Disparities in Access but not Outcomes: Medicaid Versus Non-Medicaid Patients in Multidisciplinary Chronic Pain Rehabilitation

Abstract

Purpose:

There are known disparities in chronic pain severity, treatment, and opioid-related risks amongst individuals from lower socioeconomic status, including Medicaid beneficiaries, but little is known about whether Medicaid beneficiaries benefit in a similar way from multidisciplinary chronic pain rehabilitation. This study investigated differences in clinical outcomes between Medicaid and non-Medicaid beneficiaries who completed a 3-week multidisciplinary chronic pain rehabilitation program.

Methods:

Participants (N = 131) completed a broad range of clinical measures pre- and post-treatment including pain severity, pain interference, depression, anxiety, objective physical functioning, and opioid misuse risk. Patients with Medicaid were compared with non-Medicaid patients in terms of baseline characteristics and rate of change, utilizing two-factor repeated measures analyses of variance.

Results:

There were baseline characteristic differences, with Medicaid beneficiaries being more likely to be African American, have higher rates of pain, worse physical functioning, and lower rates of opioid use. Despite baseline differences, both groups demonstrated significantly improved outcomes across all measures (p<.001) and no significant difference in rate of improvement.

Conclusions:

Results suggest that pain rehabilitation is as effective for Medicaid recipients as non-Medicaid recipients. Patients with Medicaid are particularly vulnerable to disparities in treatment, so efforts to expand access to multidisciplinary pain treatments are warranted.

Introduction

Chronic pain is a massive public health problem that affects over 50 million American adults [1] and costs over 635 billion dollars annually in healthcare utilization and lost productivity [2]. Pain is currently the most common reason for health care utilization and contributes to reduced functional status in activities of daily living, social engagement, and work limitations [1]. The burden of chronic pain is carried most heavily by disadvantaged groups in our society, with now well-established disparities in chronic pain outcomes and pain treatment amongst disadvantaged groups, particularly individuals from lower socioeconomic status (SES) and racial minorities [3–5]. These disparities are best understood as the result of complex, interwoven factors, including systemic, structural, interpersonal, and intrapersonal factors [6]. Disparities may be due in part to inaccurate and biased provider assumptions; for example, that lower-SES and Black individuals are less sensitive to pain [7,8].

Pain treatment has been evolving rapidly over the past several years due to advances in the understanding of chronic pain complexity and the opioid crisis. There has been a strong push toward treating pain with more comprehensive, multidisciplinary approaches to improve safety and treatment efficacy [9]. Historically, chronic pain has most commonly been treated pharmacologically, including with long-term opioid therapy (LTOT). However, there is no evidence of long-term efficacy of opioids for chronic pain [10]. and the risks of LTOT are now well-established, including elevated rates of opioid use disorder (OUD) and unintentional overdoses [11–13]. Unfortunately, despite some successes in reducing rates of opioid prescribing, 13 million Americans remain on LTOT, and opioid overdoses have continued rising, including from prescription opioids [14,15].

Given the well-established risks of long-term opioid therapy (LTOT) and the growing evidence supporting nonpharmacological and interdisciplinary chronic pain treatments, guidelines for the management of chronic pain, including the United States National Pain Strategy and the CDC guidelines for pain, recommend reducing the use of opioid therapy and instead using evidence-based, non-opioid, interdisciplinary approaches to the management of chronic pain [16]. Interdisciplinary pain rehabilitation programs (PRPs) are multimodal treatments that typically include a biopsychosocial approach to understanding chronic pain and an interdisciplinary team approach including cognitive behavioral therapy, physical therapy, occupational therapy, and medical management with a focus on non-opioid treatments. Uniquely, pain rehabilitation offers each of these multimodal treatments and is recommended in the Department of Health and Human Service’s Best Practices guidelines for pain management [9]. Multiple meta-analyses have supported the effectiveness of PRPs for improving chronic pain and associated disability [17–19]. Further, there is evidence that PRPs can not only improve chronic pain and related functioning, but also effectively support opioid tapering and reduce opioid-related risks [20].

Medicaid is a government-sponsored medical coverage program for eligible low-income individuals in the United States. The population of Medicaid beneficiaries has been growing rapidly, though it represents a smaller group than those covered through private health insurances in the US, and it remains a disadvantaged groups in terms of health status and access. As of September 2023, there were over 88 million individuals enrolled in Medicaid in the United States with 41 states participating in the Medicaid expansion program, which extends Medicaid coverage to a larger number of individuals with income at or below 138% of the federal poverty level [21]. Medicaid beneficiaries tend to be of a lower socioeconomic status, with the majority reporting annual household incomes of less than $36,000 (the median household income across the United States is $74,580 [22]), and are more likely to be female and younger [23]. Compared to individuals who have private insurance, Medicaid recipients have higher rates of chronic illnesses [23], including various chronic pain conditions and high impact chronic pain [24–27]. Evidence has shown that individuals from lower SES backgrounds are at even higher risk of opioid-related morbidity and mortality; for example, research has found that Medicaid beneficiaries are prescribed opioid medications at twice the rate of non-Medicaid patients [28], experience high-risk opioid prescribing at higher rates [29] and are three to six times as likely to overdose on prescription opioids [29,30]. Medicaid beneficiaries experience disparities in access to health care including longer wait times, fewer options of care providers that accept Medicaid, and lower success rates in scheduling appointments with both primary and specialty providers [31].

Given growing concern about the impacts of pain and opioid-related risks, a growing number of states with Managed Care Organization (MCO) contracts are requiring MCOs, including Medicaid MCOs, to adhere to CDC guidelines for the management of chronic pain [32]. Several state Medicaid agencies have implemented legislation that allows Medicaid recipients to access nonpharmacological pain treatments [33]. However, Medicaid beneficiaries have unique barriers to accessing evidence-based non-pharmacologic treatments for chronic pain. Although this varies geographically and plan-to-plan, limited coverage for non-pharmacologic treatments, visit caps, and poor reimbursement, which disincentivizes providers and programs from serving Medicaid beneficiaries, can contribute to disparate access to non-pharmacologic pain treatment for Medicaid beneficiaries with chronic pain [34–37].

While the impact of social determinants of health typically associated with Medicaid eligibility on access to care has been examined, there has been little examination of comparative outcomes to treatment for chronic pain patients participating in multidisciplinary pain rehabilitation programs based on insurance status. Given the disparate risks from opioid analgesic prescribing and disparate access to non-pharmacologic, non-opioid pain care among Medicaid beneficiaries, we sought to better understand outcomes from pain rehabilitation amongst individuals enrolled in Medicaid. The current study aimed to determine whether there are differences in clinical outcomes after participating in a multidisciplinary PRP between Medicaid and non-Medicaid beneficiaries with chronic pain. Based on extant literature described above, we hypothesized that Medicaid beneficiaries would demonstrate poorer status at baseline in terms of pain and functioning but show similar rates of improvement from interdisciplinary treatment.

Materials and Methods

Sample

Participants included adult patients with chronic pain who completed a 3-week PRP at the Medical University of South Carolina between January 2018 and January 2022. For the purpose of this project, participants were categorized as either Medicaid or non-Medicaid. Medical chart review was used to determine insurance type. Participants were categorized as “Medicaid” if they had Medicaid insurance listed, regardless of plan type or primary/secondary status. Participants were categorized as “non-Medicaid” if there was no Medicaid plan, of any type, listed on their insurance coverages in the medical chart. Of the 131 participants included in this study, 38 were classified as “Medicaid” and 93 were classified as having a non-Medicaid insurance. Participants were excluded if they were uninsured or had a non-Medicaid MCO insurance. There were no self-pay participants. The MUSC PRP has been able to accept patients from a variety of insurances statuses, including Medicaid-enrolled and uninsured patients, in part due to grant support.

Procedures

The PRP is an intensive outpatient program that includes 15 eight-hour days over three weeks and uses a cohort-based model (average 4 patients per cohort). Patients either self-refer to the program or are referred by a healthcare provider. Prior to program initiation, patients complete initial assessments with a health psychologist, nurse, physician, physical therapist, and occupational therapist. A biopsychosocial pain assessment conducted by the psychologist evaluates presence of chronic pain and impact on functioning, comorbid mental health or substance concerns, and appropriateness for the program. Exclusion factors include declining to participate, presence of severe psychiatric disorder that would interfere with participation (e.g., active psychosis; severe depression requiring stabilization), unwillingness to discontinue opioid medication(s), or disruptive or otherwise concerning behavior that is expected to interfere with participation or the treatment milieu. If patients are deemed appropriate program candidates by the psychologist, they undergo a medical assessment for clearance to participate in the program, including a medical record review, medication review with a focus on polypharmacy, and history and physical exam with focus on a cardiac, orthopedic, and neurologic assessment and exercise clearance. A battery of self-report measures is administered at baseline and re-administered at discharge (day 15).

The program is delivered by a multidisciplinary team and integrates health psychology, medicine, nursing, physical therapy, and occupational therapy. The PRP emphasizes a biopsychosocial model of chronic pain treatment, with emphasis on evidence-based, nonpharmacological strategies and self-management. Group psychology sessions are delivered by health psychologists and include a variety of topics, including chronic pain education, activity pacing, relaxation skills, stress management, pain and relationships, and sleep hygiene, among many others. Approaches from Acceptance and Commitment Therapy (ACT) for chronic pain are emphasized, including values-based behavioral activation, mindfulness, cognitive defusion, and acceptance of chronic pain. Patients who enter the program on prescription opioids taper their medication while in the program. For additional details about the MUSC PRP, see Crouch et al. [21].

Measures

Pain Severity and Interference

Pain severity and interference were measured using the Brief Pain Inventory-Short Form (BPI) [38]. Respondents were asked to rate their worst (in past 24 hours), least (in past 24 hours), average, and current pain intensity on a scale of 0 (no pain) to 10 (pain as bad as you can imagine). Respondents then rated the degree to which pain has interfered in seven life areas (general activity, mood, walking ability, normal work, relations with other people, sleep, and enjoyment of life) in the past 24 hours. Pain interference was rated on a scale of 0 (does not interfere) to 10 (completely interferes). The BPI is a widely used, well-validated and reliable measure of pain and pain interference in many chronic pain populations (α = .80 –.92) [39].

Depression and Anxiety

Symptoms of depression and anxiety were measured using the Brief Symptom Inventory-18 (BSI-18) [40]. The BSI-18 is an 18-item, self-report measure that assesses the presence of severity of three psychological symptoms categories (Somatization, Depression, and Anxiety). The measure also produces a Global Severity Index score that measures overall psychological distress. For the purpose of this study, only the Depression and Anxiety subscales were used, which each have 6 Likert-type items and a possible subscale range of 0–24, with higher scores indicating higher levels of depression and anxiety symptom. The BSI-18 has been validated for use in multiple physical and mental health populations (α = .61–.94 with majority of scales over .80) [41].

Physical Functioning

Physical functioning was assessed using the 6-Minute Walk Test (6MWT). The 6MWT measures the distance (in feet) that an individual can walk on a flat, hard surface in a period of six minutes. It has been well-validated as a measure of functional exercise capacity across many chronic illnesses [42]. Individuals were instructed to self-pace their intensity and stop to rest if needed. The 6MWT was supervised and measured by the staff physical therapist and was conducted on an indoor, circular track. Clinically significant change is defined as a 50–100 ft increase.

Risk for Opioid Misuse

Opioid misuse risk was measured using the Current Opioid Misuse Measure (COMM) [43], which is a 17-item self-report measure. Items assess the frequency of certain aberrant opioid-related behaviors and known risk factors for opioid misuse over the past 30 days. Response options range from 0 (never) to 4 (very often). It is a widely used measure with demonstrated predictive validity and reliability among patients with chronic pain (α = .96) [43,44]. The range is 0–68, and a cutoff score of 9 has been recommended as indicative of significant opioid misuse risk. For the purpose of this study, the pretreatment administration of COMM followed standard instructions while the posttreatment administration asked for responses to be reflective of the past week rather than the past 30 days. This alteration was made to capture the most recent changes in responses to items rather than capturing responses based on functioning before or at the beginning of the program.

Statistical analyses

Differences in baseline medical and sociodemographic variables between Medicaid versus non-Medicaid participants were evaluated using Chi-square analyses for categorical variables and independent samples t-tests for continuous variables. First, paired samples t-tests were utilized to examine changes in outcomes from pre- to post-treatment among Medicaid and non-Medicaid patients separately. Then, two-factor repeated measures analyses of variance (ANOVA) were utilized to examine whether mean changes in treatment outcomes differed for Medicaid compared to non-Medicaid patients. We examined main effects of time (i.e., pre- to post-treatment changes) and interactions between time and group, to assess whether pre- to post-changes differed by Medicaid status. All tests were two-tailed with alpha set at .05. Effect sizes for ANOVAs were calculated using partial eta squared (η_p²); effect sizes are considered small at η_p² = 0.01, medium at η_p² = 0.06, and large at η_p² = 0.14 [45]. Effect sizes for t-tests were calculated using Cohen’s d and interpreted as small (d = 0.2), medium (d = 0.5), and large (d = 0.8) [43]. All analyses were conducted using IBM SPSS Statistics, Version 27.

Results

A total of N = 139 patients completed the program and pre- and post-treatment evaluations during the specified time period. Over one-quarter of the participants (29.0%) were enrolled in Medicaid. Eight participants were not included due to being uninsured (n = 6) or having a non-Medicaid MCO insurance (n = 2), leaving a full sample of N = 131 participants. There was no missing data in variables of interest. Demographic and medical variables, stratified by insurance type, are displayed in Table 1.

Table 1.

Sociodemographic and Pain Variables at Baseline, by Insurance Status

	Medicaid (n = 38)		Non-Medicaid (n = 93)
	n (%)	M (SD)	n (%)	M (SD)	Between-group difference
Gender					χ2=0.38, ns
Female	32 (84.2)		74 (79.6)
Male	6 (15.8)		19 (20.4)
Age		46.95 (12.29)		56.43 (13.56)	t=3.73**
Race					χ2=34.49**
Caucasian	12 (31.6)		76 (80.9)
African American	26 (68.4)		15 (16.0)
Other	0 (0.0)		2 (2.1)
Ethnicity					χ2=2.43, ns
Hispanic	3 (7.9)		2 (2.1)
Non-Hispanic	35 (92.1)		91 (96.8)
Primary diagnosis					χ2=8.70, ns
Back pain	18 (47.4)		44 (46.8)
Fibromyalgia	6 (15.8)		12 (12.8)
Arthritis	1 (2.6)		12 (12.8)
Chronic neck pain	3 (7.9)		7 (7.4)
Chronic abdominal pain	2 (5.3)		7 (7.4)
Neuropathy	3 (7.9)		3 (3.2)
Other	5 (13.2)		9 (9.6)
On opioids	12 (31.6)		53 (56.4)		χ2=10.65**

^**p<.001 (two-tailed)

Comparison of Baseline Characteristics, by Insurance Type

Medicaid participants were younger (M = 46.95; SD = 12.29) than non-Medicaid participants (M = 56.43; SD = 13.56), t(129) = 3.73, p<.001, and more likely to be African American (68.4%) than non-Medicaid participants (16.1%), χ² = 34.49, p<.001. Medicaid participants were less likely to be on opioids (31.6%) than non-Medicaid participants (56.4%), χ² = 10.65, p<.01. There were no significant differences in gender, ethnicity, or primary pain diagnosis, though the cohort was overall largely female and non-Hispanic.

In terms of clinical variables, at baseline, Medicaid participants demonstrated higher pain severity [t(130)=−3.65, p < .001], poorer physical functioning on the 6MWT [t(130)=3.03, p < .01], and higher levels of anxiety [t(130)=−2.45, p < .05]. There were no significant differences in baseline pain interference, depression, or opioid misuse risk (p > .05).

Comparison of Patient Changes in Treatment Outcomes, by Insurance Type

Examining treatment outcomes among Medicaid and non-Medicaid groups separately via paired samples t-tests, both groups showed statistically significant improvements across all treatment outcomes. See Figure 1. In terms of clinical significance, 86.8% of the Medicaid group and 86.2% of the non-Medicaid group demonstrated clinically meaningful improvement in physical functioning (at least 50ft increase on the 6MWT).

Figure 1.

Pre- to Post-Treatment Changes in Outcomes, by Insurance Type

*Statistically significant within group different from pre to post, p < .05

This figure demonstrates pre-treatment to post-treatment changes in clinical outcomes, separated by Medicaid status. Similar slopes of improvement are observed for the Medicaid and Non-Medicaid groups, with both groups showing statistically significant changes over time.

Repeated measures ANOVAs demonstrated that across all outcomes, there was a significant effect of time, indicating that all participants showed significant changes from pre- to post-treatment (See Table 3). However, interactions between time and group (insurance type) were nonsignificant across all outcomes (p > .05), indicating that rate of change did not differ by insurance type.

Table 3.

Analyses of Variance (ANOVAs): Main and Interaction Effects

Variable	F	df	p	ηp2
Pain severity
Time	21.04	1,129	<.001	.14
Insurance type x Time	0.90	1,129	.35	.01
Pain interference
Time	146.46	1,129	<.001	.53
Insurance type x Time	0.23	1,129	.63	.00
Depression
Time	117.96	1,129	<.001	.48
Insurance type x Time	0.05	1,129	.83	.00
Anxiety
Time	56.18	1,129	<.001	.30
Insurance type x Time	1.16	1,129	.28	.01
Physical functioning (6MWT)
Time	163.41	1,129	<.001	.56
Insurance type x Time	0.24	1,129	.63	.00
Opioid misuse risk
Time	93.96	1,129	<.001	.42
Insurance type x Time	1.72	1,129	.19	.01

Note. All tests were two-tailed. The p-values are bold when they are less than or equal to the significance level cut-off of .05. Effect sizes were calculated using partial eta squared (ηp²).

Discussion

Despite having higher pain severity and poorer physical functioning at baseline, Medicaid beneficiaries benefitted in a significant and similar way to non-Medicaid beneficiaries from participation in a 3-week, interdisciplinary Pain Rehabilitation Program. Significant improvements were seen across an array of measures, including pain severity, pain interference, 6-minute walk test, depression, anxiety, opioid misuse risk. There were no significant differences in rate of improvement between Medicaid versus non-Medicaid participants. This is particularly noteworthy, as those with Medicaid are at higher risk for (1) high-impact chronic pain, (2) risks from opioid prescribing and (3) poorer outcomes from surgical correction of pain [46–48], indicating that effective treatments for this patient population are sorely needed.

Differences at baseline between Medicaid beneficiaries and non-Medicaid beneficiaries in this sample were evident and consistent with prior research. Medicaid beneficiaries were more likely to be African American, have higher rates of pain severity, worse physical functioning, and lower rates of pain treatment (opioid medication) despite higher pain. These findings add support to the growing body of literature highlighting disparities in chronic pain treatment [7,8,49]. Increasing access to care for Medicaid beneficiaries is an important step in reducing these disparities. Promisingly, the results of the current study provide broad support for the benefits of pain rehabilitation across a variety of domains of functioning for both Medicaid and non-Medicaid beneficiaries.

Poor access to quality healthcare impacts racial and ethnic minority populations at higher rates than whites, even in Medicaid-insured populations and expansion states, with patient and provider level barriers playing a role in disparities [50]. Racial inequities are inextricably interwoven into socioeconomic and access inequities; thus, we intentionally did not control for baseline group differences, including race, as it has been noted that attempting to “control” for race when examining the impact of SES is misguided, both phenomenologically and statistically [51]. Findings of the current study should be interpreted with understanding that racial disparities are intertwined with Medicaid disparities. That noted, it is promising that the majority Black (68.4%) Medicaid group showed significant treatment improvements, given that Black patients face established disparities in pain functioning and pain treatment [8,52,53].

Treatment of chronic pain is complex, and an interdisciplinary rehabilitation model is well-suited to address the biopsychosocial aspects of chronic pain, providing both specific skills and evidence-based treatments as well as a supportive and interactive treatment milieu [54], which could be particularly important for marginalized groups. The patient-provider relationship in chronic pain treatment plays an important role in outcomes [57]. Many patients with chronic pain report feeling disregarded and marginalized while providers report feeling frustrated and manipulated. Non-specific therapeutic factors, such as trust and positive regard, play a key role in positive pain outcomes [55–56]. Pain rehabilitation, by its very nature, involves daily interactions with providers and other patients with chronic pain building a sense of understanding, trust and compassion. This may be particularly important in opioid tapering treatment [58].

Limitations of this study include the study sample representing almost exclusively residents from South Carolina in the US, a Medicaid non-expansion state. This population may not be representative of chronic pain populations in Medicaid expansion states. The findings also may not generalize to other healthcare systems in non-US countries. Further, income data was not collected, limiting ability to determine the role of other SES factors in outcomes. It is important to note that Medicaid beneficiaries were compared to a group of patients with a variety of other insurances, including both private and federal coverages, which is a heterogenous group. However, the primary purpose of this study was not to compare a variety of insurance types, but rather to specifically determine whether patients with Medicaid benefitted from interdisciplinary pain treatment in broad comparison to those with insurance coverages that generally provide greater access to care. Finally, this study did not include long-term follow-up, so it is unclear whether treatment gains were maintained over time for either group.

Conclusion

This study demonstrated that despite disparate access to nonpharmacological and opioid-sparing pain treatments [34,35,26], Medicaid beneficiaries who participated in a pain rehabilitation program demonstrated significant and similar treatment improvements compared to non-Medicaid participants, across a variety of functional and symptom domains. These results provide support for the importance of expanding access to these evidence-based chronic pain treatments for Medicaid beneficiaries.

Table 2.

Pre- and Post-Treatment Outcomes, by Insurance Type

	Medicaid (n = 38)		Non-Medicaid (n = 93)
Outcome Variable	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)
Pain Severity	6.95 (1.75)	6.29 (1.84)**	5.82 (1.55)	4.82 (1.95)**
Pain Interference	6.75 (2.12)	4.21 (2.80)**	6.04 (1.99)	3.19 (2.14)**
Depression	10.39 (6.62)	4.87 (4.66)**	8.45 (5.86)	3.14 (3.65)**
Anxiety	9.89 (6.77)	5.82 (5.51)**	7.16 (5.36)	4.11 (4.41)**
6MWT*	970.53 (413.36)	1219.59 (407.77)**	1204.89 (396.71)	1492.75 (399.80)**
Opioid misuse risk	11.76 (7.40)	6.82 (6.29)**	10.94 (7.03)	4.44 (3.59)**

^*6-minute walk test

^**Significant within-group pre- to post-treatment change at p<.001 (two-tailed)

Implications for Rehabilitation.

• Medicaid beneficiaries, who tend to be from lower socioeconomic status (SES), with chronic pain have poorer baseline functioning compared to non-Medicaid beneficiaries, including worse pain severity, poorer physical functioning, and higher levels of anxiety.

• Medicaid beneficiaries appear to benefit significantly and in a similar way to non-Medicaid individuals from participating in multidisciplinary, non-pharmacological chronic pain rehabilitation.

• Expanding access to evidence-based chronic pain rehabilitation, including nonopioid and multidisciplinary treatments, is a key component to address the chronic pain and opioid crises that differentially impact individuals from lower SES.