Associations Between Race and Socioeconomic Status, Lower Extremity Strength, and Patient-Reported Outcomes Following Anterior Cruciate Ligament Reconstruction

This was a single-site study including patients who had undergone an ACLR from a single hospital setting. Inclusion criteria included the ACL injury being the first ACL injury ever sustained by a patient, no other structural damage to the knee (ie, no other ligament damage), and an uncomplicated ACLR procedure, meaning no follow-up procedures such as revisions, debridements, or chondroplasties were performed. The patient’s race was obtained from electronic medical record (EMR) data. Race was categorized following the same classifications as used by the EMR: American Indian or Alaska Native, Asian, Black or African American, Hispanic or Latino, White, and Other or Unknown. Patients were only able to select 1 option for their race within the EMR. Each patient’s address from the EMR was used to determine ADI values. This study was approved by the local institutional review board, and all patients involved in this study provided informed consent.

Following ACLR at our institution, patients undergo standardized return to physical activity testing to monitor physical recovery and help establish a safe clearance for return to sport. Testing was completed based on referral by the treating surgeon and was typically completed for most patients initially at 3 to 4 months post-ACLR, with a follow-up assessment 5 to 8 months post-ACLR. The number of assessments varied between patients based on their willingness to complete testing as well as their rehabilitation progress, as patients who performed well were often not required to return. As part of their assessment, patients perform knee flexion and knee extension isokinetic strength testing as well as hop tests, including both single hops and triple hops. The strength and hopping metrics are then compared between the reconstructed and contralateral legs to establish a Limb Symmetry Index (LSI) between the 2 limbs. The LSI metric is calculated as (reconstructed limb value/contralateral limb value) × 100%. The LSI is expressed as a percentage, with 90% accepted as sufficient symmetry during strength and hopping tasks.²⁰ 

Each patient completed concentric knee extension and flexion strength testing at 90°/s on a Biodex System 3 Dynamometer, a machine that measures isokinetic strength. Patients completed 8 repetitions of knee extension and 8 repetitions of knee flexion, and the peak knee extension and peak knee flexion torques were used for analysis, respectively. All testing was performed first on the uninvolved limb and then the involved limb. Participants completed practice trials on each limb to ensure familiarization before testing. We chose to analyze the peak instead of the average across trials because our primary interest is in evaluating maximal torque generating capacity of each muscle group. This approach has been widely used in the ACLR assessment literature.^21–23  Testing was always conducted on the uninvolved limb first followed by the involved limb.

Patients also completed the single-leg hop for distance.²⁴  Each patient completed 3 single-leg hops per limb along a 6-m tape measure, alternating between the uninvolved and involved limb. Hop distance was measured from the beginning of the tape measure to the landing position of the heel, and the average distance of the 3 trials was used for analysis. If patients could not hold the landing position for at least 2 seconds, the test was repeated.

All patients completed a battery of PROs during their assessment visit. The International Knee Documentation Committee (IKDC) was used to determine subjective knee function.²⁵  The IKDC consists of 3 categories: symptoms, athletic activity, and knee function.²⁵  The Knee Injury and Osteoarthritis Outcome Score (KOOS) was used to assess the severity of each patient’s knee symptoms and functional disability.²⁶  The KOOS questionnaire consists of 5 measures: Pain, Symptoms, Activities of Daily Living, Sport and Recreation Function, and Knee-Related Quality of Life.²⁶  Each of these outcome measures is scored on a scale of 0 to 100, with 100 representing ideal status. Previous studies have found that both of these measures are correlated with post-ACLR outcomes. Regarding IKDC, prior work has shown that individuals who completed a full return to their pre-ACLR activity levels had higher IKDC scores than individuals who did not make a complete return to pre-ACLR activity levels.²⁷  The KOOS questionnaire has been found to be reliable and valid for evaluating patient progress following ACLR across multiple studies.^26,28 

The ADI is a measure that allows for stratification of geographic areas by socioeconomic disadvantage on both state and national levels. It includes measures for factors including income, education, employment, and housing quality.^18,29  The ADI has been used previously in a variety of settings, including, but not limited to, a predictive study regarding readmission to a large urban hospital and a retrospective review of mortality related to COVID 2019.^30,31  Specifically in orthopedic settings, the ADI has been correlated with worse pain and function scores as well as poorer outcome measures following pediatric orthopedic procedures.¹²  The EMR was also used to determine each patient’s current listed home address, which was then used to find their ADI. Area Deprivation Index values can range from 1 to 99, with higher numbers representing greater disadvantage and lower numbers indicating lesser disadvantage.

All statistical analyses were performed in Stata 17 (StataCorp). Comparisons of categorical data were made via chi-squared analysis. Because continuous data were not normally distributed, sex comparisons were performed using Mann-Whitney U tests, and associations were calculated using Spearman ρ correlations using Stata 17 to determine covariance between ADI and KOOS measures, the IKDC, and limb symmetry on strength measurements.³²  Correlation coefficients were classified as weak (0.0–0.25), fair (0.26–0.50), moderate (0.51–0.75), and strong (≥0.76).³³  The α level was set at P < .05.

A total of 340 participants following ACLR were enrolled in this study, with full demographic information displayed in Table 1. The population in this study was comprised of student-athletes as well as community members, with a large number of the participants being involved in athletics at some level. Four hundred sixty-three patients were initially included. One hundred twenty-three patients were excluded for lack of ADI, race, or KOOS data or for failing to meet the 3 ACLR criteria of being primary (first ever ACL injury), isolated (no other cartilage or ligament damage), and unilateral, leaving 340 patients (Figure 1).

The racial demographics of the sample are shown in Table 1. The median national ADI value of the sample was 32.5 (interquartile range, 19–47), with a minimum value of 1 and a maximum value of 86. Visualization of the distribution of the ADI can be found in Figure 2, and further patient demographics are available in Table 1.

Lower extremity strength, single-leg hop, and PRO measures are summarized in Table 2. The relationships between all measured variables and PROs with the ADI were examined. Only patient outcomes from the IKDC (ρ = 0.11, P = .04) were statistically significantly correlated with the ADI (Figure 3). The ADI was not statistically significantly correlated with knee extension LSI (ρ = –0.09, P = .09), knee flexion LSI (ρ = –0.01, P = .89), single-leg hop LSI (ρ = –0.04, P = .54), KOOS Symptoms (ρ = 0.09, P = .10), KOOS Pain (ρ = 0.06, P = .28), KOOS Activities of Daily Living (ρ = 0.06, P = .28), KOOS Sport (ρ = 0.06, P = .24), and KOOS Quality of Life (ρ = 0.10, P = .06).

Our results suggest that post-ACLR outcomes may be related to the degree of a patient’s socioeconomic disadvantage as measured by the ADI, with more disadvantaged patients reporting better knee-related function and quality of life. As outcomes following ACLR have often been examined on the basis of sex demographic variables, PROs, strength, and hop data were compared by sex. Age, height, mass, and knee extension strength of the quadriceps varied by sex. There was a statistical difference in median age by sex of approximately 2 years, and both sexes showed unsatisfactory strength in quadriceps function using traditional LSI interpretations.³⁴  Therefore, it was determined that it would be appropriate to focus on the role of the ADI across both sexes in this sample.

The primary finding of this analysis was the inverse relationship between the ADI and IKDC score. These findings run contrary to much of the existing literature, which has generally shown poorer outcomes following orthopedic procedures in more disadvantaged patients.^12,13,19  Unfortunately, the mechanisms to explain these findings are unclear, although it warrants further investigation to determine what underlying factors may explain more disadvantaged patients having better PROs or if this finding is potentially due to reporting bias. It is important to note that statistically significant associations were observed between the ADI and only a small portion of PRO measures. There were no other statistically significant associations between the ADI and other outcomes of interest for this study, including other metrics for PRO measures and functional measures. Further, the associations that did exist between the ADI and PROs were classified as weak.

Although the ADI has been shown to be associated with outcomes in other areas of health care, it does not appear to be well correlated with postoperative lower extremity strength measures and PROs following ACLR. Additionally, it is important to highlight that the measure found to be related to the ADI, IKDC, is a subjective, patient-reported measure, whereas all strength metrics used to gauge ACLR were found to have no correlation with the ADI. Given the skewed racial distribution of participants, we were not able to test the role of race as an independent predictor of post-ACLR outcomes or its interaction with the ADI. Relative to national demographics, White patients were overrepresented by roughly 15%, Hispanic patients were underrepresented by roughly 15%, and African American patients were underrepresented by 1.5% in our sample.³⁵ 

The weak correlation between the ADI and post-ACLR outcomes may be impacted by the elective nature of the ACLR procedure. Given the economic and social burdens of undergoing ACLR, individuals of lower SES may be less likely to pursue this treatment option in situations where it is not deemed necessary or if they do not have the financial stability to afford the costs of surgery and sufficient physical therapy and missed time from work. It is also possible that the decreased number of more disadvantaged patients within this sample could reflect an inability for this patient population to see a physician or physical therapist if they do sustain an ACL injury. An additional possibility is that patients with higher ADI values may have lower levels of involvement in sports and other activities that have increased potential for causing ACL injuries. Socioeconomic barriers that can limit sports involvement include cost, household factors such as material deprivation, and neighborhood factors such as lack of access.³⁶ 

Previous studies in this area have yielded mixed evidence regarding the strength of relationships between race, SES, social determinants of health, and postoperative ACLR outcomes.^9,10,37,38  Prior research has suggested that patients of lower SES have poorer post-ACLR outcomes than patients of higher SES.¹³  Although much of the existing literature consistently points to lower SES resulting in poorer outcomes following ACLR, there is less conclusive evidence regarding the correlation between race and post-ACLR outcomes. In 1 study, African American and Hispanic patients were found to have greater deficits in strength during rehabilitation and a greater risk for hospital admission following ACLR than White patients.⁸  However, there are other studies that found that African American, Asian, and Hispanic patients had a lower risk of future ACLR after an initial surgery than White patients.³⁷  Previously published studies have found no differences in post-ACLR outcomes between races. One study found that, despite African American and Hispanic patients averaging fewer physical therapy visits than White patients, there was no difference in postoperative outcomes, namely graft rupture, a contralateral ACL injury, or a new meniscus injury, between races.¹⁰  Ultimately, it appears that more work is needed to fully realize what impact, if any, race has on post-ACLR outcomes.

A potential area for future work is the collection of insurance data, as this may allow for a more accurate assessment of an individual’s ability to access care and rehabilitation resources following ACLR. Insurance data were not included in this patient sample, as a significant portion of the patients did not have verifiable insurance information in their EMR. Prior work regarding insurance status and its association with post-ACLR outcomes is more robust than the ADI and other measures of SES. Patients using government insurance as opposed to private insurance experienced greater delays in time from injury to first appointment, injury to surgery, and injury to return to play.³⁹  As mentioned earlier, both spending sufficient time in physical rehabilitation and regaining sufficient strength have been shown to reduce the likelihood of reinjury following ACLR. However, the rehabilitation resources needed to achieve these benchmarks are not readily available to all patients. When comparing patients using private insurance and public insurance, those who used private insurance received magnetic resonance imaging twice as fast after their injury and underwent ACLR almost 40% faster than publicly insured patients.⁴⁰  Patients using Medicaid insurance had their coverage accepted at almost 50% fewer physical therapy clinics than patients who used private insurance and had to wait 3 days more on average than privately insured patients before their initial physical therapy visit.⁴¹  In addition to attending fewer postoperative physical therapy sessions, patients using Medicaid were less likely to return to sport following ACLR than privately insured patients.⁴²  Another potential area for future study is the inclusion of data from other institutions that serve more diverse patient populations to better determine any impact that race and the ADI may have on post-ACLR outcomes.

One of the primary limitations of this study was the homogeneity of the sample population, with over three-quarters of the patients being White and a median national ADI value of 32.5. These findings are not entirely surprising, as they mirror the demographics of the population surrounding the institution from which the sample population was recruited. Based on the most recent census data, the racial demographics of the metropolitan area surrounding the institution are as follows: 74% White, 10% African American, 4% Asian, 6% Hispanic, and 4% identifying as 2 or more races.⁴³  Another limitation was that the ADI, although correlated with many different health outcomes both in our study and others, may not fully encapsulate an individual’s SES and ability to access care following ACL injuries and ACLR. Additionally, the ADI is not able to capture factors such as a patient’s insurance status or baseline strength and activity level, all of which may impact post-ACLR outcomes either directly or indirectly.⁴⁴  Another limitation of the ADI is that it is based upon the address provided by the patient. At an academic institution like ours that serves many individuals who are attending college, the address listed may be the address of their student housing and thus not an accurate representation of their ADI. In our sample, 150 individuals were between the ages of 18 and 22, presumably many of whom were attending college and not living at their home address. Given that the area surrounding our institution is largely comprised of low ADI values, the average ADI of our sample may be lower than it would have been if students had used their home address.

Ultimately, patient ADI values were found to have statistically significant, although weak, correlations with IKDC measures. These results suggest that patients of greater socioeconomic disadvantage have better post-ACLR outcomes as measured by the IKDC. Other outcomes of interest regarding patient strength and PROs were not found to have a statistically significant relationship with the ADI. Although this is potentially an encouraging finding, it remains true that timely interventions following ACL tears and having sufficient rehabilitation are important factors in maximizing patient outcomes following ACLR, and some patients will be limited in their ability to access these resources. It is important to remain cognizant of the disparities that exist across many areas of health care, including orthopedics, and continue to work toward more equitable access to health care for all individuals.