Racial Disparities in Total Ankle Arthroplasty Utilization: A National Database Analysis

Abstract

Background: Total ankle arthroplasty (TAA) is successful at reducing pain and improving patient satisfaction. A paucity of literature exists regarding racial disparities in TAA. The aim of this study was to update the literature, analyze utilization rates, and detect differences in postoperative outcomes between the racial/ethnic groups. Methods: A retrospective study was performed utilizing the National Surgical Quality Improvement Program (NSQIP) database between the years 2012 and 2018. The postoperative complications were identified as outcomes. Patients were categorized based on race/ethnicity for comparison. A p-value less than 0.05 was considered significant. Results: 1164 patients met criteria, 1051 (90.3%) were White, 113 (9.7%) were Non-White, a nearly 10-fold difference in utilization rate. The mean age of White patients undergoing TAA was older than Non-White patients, 63.7 and 55.3 years respectively (p-value < 0.01). Incidence of postoperative complications showed no significant differences based on racial group. Conclusions: Our results found a nearly 10-fold difference in White patients undergoing TAA compared to other racial/ethnic groups. This highlights the continued racial disparities present in TAA. Further efforts are needed to improve the proportion of minority populations who undergo TAA to bridge the current racial disparities present in the field of TAA.

1. Introduction

Ankle osteoarthritis (OA) is a painful condition that is estimated to occur in 6% of the general population [1]. Pain and functional deficits associated with OA can pose a significant burden to patients, leading to limited mobility, reduced quality of sleep, fatigue, depressed mood, and loss of independence [2,3,4,5,6]. Furthermore, patients with ankle OA have been associated with considerable socioeconomic burden, with some studies estimating the overall cost of healthcare due to OA is equivalent to 1.0–2.5% of gross domestic product in developed countries [7]. The overall prevalence of OA is only predicted to increase in the future, as increased patient age, rising rates of obesity, and high rates of traumatic injury are estimated to contribute to increased incidence of OA annually [8,9,10,11].

As the prevalence of OA has continued to grow over time, the utilization of total ankle arthroplasty (TAA) as a therapeutic intervention of OA has similarly increased. Compared with ankle arthrodesis, TAA offers superior outcomes in regards to range of motion, functional scores, and overall patient satisfaction, leading to growing rates of utilization for treatment of end-stage OA. As a result, annual volumes of TAA have grown at a significant rate, increasing by 670% over the course of a decade [12]. In general, TAA has proven to be very reliable and successful procedure in regards to pain reduction and patient satisfaction, with studies citing increases in American Orthopaedic Foot & Ankle Society Ankle-Hindfoot Scale (AOFAS) and short form healthy survey (SF-36) scores and decreases in visual analog scale for pain (VAS), Short Musculoskeletal Function Assessment (SMFA) function index, and SMFA bother index scores [13,14,15]. Though historically, indications for TAA were mainly limited to an older patient with low functional demands, minimal deformity, and isolated ankle arthritis, improvements in surgical technique and implant design have allowed for expanded indications across a larger, more complex patient demographic [16,17,18]. In the current era, TAA is indicated for any patients with end-stage ankle OA who has sufficient vascularized bone stock and adequate soft tissue coverage [19,20]. In patients who desire greater ankle mobility or have degenerative changes in surrounding joints, TAA may provide greater benefits compared to ankle arthrodesis [21,22]. However, although patient indications for TAA have expanded and utilization rates have increased, it is uncertain if there are disparities in utilization and subsequent outcomes of TAA across patient demographics. As a result, further investigations are needed to distribution of demographics in patients who undergo TAA.

The prevalence of racial and ethnic disparities associated with utilization rates, postoperative complications, and patient outcomes in the United States is an important topic of concern within orthopedic literature, particularly in association with total joint arthroplasty (TJA). While the underlying causes leading to disparity in utilization rates across the patient population are uncertain, various factors such as socioeconomic status, patient knowledge, patient preference, willingness to undergo arthroplasty, patient expectation of post-arthroplasty outcome, religion/spirituality, and physician-patient interaction have been explored previously in the literature. Previous studies assessing patient demographics in TAA have discovered a striking disparity in utilization rates of TAA across racial and ethnic groups. However, further investigations are warranted to fully assess the degree of disparity in utilization rates and outcomes in TAA across all racial and ethnic populations. As a result, the purpose of this study was to compare utilization rates of TAA and analyze differences in postoperative outcomes between patient racial and ethnic groups using a large national patient database. We hypothesize that these racial disparities will be evident in the utilization rates and postoperative outcomes of TAA in the United States, and we seek to consider underlying factors that may drive these disparities, including access to healthcare, patient-physician communication, and hospital setting.

2. Materials and Methods

A retrospective cohort study was performed utilizing the National Surgical Quality Improvement Program (NSQIP) database. The NSQIP database is composed of nationally sourced, prospectively collected, patient level aggregate data, which is collected for the purpose of advancing the quality of care provided to surgical patients. The NSQIP database collects data from a large range of participating hospitals and includes 30-day postoperative outcomes for each patient. This database also tracks preoperative variables such as patient demographics, comorbidities, surgical history, as well as perioperative data such as operative time, length of stay, and postoperative complications.

To create our cohorts for analysis, patients were identified for inclusion using the Current Procedural Terminology (CPT) codes 27700 and 27702, corresponding to arthroplasty ankle and arthroplasty ankle with implant, respectively. All patients who underwent a primary TAA between January 2012 and December 2018 were initially included. Patient records with identified racial and/or ethnic identity were selected for analysis, which included identities of non-Hispanic White, non-Hispanic Black, Asian, Native American, and Hispanic. Two patient cohorts were created for analysis: non-Hispanic White and non-White (non-Hispanic Black, Asian, Native American, and Hispanic). A total of 1164 patients met inclusion criteria for this study. Once our patient cohorts were identified, we queried the database for 30-day postoperative complications to compare differences in patient outcomes. The complications that were assessed included superficial wound infection (Yes or No), deep wound infection (Yes or No), wound dehiscence (Yes or No), pulmonary embolism (Yes or No), deep vein thrombosis (Yes or No), blood transfusion (Yes or No), unplanned reoperation (Yes or No), and readmission (Yes or No). The NSQIP database only captures 30-day readmissions and reoperations that are related to the index surgery; all other readmissions and reoperations unrelated to the index surgery assessed are excluded from outcome analysis.

We controlled our outcome analysis for potential confounding covariates, including age at surgery, gender, current tobacco use, and preoperative diagnosis of diabetes. Summary statistics describing patient characteristics and demographics were reported using mean, standard deviation (SD), and percentage. To compare baseline characteristics of non-Hispanic white and non-white subjects, t-test for continuous variables and chi-square test for categorical variables were employed. Penalized logistic regression to consider small numbers of the postoperative complications was used to identify factors associated with incidence of the complications, and results were reported as odds ratios. A p-value less than 0.05 was considered statistically significant.

3. Results

A total of 1164 patients with known race undergoing TAA from 2012–2018 were included in this analysis. Of these patients 1051 (90.3%) were non-Hispanic White (White), and 113 (9.7%) were non-Hispanic Black, Asian, Native American, and Hispanic (non-White). There was a statistically significant difference (p < 0.01) in patient age, with a mean age of 63.7 (SD: 10.7) years in White patients, compared to 55.3 (13.2) years in Non-White patients. There was a similar distribution of males and females in each group and similar rates of diabetes mellitus in both groups. There was a higher rate of current tobacco use in the non-White cohort compared to the White cohort (14.2% vs. 8.8%, p-value = 0.06). A complete summary of all demographic information between the two cohorts is reported in

Table 1. Patient Characteristics in the White and Non-White Patient Cohorts.

	White (N, %) 1	Non-White (N, %)	p-Value
N	1051 (90.3%)	113 (9.7%)
Age (years)	63.7 (±10.7)	55.3 (±13.2)	<0.01
Gender
Male	562 (53.5%)	55 (48.7%)	0.33
Female	489 (46.5%)	58 (51.3%)
Current Smoker:
Yes	92 (8.8%)	16 (14.2%)	0.06
No	959 (91.2%)	97 (85.8%)
Diabetes Mellitus
Yes	125 (11.9%)	15 (13.3%)	0.67
No	926 (88.1%)	98 (86.7%)

¹ Note: White = non-white Hispanic, Non-White = non-Hispanic Black, Asian, Native American, and Hispanic.

.

Regarding postoperative complications, there were 5 incidences of superficial wound infection in the White cohort (0.5%) and 0 incidences of superficial wound infection in the non-White cohort (0.0%) at 30 days postoperatively. There were no incidences of deep wound infection in either group in the early postoperative period. There were 2 incidences of wound dehiscence in the White cohort (0.2) compared to 0 incidences in the non-White cohort (0.0%). Two patients (0.2%) in the White cohort experienced a pulmonary embolism (PE), compared 0 patients in the non-White cohort (0.0%). There was 1 incidence of postoperative deep vein thrombosis in both the White (0.1%) and non-White (0.9%) cohort. Postoperative blood transfusion was necessary for 3 patients in the White cohort (0.3%) and 1 patient in the non-White cohort (0.9%). Unplanned reoperation within 30 days occurred in 8 patients in the White cohort (0.8%) and 1 patient in the non-White cohort (0.9%). There were 12 incidences of readmission within 30 days in the White cohort (1.2%), compared to 0 incidences in the non-White cohort (0.0%) (

Table 2. 30-Day Postoperative Complications.

Complications	White N, %	Non-White N, %	Odds Ratio (OR)	p-Value	95% Confidence Interval
Superficial Wound Infection	5 (0.5%)	0 (0.0%)	0.99	1.00	0.05–19.43
Deep Wound Infection	0 (0.0%)	0 (0.0%)	0.20	0.39	0.00–8.15
Wound Dehiscence	2 (0.2%)	0 (0.0%)	0.31	0.46	0.01–6.94
Pulmonary Embolism (PE)	2 (0.2%)	0 (0.0%)	0.80	0.88	0.04–16.45
Deep Vein Thrombosis (DVT)	1 (0.1%)	1 (0.9%)	0.13	0.10	0.01–1.47
Blood Transfusion	3 (0.3%)	1 (0.9%)	0.24	0.18	0.03–1.93
Unplanned reoperation	8/977 (0.3%)	1/108 (0.9%)	0.76	0.76	0.13–4.50
Readmission	12/973 (1.2%)	0/109 (0.0%)	2.90	0.94	0.17–50.24

).

In comparison of relative risk associated with postoperative complications, there were no significant differences in the odds ratio for any complication assessed between the two cohorts. Furthermore, there were no increased risks associated with developing superficial wound infection, deep wound infection, PE, DVT, requiring a blood transfusion, requiring another operation, or requiring readmission to the hospital within 30 days in the non-White cohort. A full summary of complication rates and relative risks are summarized in Table 2.

4. Discussion

The primary aim of this study was to identify any differences in utilization of TAA across a patient population using a large, nationwide database. The secondary aim was to assess for any differences in incidence of adverse outcomes, stratified by racial or ethnic group. Our study found a striking discrepancy in utilization rates across racial and ethnic backgrounds, with a 10-fold increase in utilization of TAA in non-Hispanic, White patients as compared to all other racial and ethnic groups. In contrast, however, we found no significant difference in the rates of complications between the patient groups, with the White and non-White cohorts demonstrating similar incidences of superficial infection, deep infection, wound dehiscence, postoperative bleeding requiring transfusion, DVT, PE, reoperation, and readmission within 30 days of total ankle replacement.

The striking differences in utilization of TAA across racial and ethnic groups found in this study contrast with the national demographic breakdown across the United States. In a demographic breakdown of patients with ankle OA, 89% of the population are non-Hispanic white patients, while non-white populations make up the following 11% [23]. Furthermore, there was significant disparity found between the two study groups, reporting a utilization rate of 0.14 in White populations versus 0.07 in Black populations in 1998 that further worsened to 1.17 versus 0.33 in White versus Black populations by 2011 [23]. However, 90.3% of patients who underwent TAA in our analysis were white, compared to 9.7% of non-white patients, demonstrating the discrepancy of TAA utilization by racial group for patients presenting with a diagnosis of ankle OA. Our findings of differences in utilization rate compared with national patient demographic distribution are consistent with the results from previous studies assessing racial and ethnic disparities in orthopedics. In one retrospective review of racial disparities in TAA using the National Inpatient Sample, there was 2-fold higher utilization rate of TAA found in the White cohort compared to the Black cohort in 1998, which continued to worsen to a 4-fold difference between the two groups by 2011 [24]. Our study further supports these previous findings, and highlights that racial disparities for utilization of TAA have continued to persist within the modern era. Unfortunately, these differences in utilization rates stratified by race permeate throughout all subsets of TJA, and Orthopedics in general. For example, in total shoulder arthroplasty (TSA), previous studies examining this issue in racial disparities have reported that the proportion of black patients undergoing TSA account for only 3–5% of all cases of shoulder arthroplasty [25,26]. In total knee arthroplasty (TKA), lower utilization rates among non-white populations have similarly been demonstrated, with one study reporting a rate of 4.65 white patients undergoing TKA per 1000 annually, compared to a rate of only 3.90 per 1000 annually for non-white patients [27]. Similarly, further reports have found that white patients account for 92% of all cases of primary total hip and knee arthroplasty, while non-white patients only account for 8% of all cases [28]. Our findings only further validate this trend reported within TJA, demonstrating the existence of racial disparity in the utilization of total joint arthroplasties among non-white patient populations. Action must be taken by health care providers and policy makers to combat these disparities and attempt to mitigate the racial gap in utilization rates.

Racial disparities regarding healthcare access and appropriate care have been a topic of concern within the field of orthopedics and within TJA specifically. However, despite numerous studies documenting the racial gap in utilization within TJA over the years, these disparities only continue to persist [12,28,29,30,31,32,33,34,35,36,37]. Though the underlying causes of these trends are complex and multifactorial, several factors have been theorized to contribute to these discrepancies in utilization and outcomes following TJA by racial and/or ethnic group.

One of these factors that has been investigated is socioeconomic status, including insurance coverage, access to specialty care, and access to high-volume hospitals. While the prevalence of surgeons that preform ankle arthrodesis is relatively well distributed throughout the United States, surgeons who preform TAA are disproportionally concentrated within high-volume centers located in large, metropolitan areas [38]. As a result, poor, minority patients may encounter barriers to TAA due to travel distance, lack of transportation, insufficient insurance coverage, or difficulty obtaining surgeon referral. Minority patients, as a result, may be subject to seek treatment at a low-volume hospital, where an intervention such as ankle arthrodesis, instead of TAA, may be more reliably performed and recommended [39]. However, these differences in utilizations rates by racial group may only partially be attributed to differences in socioeconomic status. In fact, in a previous investigation of racial disparities in TKA, authors found that when controlling for income, utilization rates for TKA remained significantly lower among minority populations in comparison to White populations [32].

Other factors that may contribute to the racial disparities in TJA utilization rates are insufficient communication and medical mistrust between the patient and the provider. The patient–provider relationship is imperative to establish trust and ensure open, free communication between the two parties. However, lack of diverse representation or differences in cultural, ethnic, or socioeconomic backgrounds may lead to patient distrust and physician bias that ultimately harms the patient–provider relationship [40]. In a report by the Institute of Medicine, it was determined that providers were less likely to offer alternative therapies or treatments in patient who seemed distrustful or unwilling to undergo treatment [41]. Moreover, in another study assessing patients seeking treatment for osteoarthritis in the Veterans Affairs administration, it was found that African American patients were significantly less likely to receive a recommendation for surgery compared to White patients, even when controlled for patient age, despite having similar rates of osteoarthritis [29]. It is possible that bias or lack of empathy may cause physicians to underestimate patient’s osteoarthritis severity, level of pain, or impact on function, and thus impact their recommendations for surgery [42].

Moreover, differences in patient education regarding treatment options, alternative therapies, and surgical outcomes between racial and ethnic groups, may also affect the individual’s choice for therapy and intervention for their condition. Cultural differences in patient perception of diseases and the role of self-healing can skew the individual’s desire to pursue for intensive interventions. Compared to White patients living with osteoarthritis, Black patients have been demonstrated to assign greater credence in the power of prayer and use of over-the-counter medications for treatment of their condition [43]. Furthermore, lack of advocacy and representation within the medical field can disproportionally affect minority groups, resulting in insufficient or incorrect understanding of medical interventions available [35]. Fear and anxiety regarding pain, operative complications, and postoperative outcomes can significantly influence individuals’ decision to pursue surgery; this factor may likely contribute to overall racial disparities of utilization in TJA [44]. In the context of joint arthroplasty, preoperative fear and anxiety has been shown vary significantly between White and Black patients, with Black populations demonstrating significantly higher scores for fear and anxiety during their preoperative consultations [45]. Further research is necessary to fully identify the factors that contribute to the racial disparities found in TAA specifically, as well as in TJA and within the healthcare system.

Our study found no statistically significant difference in the incidence rates of superficial infection, deep infection, wound dehiscence, postoperative bleeding requiring transfusion, DVT, PE, reoperation, and readmission at 30 days postoperatively between our White and non-White patient cohorts. These similar rates of complications following TAA between our two cohorts, however, contrasts with previous findings reported in the literature. In a retrospective review of postoperative outcomes following TAA in White versus Black patients, it was determined that Black patients had higher rates of prolonged hospital stays, in-hospital blood transfusions, and overall rates of postoperative revisions [36]. However, this previous study only assessed patients who underwent TAA in the inpatient setting, thus possibly skewing results by representing a patient cohort with more medical comorbidities, increased risks of complications, and greater demands for postoperative care. Our analysis, by contrast, included both inpatient and outpatient TAAs, which may explain the similar rates of complications. Moreover, several studies investigating racial disparities in postoperative outcomes following TJA have similarly reported higher rates of mortality, readmissions, and complications at 30 and 90 days postoperatively among African American, Hispanic, and mixed-race patients when compared to White patients [27,46,47,48]. However, our findings did not support these trends reported previously in the literature, suggesting that racial disparities may not significantly affect rates of complications in the early postoperative period.

We believe the results obtained from this study are important for several reasons. First, this study uses patient information and outcomes from a large national database that includes data from numerous hospital systems throughout the country, which strengthens our findings generalizability. Our study serves as a sufficient representation of the patient population undergoing TAA in the United States, and therefore our findings demonstrate the prevalence of racial disparities that exist regarding utilization rates for TAA. In addition, at 30 days postoperatively, we found that there were no differences in incidence of complications following TAA based on racial or ethnic identity, demonstrating that postoperative outcomes are not influenced by racial disparity in the early postoperative period. As a result, we can conclude that, although racial disparities exist in TAA in regards to utilization rates, non-white patients appear to achieve similar postoperative outcomes as white patients. Thus, greater efforts are needed to increase the proportion of minority populations that undergo TAA to better improve racial disparities in the setting of TAA.

There are several limitations of this study. First, this study is limited by the brief follow up period of 30 days that is inherent to the data provided in the NSQIP database. The 30-day postoperative period utilized in the NSQIP database serves as a metric for healthcare quality as most complications of surgery occur in the first 30 days following the index operation [49]. However, this period may not accurately capture orthopedic complications that typically occur in the long term, such as implant failure, osteolysis, or aseptic loosening. While important and clinically relevant information can be gleaned in this short time frame, longer follow up periods of 2 years or greater would provide more in depth and applicable clinical information. This study is also limited by its retrospective nature. Furthermore, our findings are reliant upon patient self-reporting of racial and ethnic identities, as well as reliant upon accurate coding and reporting of preoperative and postoperative variable. However, this study provides a broad analysis of large, nationally represented patient populations, which allows us to draw conclusions regarding the current state of racial disparities in the context of TAA.

5. Conclusions

This study aimed to investigate the differences in utilization rates and postoperative outcomes in TAA between White and non-White patient populations. Using a large, national database, our study found a near 10-fold difference between the volume of White patients undergoing TAA compared to the volume of TAA in all other racial/ethnic groups. These results highlight the persistence of racial disparities that exist in the setting of TAA, as well as in the context of TJA as a whole. Additionally, we found no significant differences in short term postoperative outcomes based on racial or ethnic group. This suggests that racial disparities are limited in regard to early postoperative complications in the setting of TAA. Greater efforts are needed to ensure proportional representation of minority groups in TAA, which may necessitate changes in preoperative patient education, increased representation of minority groups in the field of TAA, and improved patient–provider communication and understanding.