Decrease in Pediatric Anterior Cruciate Ligament Injury-related Care Disparities for Medicaid Patients From 2015 to 2020

Abstract

Introduction:

Anterior cruciate ligament (ACL) injuries are relatively common in the pediatric population, and ACL reconstruction (ACLR) followed by postoperative physical therapy (PT) is frequently considered to decrease the risk of additional knee injury and improve functional outcomes. Disparities across insurance groups in ACL surgical rates and PT utilization have been previously reported, but there is limited analysis from national databases, and even less about disparity trends over time.

Methods:

The 2015 to 2020 M157 PearlDiver database was queried for patients younger than 18 years diagnosed with ACL injury with commercial or Medicaid insurance. Factors abstracted included year, insurance plan, ACLR or not, and PT utilization and number of visits. The incidence of ACLR and PT visits were compared across insurance groups by year.

Results:

In total, 61,333 ACL injuries were identified for which ACLR was done for 21,083 (34.4%). The incidence of reconstruction was higher for those with commercial than Medicaid (35.0% vs. 30.7%, P < 0.001). Over the years of the study, this gap decreased from 7.7% in 2015 (P < 0.001) to 0.7% in 2020 (P = 0.714). For postoperative patients, those with commercial insurance had more PT sessions (26.45 vs. 22.53, P < 0.001). This gap decreased from 6.48 sessions in 2015 (P < 0.001) to 4.07 sessions in 2020 (P = 0.002). For nonsurgical patients, those with commercial insurance were more likely to receive PT (43.3% vs. 39.8%, P < 0.001) and to receive more sessions (18.76 vs. 14.35, P < 0.001). The gap in PT incidence deceased from 9.7% in 2015 (P < 0.001) to 2.4% in 2020 (P = 0.200).

Conclusion:

Medicaid pediatric patients with ACL injuries were markedly less likely to undergo reconstruction and received less PT than commercial insurance patients, but these differences decreased/disappeared over the years studied. It is reassuring to see these insurance-related disparities decreasing over time, and continued access efforts are needed.

Anterior cruciate ligament (ACL) injuries are relatively common in the pediatric population and have increased in incidence due to greater participation and intensity of youth sports.^1,2 For this patient population, early surgical ACL reconstruction (ACLR) has been established as the standard of care, improving the rate of return to athletic activity and reducing the risk of secondary meniscal and chondral injuries.^3-5 Furthermore, physical therapy (PT) has been shown to have a role in reducing the risk of reinjury and improving outcomes after ACL injury and reconstruction.^6,7

Nonetheless, disparities in the management of pediatric ACL injuries have previously been reported. One 2014 epidemiologic study of ACLRs in New York State found that reconstructions were six times more common in patients with private health insurance compared with those enrolled in Medicaid.⁸ Other studies have identified higher cost of surgery, lower socioeconomic status, lack of private insurance, and identifying as a racial minority as barriers to receiving ACLR.^9,10

Numerous disparities in PT utilization between demographic groups have been identified for a variety of conditions. One study in 2022 identified disparities by race in PT utilization for musculoskeletal pain.¹¹ Another study investigating PT after rotator cuff repair found that markedly fewer practices accepted Medicaid insurance nationally compared with private insurance.¹² In ACL injury care, disparities across racial and insurance groups were also found in the number of PT visits and functional outcomes after surgery.^1,2

Medicaid is the largest health insurance program by enrollment in the United States and disproportionately enrolls people from vulnerable, minority backgrounds.¹³ Patients with this coverage are only expected to have increased with the expansion of Medicaid following the Affordable Care Act.¹⁴ Therefore, ensuring patients with Medicaid insurance receive equitable care compared with other insurance programs may be important in addressing disparities faced by underserved populations.

This study aimed to characterize trends in pediatric ACL surgical rates and PT utilization for those with Medicaid relative to commercial insurance, providing insight into how previously described disparities have evolved over time. Understanding if any progress has been made in reducing healthcare disparities may be helpful for informing future initiatives.

Methods

Study Population

A retrospective study was done with data from between 2015 and 2020 from the M157Ortho PearlDiver Mariner database (PearlDiver Technologies), a commercially available administrative US database containing 157 million patients. As data from this database are output in deidentified/aggregated form, our institutional review board deemed studies using this database as exempt from review. The PearlDiver database is effective in analyzing differences between insurance groups and for evaluating trends over time within the field of orthopaedics.^3-7

International Classification of Diseases of the 10th Revision (ICD-10) codes were used to isolate patients with ACL injuries⁴ (Table 1). Exclusion criteria included age 18 years or older, recent history of fracture/polytrauma, infection, or neoplasm, and not being active in the database for 90 days after their diagnosis/procedure. Patient factors abstracted included age, sex, Elixhauser Comorbidity Index, and insurance plan. Year of injury was also identified.

Table 1.

Current Procedural Terminology/International Classification of Diseases Codes Queried

Description	CPT/ICD-10 Codes
ACL injury	ICD-10-D-M23601, ICD-10-D-M23602, ICD-10-D-M23611, ICD-10-D-M23612, ICD-10-D-M2351, ICD-10-D-S83512A, ICD-10-D-S83512D, ICD-10-D-S83512S, ICD-10-D-S83511A, ICD-10-D-S83511D, ICD-10-D-S83511S, ICD-10-D-M2352
ACL reconstruction	CPT-29888
Physical therapy	CPT-97001, CPT-97005, CPT-97006, CPT-97010, CPT-97014, CPT-97016, CPT-97022, CPT-97032, CPT-97035, CPT-97110, CPT-97112, CPT-97113, CPT-97116, CPT-97124, CPT-97140, CPT-97105, CPT-97161, CPT-97162, CPT-97163, CPT-97164, CPT-97530, CPT-97750

ACL = anterior cruciate ligament, CPT = Current Procedural Terminology, ICD-10 = International Classification of Diseases of the 10th Revision

ACL Reconstruction and Physical Therapy Utilization

Patients receiving ACLR were abstracted using Current Procedural Terminology code CPT-29888, and year was identified. Patients with an ACL injury diagnosis who did not receive an ACLR within 2 years were defined as having had nonsurgical treatment.

Patients receiving PT after injury and/or reconstruction were the defined based on Current Procedural Terminology codes (Table 1). Physical therapy defined as yes or no, and the number of sessions within the first year after ACL injury (in the nonsurgical group) or reconstruction (in the surgical group).

Data Analysis

For the Medicaid and commercially insured patients, the number of ACL injuries and reconstructions was tabulated by year. The number of reconstructions relative to injuries was compared.

Chi-square tests for equality of proportion were done to test for statistical significance in the difference of the number of ACLR to ACL injuries between insurance groups, as well as the difference in incidence of PT between insurance groups for both nonsurgical and postoperative ACL injury patients. Two sample t-tests were used to evaluate the significance of differences observed in the average number of PT sessions received between insurance groups for both nonsurgical and postoperative ACL injury patients.

All statistical analyses were done using the PearlDiver software, R, or GraphPad Prism version 9 (GraphPad Software). Significance was defined as P < 0.05.

Results

Study Population and Surgical Rates

For the Medicaid population, a total of 8834 ACL injuries and 2712 ACLRs were identified between 2015 and 2020 (Table 2). The ratio of reconstructions to injuries was 30.7%.

Table 2.

Anterior Cruciate Ligament Injuries and Anterior Cruciate Ligament Reconstructions Between 2015 and 2020 by Insurance Type

Year	Commercial ACL Injuries	Commercial ACL Reconstructions (% of Injuries)	Medicaid ACL Injuries	Medicaid ACL Reconstructions (% of Injuries)	P
2015	3,818	1,349 (35.3%)	601	166 (27.6%)	<0.001
2016	11,414	3,787 (33.3%)	1,942	536 (27.6%)	<0.001
2017	10,280	3,650 (35.5%)	1,754	541 (30.8%)	<0.001
2018	9,823	3,545 (36.1%)	1,520	459 (30.2%)	<0.001
2019	10,267	3,570 (34.8%)	1,867	607 (32.5%)	0.062
2020	6897	2,459 (35.7%)	1,150	403 (35.0%)	0.714
Total	52,499	18,371 (35.0%)	8,834	2,712 (30.7%)	<0.001

ACL = anterior cruciate ligament

For the commercial insurance population, a total of 52,499 ACL injuries and 18,371 ACLRs were identified between 2015 and 2020 (Table 2). The number of reconstructions to injuries was 35.0%, significantly higher than the number found in the Medicaid population (P < 0.001).

When assessed by year, the gap in ACLRs relative to ACL injuries between the insurance types decreased from 7.7% in 2015 (significantly different, P < 0.001) to 0.7% in 2020 (not significantly different, P = 0.714) (Figure 1).

Figure 1.

Diagram showing the number of anterior cruciate ligament (ACL) reconstructions divided by the number of ACL injuries diagnosed over time by insurance. The gap between Medicaid and commercial insurance steadily decreased and was no longer statistically significant by 2020.

Physical Therapy After ACLR

After ACLR, 61.1% of patients with Medicaid insurance received PT, while 62.4% of commercially insured patients received PT (P = 0.199). For those with Medicaid insurance who received PT, the mean (±SD) number of sessions of PT was 22.53 ± 16.77 sessions, significantly fewer than the mean for patients with commercial insurance, which was 26.45 ± 18.72 sessions (P < 0.001).

When assessed by year, in 2015, the average number of PT sessions after ACLR for commercial insurance was 25.64, which was 6.48 sessions more than the 19.16 sessions for Medicaid patients (P < 0.001). In 2020, the average number of PT sessions after ACLR for commercial insurance patients was 30.24, which was 4.07 sessions more than the 26.17 sessions for Medicaid patients (P = 0.002). No significant improvement was observed in the proportion of patients receiving PT between insurance groups between 2015 and 2020 (Table 3).

Table 3.

Trends in Physical Therapy Incidence and Average Utilization Following Anterior Cruciate Ligament Reconstruction Between Insurance Groups

Factor or Variable	Physical Therapy Visits Post ACL Reconstruction in Pediatric Patients
	Commercial	Medicaid	P
Total	18,362	2,712
Patients receiving PT (%)	11,458 (62.4%)	1,657 (61.1%)	0.199
PT sessions (mean ± SD)	26.45 ± 18.72	22.53 ± 16.77	<0.001
2015	1,340	166
Patients receiving PT (%)	730 (54.5%)	86 (51.8%)	0.570
PT sessions (mean ± SD)	25.64 ± 19.12	19.16 ± 12.30	<0.001
2020	2,459	403
Patients receiving PT (%)	1551 (63.1%)	270 (67.0%)	0.144
PT sessions (mean ± SD)	30.24 ± 20.50	26.17 ± 19.06	0.002

Bold text indicates P < 0.5, ACL = anterior cruciate ligament, PT = physical therapy

Physical Therapy After Nonsurgical ACL Injury Treatment

After ACL injury without reconstruction (managed nonsurgically), 39.8% of patients with Medicaid insurance received PT, significantly less than the number of commercially insured patients receiving PT, which was 43.3% (P < 0.001). Those with Medicaid insurance received on average 14.35 ± 14.50 sessions of PT, which was significantly less than that received by those with commercial insurance which was 18.76 ± 17.14 sessions (P < 0.001).

When assessed by year, 39.6% of commercial insurance patients received PT in 2015 compared with 29.9% of Medicaid patients, which corresponded to a gap of 9.7% (P < 0.001). In 2020, 39.7% of commercial insurance patients received PT compared with 37.3% of Medicaid patients, which corresponded to a gap of 2.4% (P = 0.200). No improvement was observed in the average number of sessions of PT received between 2015 and 2020 (Table 4).

Table 4.

Trends in Physical Therapy Incidence and Average Utilization After Anterior Cruciate Ligament Injury Treated Nonsurgically Between Insurance Groups

Factor or Variable	Physical Therapy Visits Post Nonsurgical ACL Injury in Pediatric Patients
	Commercial	Medicaid	P
Total	37,025	6,437
Patients receiving PT (%)	16,030 (43.3%)	2560 (39.8%)	<0.001
PT sessions (mean ± SD)	18.76 ± 17.14	14.35 ± 14.50	<0.001
2015	2,775	452
Patients receiving PT (%)	1099 (39.6%)	135 (29.9%)	<0.001
PT sessions (mean ± SD)	16.97 ± 15.61	13.79 ± 12.14	0.006
2020	4,910	826
Patients receiving PT (%)	1950 (39.7%)	308 (37.3%)	0.200
PT sessions (mean ± SD)	21.50 ± 19.92	16.23 ± 16.73	<0.001

Bold text indicates P < 0.05, ACL = anterior cruciate ligament, PT = physical therapy

Discussion

Injuries to the ACL are common in pediatric patients, and surgical reconstruction and PT have been shown to have important roles in improving functional outcomes. Understanding potential disparities in these two aspects of care is relevant in improving equity for pediatric patients facing ACL injuries. This study used a large, national, multi-insurance administrative database to address knowledge gaps in these areas.

A total of 61,333 ACL injuries meeting the study criteria were identified. A much larger number of ACL injuries were identified in the commercial insurance group compared with the Medicaid group (52,499 vs. 8,834). This difference may be due to several reasons. First, a much larger proportion of children in the United States are covered using commercial plans compared with public insurance.⁸ In addition, the PearlDiver database consists of a greater number of records and patients with commercial insurance compared with Medicaid. Finally, previous studies using other databases investigating insurance and ACL injuries found markedly higher number of commercial ACL injuries compared with Medicaid, providing additional evidence that commercial insurance patients are more likely to have/report an ACL injury.⁹

In assessing the rate of ACLR relative to ACL injury, those with Medicaid had a lower rate than those with commercial insurance (35.0% vs. 30.7%). This recapitulated findings in the literature that had been based on local/institutional data.^8,15,16 However, this study went on to find that this difference decreased/disappeared over the years of the study. The difference changed from 7.7% in 2015 (markedly different for Medicaid relative to commercial insurance) to 0.7% in 2020 (no longer markedly different). Although the clinical significance of this initial gap may not be certain, this relative difference of 20% may have contributed to poorer outcomes, and it is encouraging that the earlier-described gap in access to care has resolved to be equivalent between insurance groups.

The rate of ACLR for ACL injuries (between 28% and 40%) was lower than expected and reflects the lack of specificity in ICD-10 coding for ACL injuries. The set of ICD-10 codes used to identify ACL injuries is not uniform across the literature.^{10–12,15,16} The set of codes used to identify ACL injuries was ultimately selected from a study that studied the clinical effect of inadequate access to ACLR, but it is possible that this set of codes would include ACL injuries for which a reconstruction may not necessarily be indicated.¹⁵ Although the ACL injury cohort may include such patients, it was assumed that the distribution of such injuries would be similar across both insurance cohorts.

After ACLR, more than 60% of patients received PT. No notable difference was observed in the proportion of patients receiving PT between insurance groups. However, patients with Medicaid insurance received fewer sessions of PT on average than those who were commercially insured. Although this difference remained statistically significant in 2020, the gap seemed to be shrinking—again an encouraging finding.

For nonsurgical ACL injury treatment, the rate of PT utilization was surprisingly low in both insurance groups compared with postoperative ACLR PT, with less than 50% utilization. It was expected that PT in a nonsurgical cohort would be the mainstay of treatment and that the rate of PT utilization would be like that of patients receiving ACLR. With this analysis, patients with Medicaid insurance were markedly less likely to receive PT than their commercial insurance counterparts. However, this difference vanished across the years of the study. As with patients who had undergone ACLR, those with Medicaid insurance managed nonsurgically received on average fewer sessions of PT than those who were commercially injured. No improvement was observed in this disparity across the years of the study. Although the reasons for disparity in PT access for ACL injuries may not be explicitly known, previous studies of musculoskeletal conditions have identified Medicaid insurance and low socioeconomic status as causes of disparity in PT utilization, with one study after rotator cuff repair finding that a notable number of PT practices do not accept Medicaid insurance.^13,14

Although it is currently difficult to determine the exact reason behind these decreased disparities over time, there are several potential explanations. Possibly, increased attention to such gaps in care from previous studies may, unto itself, have led to increased efforts to address previously recognized disparities.^9,10,16 In addition, the possibility exists that Medicaid expansion through the Affordable Care Act launched in 2014 may have increased the numbers with this coverage and secondarily increased access to care/reduced disparities, as has been shown for rehabilitation in other areas.¹⁷ Nonetheless, despite the observed improvements in disparities, several disparities were found to remain in 2020 that need ongoing attention.

This study has several strengths. First, use of the PearlDiver database includes a diverse, nationally representative patient population from numerous geographical locations. The large size and diversity of the database support the generalizability of the results. In addition, the large number of patients allows for tracking of trends of disparity across time.

There are also several limitations present in this study. Most notably, the study has the limitations of a retrospective study and reliance in administrative coding. As previously mentioned, the lack of specificity surrounding ICD-10 coding for ACL injury resulted in ACLR rates that were lower than expected. Furthermore, the actual drivers of patient-specific decisions, such as the preferences of the patient, family, and/or surgeon could not be assessed. Finally, outcome differences between those with or without surgery, with or without PT, and with variable number of sessions of PT could not be determined.

Conclusion

Overall, Medicaid pediatric patients with ACL injuries were markedly less likely to undergo reconstruction and received less PT than commercial insurance patients, but these differences decreased/disappeared over the years studied. It is reassuring to see these insurance-related disparities decreasing over time, and continued access efforts are needed.