ The diagnosis of Behcet is challenging, relying primarily on the presence of multiple clinical findings, including recurrent oral ulcers, genital ulcers, uveitis, unexplained vascular pathology, and several other characteristic symptoms.⁵ A positive pathergy test, where a skin prick results in a small pustule, can also help with the diagnosis.⁵ The complex nature of BS, coupled with the immunosuppressive medications used to manage it, may predispose patients to increased perioperative adverse events after THA and TKA.⁶

Multiple studies have explored the risk factors of adverse events after orthopaedic surgeries, with factors such as age, comorbidity burden, and specific conditions like diabetes mellitus being associated with increased risks.^7,8 Similarly, systemic inflammatory conditions such as rheumatoid arthritis (RA) have been linked to higher rates of complications after THA, TKA, and other orthopaedic procedures.^9,10

Behcet syndrome, although less common than RA, presents a similar challenge in the orthopaedic setting, given the disease's multisystem involvement and the potential side effects of treatments.^3,4 For reference, in the United States, Behcet affects approximately 3 to 5 per 100,000 patients, compared with 500 to 1000 per 100,000 for RA.⁵ Notably, the incidence of Behcet is much higher in certain regions of the world, such as Mediterranean countries, particularly Turkey; the Middle East; Central Asia; and Japan, where up to 400 per 100,000 are affected.¹¹ The specific effect of BS and the medications used in its treatment, such as steroids, immunosuppressants, biologics, and anti-inflammatories,¹² on postoperative outcomes after THA and TKA is not well understood, and existing studies have not adequately addressed this gap. Existing studies include case reports of cardiopulmonary complications after TKA in BS and small non-US database studies that do not evaluate outcomes after THA in patients with BS.^13,14 Given this, understanding the relationship between BS and postoperative outcomes is of importance, particularly in high-prevalence regions, because patients with BS should be appropriately counseled and clinicians must be aware of potential risks to implement effective risk-mitigation strategies.

Thus, using the PearlDiver M157Ortho data set, this study aimed to evaluate the relative risk of 90-day adverse events and 5-year survival to revision surgery in patients with BS undergoing THA or TKA compared with those without BS. This work builds on the growing trend of leveraging large national databases to evaluate nuanced patient populations, providing insights that may not be attainable through localized studies.^10,15-17

Methods

Database and Cohort

This study used the 2010-2022 PearlDiver M157Ortho data set (Colorado Springs, Colorado). This is a large Health Insurance Portability and Accountability Act report national billing claims data set containing information on over 157 million orthopaedic patients in the United States across all payer types and sites of care. All PearlDiver data are output deidentified and aggregated at the user interface level; thus, our institutional review board deemed studies using this database exempt from review.

Current Procedural Terminology codes 27447 and 27130 were used to identify patients who underwent primary TKA or primary THA, respectively. Patients who had any diagnosis related to trauma, neoplasm, or infection on the day of surgery were excluded from the study cohort.

International Classifications of Disease codes were used to isolate TKA and THA patients with and without Behcet syndrome (BS). Several patient demographics were abstracted, including age, sex, and Elixhauser Comorbidity Index (ECI, a longitudinal measure of patient comorbidity burden generated using International Classifications of Disease codes). TKA/THA patients with and without BS were subsequently matched in a 1:4 ratio based on age, sex, ECI, and procedure performed (TKA or THA) using the PearlDiver system.

Identification of Adverse Events/Survival to Revision Surgery

90-day postoperative adverse events were identified based on International Classifications of Disease codes, as previously described.^{10,16,18,19,20,21,22,23,24,25,26} Aggregated adverse event groups included serious adverse events (pulmonary embolism, surgical site infection, deep vein thrombosis, and sepsis), minor adverse events (wound dehiscence, hematoma, acute kidney injury, transfusion, UTI, and pneumonia), or any adverse events (any serious or minor adverse event).

5-year revision surgery rates were then defined based on the occurrence of revision codes after index TKA/THA procedures. Survival to revision surgery was determined for matched cohorts of TKA/THA patients with Behcet and without Behcet and compared between the 2 groups.

Data Analysis

Univariate analysis was conducted to compare demographics of arthroplasty patients with and without BS both before and after matching. The Welch t-test was used to evaluate for statistically significant differences in patient age and ECI between groups while the Pearson chi-squared test was used to compare differences in patient sex and procedure performed. A post hoc power analysis was conducted for each outcome, and an average power was reported.

A multivariate logistic regression model controlling for patient age, sex, ECI, and procedure performed (TKA or THA) was used to determine the odds ratios (ORs) and 95% confidence intervals of experiencing adverse events in the 90 days after arthroplasty surgery in the matched cohort of patients with BS compared with those without BS.

5-year survival to revision surgery was compared between matched cohorts of arthroplasty patients with and without BS using Kaplan-Meier survival analysis. Log-rank analysis was conducted to evaluate for a statistically significant difference in 5-year survival between the 2 cohorts.

All statistical analyses were conducted using the in-built PearlDiver system. Statistical significance was designated as P < 0.05. Prim9 (GraphPad Software) and Microsoft Excel (Microsoft Corporation) were used to create all figures.

Results

Study Cohort Demographics

A total of 2,119,069 arthroplasty patients were identified in the data set from 2010 to 2021, of which 1,437,208 (67.8%) underwent TKA and 681,579 (32.2%) underwent THA. Of this group, 282 patients (0.013%) had a diagnosis of BS, among which 171 (60.6%) underwent TKA and 111 (39.4%) underwent THA (Table 1). Univariate analysis revealed that arthroplasty patients with BS were younger (average age 57.0 vs. 65.4, P < 0.0001), more female predominant (78.7% vs. 60.4%, P < 0.0001), and more comorbid (average ECI 7.03 vs. 4.26, P < 0.0001) and comprised a larger proportion of THA patients (39.4% vs. 32.2%, P = 0.0117).

Table 1.

Univariate Analysis of Unmatched and Matched Total Hip and Knee Arthroplasty Patients With and Without Behcet Syndrome

	Unmatched			Matched
	TKA/THA No Behcet	TKA/THA Behcet	P	TKA/THA No Behcet	TKA/THA Behcet	P
N	2,118,787 (98.7%)	282 (0.013%)		1127 (80.0%)	282 (20.0%)	0.8155
Age (mean ± std dev)	65.4 ± 9.3	57.0 ± 11.5	<0.0001	57.1 ± 11.5	57.0 ± 11.5
Sex
Female	1,280,616 (60.4%)	222 (78.7%)	<0.0001	888 (78.8%)	222 (78.7%)	1.0000
Male	838,167 (39.6%)	60 (21.3%)		239 (21.2%)	60 (21.3%)
ECI (mean ± std dev)	4.26 ± 3.26	7.03 ± 4.11		7.02 ± 4.09	7.03 ± 4.11	1.0000
1-3	1,025,728 (46.9%)	53 (18.8%)	<0.0001	232 (20.6%)	53 (18.8%)
4-6	643,173 (29.4%)	85 (30.1%)		340 (30.2%)	85 (30.1%)
7-9	291,325 (13.3%)	66 (23.4%)		264 (23.4%)	66 (23.4%)
10+	158,561 (7.2%)	78 (27.7%)		291 (25.8%)	78 (27.7%)
Procedure
TKA	1,437,208 (67.8%)	171 (60.6%)	0.0117	747 (66.3%)	171 (60.6%)	0.0874
THA	681,579 (32.2%)	111 (39.4%)		380 (33.7%)	111 (39.4%)

THA = total hip arthroplasty, TKA = total knee arthroplasty, ECI = Elixhauser Comorbidity Index

Bold entries represent statistically significant P values at P < 0.05.

After matching arthroplasty patients with and without BS 1:4 based on age, sex, ECI, and procedure performed, 1127 patients (80.0%) without BS were identified and compared with 282 (20.0%) with BS. Univariate analysis no longer showed any notable difference in patient age, sex, ECI, or procedure performed between the matched groups.

Incidence of Adverse Events

The incidence of any, serious, minor, and individual adverse events among matched arthroplasty patients with and without BS is listed in Table 2. A 90-day adverse event was noted for 94 (33.3%) of the 282 matched patients with BS, compared with 221 (19.6%) of the 1,127 matched patients without BS. The difference in the proportion of patients with and without BS who experienced any adverse event was statistically significant (P < 0.0001). Univariate analysis also showed patients with BS to have a higher overall incidence of serious adverse events at 14.9% versus 9.2% (P = 0.0053) (specifically sepsis [5.7% vs. 2.1%, P = 0.0013] and pulmonary embolism [3.9% vs. 1.4%, P = 0.0065]) and minor adverse events at 28.0% versus 14.6% (P < 0.0001) (specifically pneumonia [8.2% vs. 2.0%, P < 0.0001], UTI [17.4% vs. 5.5%, P < 0.0001], and hematoma [specific percentage not available, P < 0.0001]; Table 2).

Table 2.

Univariate Incidence of Adverse Events After THA/TKA in Matched Patients With and Without Behcet Syndrome

	THA/TKA No Behcet (N = 1,127)	THA/TKA Behcet (N = 282)	P
Any adverse events	221 (19.6%)	94 (33.3%)	<0.0001
Serious adverse events	104 (9.2%)	42 (14.9%)	0.0053
Sepsis	24 (2.1%)	16 (5.7%)	0.0013
Pulmonary embolism	16 (1.4%)	11 (3.9%)	0.0065
Deep vein thrombosis	36 (3.2%)	14 (5.0%)	0.1499
Surgical site infection	31 (2.8%)	<11	0.4777
Minor adverse events	164 (14.6%)	79 (28.0%)	<0.0001
Pneumonia	22 (2.0%)	23 (8.2%)	<0.0001
Urinary tract infection	62 (5.5%)	49 (17.4%)	<0.0001
Wound dehiscence	22 (2.0%)	<11	0.1074
Acute kidney injury	37 (3.3%)	16 (5.7%)	0.0588
Hematoma	12 (1.1%)	<11	0.0026
Transfusion	40 (3.5%)	<11	1.0000

THA = total hip arthroplasty, TKA = total knee arthroplasty

^aPearlDiver does not provide exact numbers of patients when an outcome is experienced by < 11 patients to preserve patient anonymity.

Bold entries represent statistically significant P values at P < 0.05.

The results of multivariate logistic regression analysis controlling for patient age, sex, ECI, and procedure performed (TKA or THA) are presented in Table 3 and Figure 1. This model found patients with BS to have significantly greater odds of experiencing any adverse event (odds ratio [OR] 2.16, P < 0.0001), serious adverse events (OR 1.78, P = 0.0051), and minor adverse events (OR 2.39, P < 0.0001).

Table 3.

Multivariate Analysis of Risk of Adverse Events After TKA/THA in Patients With Behcet Versus Those Without Behcet

N = 1409	OR (95% CI)	P
Any adverse events	2.16 (1.59, 2.91)	<0.0001
Serious adverse events	1.78 (1.18, 2.64)	0.0051
Sepsis	3.03 (1.50, 6.01)	0.0016
Pulmonary embolism	2.93 (1.29, 6.45)	0.0083
Deep vein thrombosis	1.61 (0.82, 2.98)	0.1432
Surgical site infection	1.03 (0.43, 2.19)	0.9365
Minor adverse events	2.39 (1.73, 3.28)	<0.0001
Pneumonia	4.76 (2.55, 8.94)	<0.0001
Urinary tract infection	3.71 (2.46, 5.58)	<0.0001
Wound dehiscence	1.87 (0.83, 3.93)	0.1120
Acute kidney injury	1.81 (0.95, 3.34)	0.0619
Hematoma	1.64 (0.51, 4.50)	0.3629
Transfusion	0.98 (0.45, 1.92)	0.9540

THA = total hip arthroplasty, TKA = total knee arthroplasty

Bold entries represent statistically significant P values at P < 0.05.

Figure 1.

Plot demonstrating risk of adverse events after arthroplasty surgery (THA/TKA) in patients with Behcet syndrome versus those without BS. Black bars represent statistical significance, and gray bars indicate no statistical significance. THA = total hip arthroplasty, TKA = total knee arthroplasty

Among adverse events classified as serious, patients with BS had greater odds of experiencing sepsis (OR 3.03, P = 0.0016) and pulmonary embolism (OR 2.93, P = 0.0083) compared with those without BS. Among adverse events classified as minor, patients with BS had greater odds of experiencing pneumonia (OR 4.76, P < 0.0001) and UTI (OR 3.71, P < 0.0001). Patients with Behcet syndrome did not have markedly greater odds of experiencing postoperative surgical site infection, deep vein thrombosis, wound dehiscence, hematoma, acute kidney injury, or transfusion.

Survival to Revision surgery

The results of Kaplan-Meier analysis for 5-year survival to revision after arthroplasty among patients with and without BS are shown in Figure 2. 5-year survival (standard error) among patients with BS was 95.1% (1.8%) compared with 94.3% (0.81%) among patients without BS. A log-rank test revealed no statistical difference in the rate of 5-year survival to revision surgery between the 2 groups (P = 0.3).

Figure 2.

Plot demonstrating Kaplan-Meier 5-year survival analysis to revision surgery after arthroplasty (THA/TKA) among patients with and without Behcet demonstrating. THA = total hip arthroplasty, TKA = total knee arthroplasty

Power Analysis

Post hoc power analysis revealed the average power for the study to be 0.556, suggesting moderate overall ability of this study to detect differences in adverse events between TKA/THA patients with and without BS.

Discussion

At present, there is a scarcity of clinical research examining surgical outcomes in patients with BS undergoing THA or TKA. Existing studies include a single case report demonstrating a patient with BS who experienced massive pulmonary embolism after THA¹⁴ and a non-US database study with only 1 year of follow-up that examined outcomes after TKA in approximately 300 patients with BS.¹³ The purpose of this study was to determine the correlation of BS with short-term complication rates and longer term implant survival after TKA and THA. After controlling for patient age, sex, comorbidity burden, and procedure performed (THA or TKA), our data suggest that patients with BS are at increased risk of developing postoperative pneumonia, sepsis, UTI, and pulmonary embolism. Notably, there was no statistically significant difference in 5-year survival to revision surgery after THA/TKA between patients with and without BS.

In our extensive study involving a large cohort of over 2 million patients undergoing THA and TKA, we observed that a small fraction had BS, representing just 0.013% of the study population. This incidence rate is in line with previously reported incidence rates of BS in the United States, ranging from 0.005% to 0.011%.^27,28 Slight differences in incidence could be attributed to various factors, such as variability in the specific ethnicities of the patient population under study—with those who have Mediterranean, Middle Eastern, Central Asian, or Japanese heritage having much higher rates of BS than other populations.¹¹ Importantly, the small proportion of patients with this condition in our study exemplifies the critical need for using large-scale databases to accurately study condition-specific outcomes. Such expansive data sets allow for a more robust statistical analysis, thereby providing more reliable conclusions that can inform clinical practice.

Once the populations were matched and variables controlled for, there were differences in perioperative outcomes for those with versus without BS. Of the adverse outcomes analyzed, BS patients had greatest increased odds of experiencing postoperative pneumonia. This aligned with known irreversible parenchymal lung damage associated with patients with BS that predisposes them to pneumonia.^29,30 Importantly, the chronic use of immunosuppressive medications may also compound the risk of pneumonia in BS.^31,32

The increased odds for UTI for those with BS may align with genital ulcers that are a well-documented complication of BS.^32,33 This UTI association was similarly reported in patients with Behcet undergoing orthopaedic surgery in general, as published by Zhang et al.¹³ The long-term use of immunosuppressive therapies in this patient population may also contribute to the increased rates of pneumonia, UTI, and subsequently sepsis seen in this patient population.^34,35

Pulmonary embolism was the other individual perioperative adverse outcome associated with BS. There already exists a body of literature highlighting the pulmonary vascular disease as a clinical feature of BS.^36-38 Pulmonary artery involvement is the most common manifestation of arterial involvement in BS causing any number of complications including pulmonary artery aneurysm, pulmonary artery thrombosis, and subsequent pulmonary embolism.^39,40 While BS generally has a predilection for the venous rather than the arterial system, pulmonary arteries are an exception to this general rule because they resemble venous structures due to their thinner walls, lower elasticity, and lower pressures.⁴⁰ Notably, active immunosuppression has been found to reduce the risk of pulmonary emboli in patients with BS, with the caveat that long-term corticosteroid use has been associated with increased risk of venous thromboembolism.^41,42 As such, patients with BS may benefit from an immunosuppressive regimen, but this must be balanced by the other adverse outcomes discussed above.

This study revealed no difference in the rate of revision surgery within a 5-year period after THA or TKA, indicating that while patients with BS carry a markedly higher risk of short-term complications, their risk on long-term issues are similar to those without the condition. Notably, the complications seen in this population seem to be largely medical rather than surgical. This is encouraging that after mitigation of short-term complications, longer term implant survival may not be affected.

This study has several limitations. Notably, administrative data are limited based on the accuracy of coding. Additional administrative data do not provide patient-reported, surgeon-reported, or radiographic outcome measures; thus, these clinically relevant measures are unavailable in this study. In addition, medical therapies administered during the perioperative period including immunosuppression or anticoagulation were not controlled for in the analysis. Finally, surgical approach, hip-specific outcome measures, and adverse surgical outcomes were not assessed.

In conclusion, patients with BS undergoing TKA/THA have increased risk of postoperative complications including pulmonary embolism, pneumonia, and UTI. Increased care must be taken during the perioperative period to mitigate these risks. Specifically, given the greater risk of pulmonary embolism, patients with BS undergoing TKA/THA may be considered for stronger venous thromboembolism prophylaxis. Furthermore, given the greater risk of infection in patients with BS, notably sepsis, pneumonia, and UTI, likely due to immunosuppressive medications used to treat the disease, holding these medications in the perioperative period could be considered. In addition, close postoperative monitoring is required in this group to detect early signs of infection. Ultimately, while there is an increased risk of postoperative complications, the longer term similar revision rates should be encouraging to those with this condition and those caring for them.