1 Introduction

Disparities in healthcare utilization and outcomes are known to exist throughout the health care system for minority groups including racial and ethnic minorities. Many factors contribute to ongoing healthcare disparities including social determinants of health, individual behavior and biology, economic and educational disparities, access to healthcare and variable quality of healthcare, racism and discrimination; identifying and addressing these disparities is crucial to achieving health equity and improving healthcare for individuals and overall.¹ According to the Agency for Healthcare Research and Quality's (AHRQ) annual report on disparities, overall there has been improvement in many areas, although there are persistent disparities that need to be addressed.²

Rotator cuff tears are common injuries that lead to pain, loss of function and decreased quality of life in middle aged and older adults.³ Certain populations with this injury can be managed non-operatively with satisfactory improvements in pain, range of motion (ROM) and function, although many patients proceed with and benefit from surgical repair.^4,5 Rotator cuff repair (RCR) has been proven to be a safe and effective surgical procedure with clear benefits for individual patient and society as a whole.^3,6 There are a number of factors that have been shown to adversely affect patient outcomes following RCR including age, diabetes, and size of tear.^7–11 A number of studies in spine surgery and total joint arthroplasty, including total shoulder arthroplasty have shown disparities and poorer outcomes in minority patients undergoing these procedures, although there is little published research into whether this holds true when looking at RCR.^12–16 This study aims to examine the influence of race on short and medium term outcomes of patients undergoing arthroscopic rotator cuff repair by using two large research databases, the American College of Surgeons National Quality Improvement Program (NSQIP) and TriNetX. We hypothesize that, in alignment with other orthopedic specialties, there are disparities and poorer outcomes in minority patients undergoing RCR.

2 Methods

This study was deemed institutional review board exempt by the institutions clinical research committee. The NSQIP and TriNetX research databases were retrospectively queried for rotator cuff repairs that occurred from January 2016 through December 2018. All queries were developed using current procedural terminology (CPT) and international classification of disease 10th edition (ICD-10) codes.

NSQIP was queried for all patients within the date range who underwent arthroscopic rotator cuff repair (CPT code 29827). Patients who had no race or ethnicity documented were excluded. All races other than white were grouped into the minority group. Univariate comparisons between white and minority patients were performed of comorbidities and risk factors that have been previously published or deemed clinically relevant to contributing to complications following this procedure; univariate comparisons between white and minority patients were also performed on 30 day postoperative outcomes. The sample was then propensity score matched (PSM) on every comorbidity and risk factor used in the first analysis. The matched sample was then used to repeat the analysis of 30 day outcomes.

The TriNetX database was then queried using similar methodology including all patients who underwent an arthroscopic rotator cuff repair (CPT code 29827) who did not have a history of previous arthroscopic rotator cuff repair (CPT code 29827). The risk factors captured in the NSQIP database were translated into ICD-10 codes in order to query the TriNetX database. Univariate comparisons of risk factors and outcomes between white and minority were then performed. The 2-year postoperative outcomes evaluated included repeat arthroscopic rotator cuff repair (CPT code 29827), open acute rotator cuff repair (CPT code 23410), open chronic rotator cuff repair (CPT code 23412), reconstruction of complete shoulder rotator cuff avulsion (CPT code 23420), adhesive capsulitis (ICD-10 code M75.0), superficial incisional surgical site infection (ICD-10 code T81.41), deep incisional surgical site infection (ICD-10 code T81.42) and deep joint space surgical infection (ICD-10 code T81.43). This sample was then PSM on comorbidities and risk factors, and a matched analysis of 2-year outcomes was performed.

3 Results

The initial NSQIP sample included 16,467 patients, 12,731 of which were white and 3716 were a minority race. There were significant differences between groups in age, body mass index (BMI), sex, history of diabetes, smoking, chronic obstructive pulmonary disease (COPD), hypertension, and bleeding disorder (p < .001 except for smoking, p = .044) (Table 1). After PSM there were 3716 patients in each group and the only statistically significant risk factor between groups was a higher rate of COPD among white patients (p = .036) (Table 3).

Table 2 and 4 demonstrates the difference in operative characteristics and 30 day outcomes before and after propensity score matching. The only significant difference between white and minority patients was time in the operating room (OR), with minority patients in both the matched and unmatched cohorts having longer OR times. Prior to PSM the mean case duration was 92.3 ± 48.8 min for minority patients and 88.3 ± 44.7 in white patients (p < .001). After PSM the mean case duration was 92.2 ± 48.8 min for minority patients and 87.6 ± 46.1 min for white patients (p < .001).

The initial TriNetX sample included 28,418 patients, 24,231 of which were white and 4187 were a minority race. Prior to PSM there were significant differences between groups in age, BMI, sex, diabetes, nicotine dependence, CHF, hypertension and bleeding disorder (all p < .001). After PSM there were 4185 patients in each group and the patients in the minority group had higher rates of COPD (p = .007), CHF (p = .006) and bleeding disorder (p = .008) (Table 5).

Table 6 shows the difference 2-year outcomes between groups after PSM. There were no significant differences between groups in the rates of repeat open or arthroscopic rotator cuff repair, surgical site infection or adhesive capsulitis.

4 Discussion

In this study, examining comorbidities and short and medium term outcomes for patients undergoing arthroscopic rotator cuff repair, the only significantly different outcome between white and minority patients was operative time with minority patients experiencing significantly longer operative times; this finding remained significant even after propensity score matching. There were significant differences in comorbidity burden prior to propensity score matching in both database samples with minority patients having higher rates of most comorbid conditions that we examined. After the propensity score match was performed, the majority of the significant differences in comorbidities between groups were eliminated. The complication rate at 2 years was low overall, with no significant differences between white and minority patients. These results suggest that while minority patients undergoing RCR have higher comorbidity burden, they experience similar postoperative outcomes as lower-risk white patients.

To our knowledge, this is the first study examining differences in outcomes following RCR by race. The majority of research into racial disparities in orthopedic procedures has focused on larger procedures such as joint arthroplasty and spine surgery. Our finding of longer operative time in minority patients compared with white patients is consistent with prior studies in spine surgery and total hip arthroplasty.^13,18 A recent study by Yin et al. found that in patients undergoing total shoulder arthroplasty, African American patients had significantly longer operative time when compared with white patients, even after propensity score matching.¹⁶ Decreased operative time in rotator cuff repair has been shown to decrease early postoperative pain and may have benefits with regards to fewer anesthetic side effects, thus improving early postoperative recovery.^19,20 The difference in operative time in this study was only 4.5 min which was statistically significant but is of unclear clinical significance. There are likely a number of factors that contribute to the difference in operative time, which cannot be determined from this study; there is some evidence in other specialties that minority patients are more likely to receive care from lower quality and lower volume institutions which may contribute to the differences identified in this study.^21,22

Certain comorbidities including diabetes, hypertension, increased BMI and smoking have been associated with adverse outcomes following orthopedic surgeries, including rotator cuff repair.²³ In our unmatched study populations, minority patients had a higher BMI and significantly higher rates of diabetes and hypertension. While these did not translate into poorer outcomes for minority patients in this study, this could be a cause for concern. It is well documented that patients with diabetes who undergo RCR are at increased risk of a number of different postoperative complications including postoperative infection, adhesive capsulitis and repair failure.^24–26 A recent study by Borton et al. found that patients with diabetes undergoing arthroscopic rotator cuff repair had twice the risk of repair failure and four times the risk of frozen shoulder compared with nondiabetic patients.²⁷ Obesity has also been implicated as an independent risk factor in repair failure and increased postoperative complications in RCR.^28,29 Hypertension has been shown to increase the risk of unanticipated admission following RCR as well as readmission in the first 30 days following surgery.^30,31 Rates of smoking were also significantly different in the unmatched samples; white patients in the NSQIP database had higher rates of smoking and minority patients in the TriNetX database had higher rates of nicotine dependence. A recent study by Naimark et al. found that patients who smoke tend to present with larger tears and worse initial outcome scores and experience lower functional improvement following RCR.³² Based on these studies, and the results of the current study demonstrating increased comorbidity burden in non-white patients, interventions aimed at medically optimizing this high risk population should be considered. Preoperative optimization programs aimed at improving modifiable risk factors such as diabetes control, BMI reduction, and smoking cessation have been demonstrated to effectively improve the value of care delivered in other orthopedic surgery populations.³³ Targeted use of these programs for high risk patients undergoing RCR hold promise to further improve the health status of non-white patients and reduce disparities.

While the current study evaluated disparities in outcomes, multiple prior studies have evaluated disparities in access to care across a variety of patient populations. According to the AHRQ 2019 National Healthcare Quality and Disparities Report, there has either been no change or worsening quality in 5 measures examining cost and access to healthcare when comparing white patients to black patients.² In a recent review by Ziedas et al. examining access to care in patients undergoing anterior cruciate ligament reconstruction, black race was one of the contributing factors to delayed care and inferior outcomes.³⁴ A number of studies have also shown decreased utilization in orthopedic procedures among minority patients, especially in arthroplasty and spine surgery.^14,35–38 In this study, 22.7% of the patients in the NSQIP database and 14.7% of the patients in the TriNetX database were minority, which is less than expected based on the racial breakdown of the United States population. This may indicate that access to and/or utilization of rotator cuff repair is limited among minorities in the United States.

This study does have a number of limitations. First, the inherent biases of a retrospective, observational study limits the generalizability to a wider population. Second, the use of administrative databases that rely on coded data may not provide a representative sample. Both databases used in this study require institutions to opt in to provide data which may bias the samples to larger facilities with an increased focus on quality improvement and research. Future studies using larger national databases that are reflective of the broader population undergoing RCR are required to confirm our findings. Third, there were differences in data definitions between the NSQIP and TriNetX databases which could have led to differences in mapping risk-factors between the two datasets. Finally, the significant finding in this study, longer operative time among minority patients, is likely multifactorial and causal inference cannot be made with a database study. Despite these limitations, this study adds value as one of the first to evaluate disparities in the rotator cuff repair population.

5 Conclusion

Before propensity score matching there were significant differences in comorbidities between white and minority patients in both the NSQIP and TriNetX databases, as well as increased operative time for minority patients in the NSQIP database. After propensity score matching the only significant short-term outcome between white and minority patients undergoing RCR was a difference in operative time; there were no differences in medium-term outcomes. This may indicate that the source of racial disparities is one of access to healthcare rather than an innate difference in the patients themselves. Further study is needed to elucidate this issue.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT author statement

Andrea Johnson: Writing – original draft, Writing – review and editing, Visualization; Abigail Parkison: Writing – original draft; Benjamin Petre: Conceptualization, Supervision; Justin Turcotte: Methodology, Formal Analysis, Investigation, Resources, Data Curation, Writing – review and editing, Supervision, Project administration; Daniel Redziniak: Conceptualization, Writing – review and editing, Supervision.