Health Service Utilization and State Costs Among Adults Aging With Early-Acquired Physical Disabilities in Medicaid Managed Care

Abstract

Objective: We evaluated the impact of Medicaid managed care (MMC) on health service use and state costs among adults with early-acquired physical disabilities. Method: Using claims data, we tracked utilization of the emergency department (ED), inpatient admissions, outpatient physician visits, and state expenditures on enrollees who transitioned to MMC (n = 881). The inverse propensity score weight and a difference-in-differences regression model were used to estimate the impact of MMC using their counterparts who remained in fee-for-service (n = 1,552) as the comparison group. Results: MMC reduced ED use by 3.2% points/month (p < .001). Relative to younger enrollees (age ⩽45 years), MMC reduced inpatient admissions of older enrollees (age ⩾46 years) by 3.3% points/month (p < .001), and state expenditures by US$839/month (p < .01). Discussion: MMC could reduce the hospital service use of and state spending on enrollees with early-acquired physical disabilities. This impact may vary depending on the enrollees’ age.

Keywords

emergency department quasi-experimental design propensity score difference-in-difference

Introduction

Adults who acquired physical disabilities early in their life use health services more frequently than the general population. Previous studies found that the inpatient hospital admissions and outpatient visits to physicians’ offices by adults with cerebral palsy (CP) were 9 times and 2 times, respectively, higher than those of adults in the general population (Young et al., 2007; Young et al., 2005). Adults with multiple sclerosis (MS) were twice as likely to visit the emergency department (ED) and 3.5 times more likely to be admitted to an inpatient unit of the hospital than their counterparts without MS (Asche, Singer, Jhaveri, Chung, & Miller, 2010). Adults with spina bifida (SB) made an average of 24 outpatient visits to physicians per year (Ouyang, Grosse, Armour, & Waitzman, 2007). Their utilization of hospital inpatient units was 11 times higher than that of the general population (Young et al., 2005). Adults with muscular dystrophy (MD) reported 36 outpatient visits per year (Ouyang, Grosse, & Kenneson, 2008), and higher usage of the ED and inpatient hospital units than those without MD (Mann, Royer, McDermott, et al., 2015). As a result of their persistent use of health services, health service costs for these adults were 3 to 10 times higher than that of the general population (Asche et al., 2010; Ouyang et al., 2007; Ouyang et al., 2008).

In an effort to control state expenditures and to provide better access to health services for enrollees, states have been transitioning Medicaid enrollees with disabilities from the fee-for-service (FFS) model to the Medicaid managed care (MMC) model (National Council on Disability, 2013; Sparer, 2012). Persons with disabilities constitute a relatively small subset of Medicaid enrollees, but they tend to utilize more acute health services than other enrollees and use a wide range of long-term services. Thus, they consume a disproportionate share of the state’s Medicaid expenditures (Kaiser Commission on Medicaid and the Uninsured, 2015). Under the MMC model, states pay managed care organizations (MCOs) a predetermined, fixed monthly premium for each enrollee (i.e., capitation payment) to stabilize the state’s Medicaid expenditures and increase the predictability of future spending (Sparer, 2012; Yamaki, Wing, Mitchell, Owen, & Heller, 2018). MCOs are supposed to use a variety of strategies to ensure that the enrollees have access to a continuum of high quality health care through their network of providers. Such strategies include the provision of care coordination, assignment of primary care physicians, referrals to specialists and ancillary service providers, and the establishment of financial incentives for providers (Bowers, Owen, & Heller, 2017). These approaches are expected to increase the utilization of community-based health services and decrease that of hospital-based health services, which inflate state Medicaid expenditures (Sparer, 2012). In 2016, 16 states provided acute health services for persons with physical disabilities through MMC (Smith et al., 2016).

The purpose of the present study was twofold. First, we examined how MMC affected the health service utilization and the state expenditures for adults with early-acquired physical disabilities. Although the population with disabilities includes heterogeneous subgroups, who may respond quite differently to MMC, past researchers often overlooked the divergence within the overall population of MMC enrollees with disabilities (Burns, 2009a, 2009b; Caswell & Long, 2015). Only a handful of researchers have investigated the MMC impact on a specific disability group (Shern et al., 2007; Yamaki, Wing, Mitchell, Owen, & Heller, 2019). None of studies mentioned above examined the MMC impact specifically on enrollees with early-acquired physical disabilities. We hypothesized that both enrollee utilization of hospital-based services and state expenditures would decrease under MMC, compared with the corresponding rates and amounts under FFS. Conversely, we hypothesized that enrollee utilization of community-based health services would increase under MMC.

Second, we examined the differential effect of MMC among older versus younger enrollees with early-acquired physical disabilities. The number of older individuals who are “aging with physical disability” is growing, and their health care needs would be different from their younger counterparts and the general population who are “aging into physical disability” (Molton & Yorkston, 2017). Their health service utilization and the associated state spending can be higher than these groups due to, for example, the possible progressive nature of their disability (Klingbeil, Baer, & Wilson, 2004; Yorkston, McMullan, Molton, & Jensen, 2010), the presence of secondary health conditions (Jain, Higgins, Katz, & Garshick, 2010; Jensen et al., 2013; Kinne, Patrick, & Doyle, 2004; Klingbeil et al., 2004; Marrie & Hanwell, 2013; McNalley et al., 2015; Molton et al., 2014), or the onset of age-related chronic health conditions (Charlifue, Jha, & Lammertse, 2010; Klingbeil et al., 2004; Marrie & Hanwell, 2013; Webb, 2010). Although the life expectancy of people with early-acquired physical disabilities has improved greatly (Haak, Lenski, Hidecker, Li, & Paneth, 2009; Liptak, 2008; Teoh, Geelhoed, Bayley, Leonard, & Laing, 2016; Webb, 2010), much of the research focuses on the health service utilization of young adults who transitioned from pediatric to adult services (Binks, Barden, Burke, & Young, 2007; Mann, Royer, McDermott, et al., 2015; Mann, Royer, Turk, et al., 2015; Young, Anselmo, Burke, McCormick, & Mukherjee, 2014; Young et al., 2007). Few researchers have investigated the health service utilization among older adults with early-acquired physical disabilities (Asche et al., 2010; Ouyang et al., 2007). We explored whether MMC had a different impact on the older enrollees who would be high users of health services relative to their younger counterparts.

Method

Data

The data for this study were extracted from data sets obtained from the state Medicaid agency and the two MCOs for the evaluation of the Integrated Care Program (ICP), part of Illinois’ Medicaid reform initiative aimed to control the state spending by expanding coordinated care (State of Illinois, 2011). Under the ICP, Medicaid services for all “seniors and persons with disabilities (SPD)” enrollees, excluding children below 18 years old and those who were dually eligible for Medicare, residing in suburban counties around the City of Chicago were transferred to one of two for-profit MCOs, selected via competitive procurement process, starting in May 2011 (Illinois Department of Healthcare and Family Services, 2016). The state data set included demographics, administrative information, and FFS claims of Medicaid enrollees who met aforementioned ICP eligibility and resided either in the Chicago suburbs (i.e., treatment region) or in the city of Chicago (i.e., comparison region) at least 1 month during the 9-month period from July 2010 to March 2011, and the 38-month period from May 2011 to June 2014. Starting in May 2011, this data set included a state’s monthly capitation payment amount to the MCOs for each enrollee who transitioned to MMC. The MCO data set included claims of the enrollees in the Chicago suburbs who transitioned to MMC during the 24-month period between January 2012 and December 2013.

Subjects

We identified any enrollee who were aged 19 years or older in July 2010, continuously enrolled in Medicaid for 33 months before and after ICP described later in “design” section, and had a pre-ICP period claim that included International Classification of Diseases 9th revision (ICD-9) codes for MS (340), hemiplegia (342), CP (343), paralysis (344), MD (359.0, 359.1, and 359.21), SCI (952), SB (741), polio (045 and 138), and osteogenesis imperfecta (756.51) as subjects with early-acquired physical disabilities. Subjects who resided in the Chicago suburbs in July 2010 were labeled as the treatment group. Those who resided in the city of Chicago were labeled as the comparison group. Using the intention-to-treat approach (Gupta, 2011), subjects were locked into the initially identified group even though they moved from the suburbs to the City or vice versa during the study period. Enrollees who resided in Intermediate Care Facilities for people with intellectual/developmental disabilities (ICF/DDs) or nursing facilities were excluded to focus on the impact of MMC on community residents and state costs supporting them. Appendix A graphically summarizes the flow of the sample selection. For the age-based analysis, enrollees aged 45 years or younger were labeled as the young group, whereas enrollees aged 46 years and above were labeled as the older group. We chose a lower than usual age cut off to take into account an early onset of age-related health conditions they often experience (Glew & Bennett, 2011; Haak et al., 2009; Turk, 2009) and the progressive nature of some of their primary disabilities (Klingbeil et al., 2004; Yorkston et al., 2010).

Variable Descriptions

Health service utilization

We measured hospital-based service utilization using ED visits and inpatient hospital admissions, and community-based health services utilization using ambulatory visits to primary care physicians (PCPs). We first identified these service utilization claims using specifications set by the Health care Effectiveness Data and Information Set (HEDIS), a performance measurement of health services used by the majority of health insurers in the United States (National Committee for Quality Assurance, 2016). HEDIS measures were adopted by the state as standardized quality measures to evaluate the performance of the two MCOs (Illinois Department of Healthcare and Family Services, 2012b), and used in the state’s project evaluating the ICP (Heller et al., 2014). We made some modifications of the HEDIS specification specific to the present study to include claims related to mental health conditions and to accommodate some limitations in our claims data as described below. We aggregated the identified claims to create outcome variables that represent a number of access to each service per subject per month.

ED visit

Using the HEDIS specification, any claims that met either one of the following two criteria were screened initially: (a) a claim included one of the following five Current Procedural Terminology (CPT) codes, 99, 281-99, 285 (Note: CPT code represents medical, surgical, and diagnostic procedures performed by the health service provider) or (b) a claim included one of 5,775 CPT codes defined as the “ED Procedure Code Value Set” and indicated “emergency room/free standing clinic” as “place of service” (Illinois Department of Healthcare and Family Services, 2012b). Per the HEDIS specification, we included only the first ED visit when enrollees visited ED multiple times a day, and excluded the ED visit followed by the same day inpatient hospital admission. Although the HEDIS specification calls for the use of the revenue code, a billing code for hospital services, we did not use it because the code was not available in the FFS claims data from the state. We did not exclude ED visits related to mental health and substance abuse, both of which HEDIS suggested to exclude, as our preliminary analysis had suggested that a significant proportion of hospital-based service utilization by enrollees with disabilities were associated with these exclusion criteria; therefore, removing them would undercount their ED utilization.

Inpatient hospital admission

Based on the HEDIS specification for “inpatient utilization (IPU),” we initially screened claims that met all of the following four inclusion criteria: (a) admission date and (b) discharge date were indicated, (c) the “inpatient admission indicator” was “Yes,” and (d) the “provider type” was “general hospital.” Then, the HEDIS specification calls for applying the IPU exclusion criteria, which removes (a) admissions related to mental health and behavior disorders, (b) infant delivery, and (c) surgery using either the discharge diagnosis or Diagnostic Related Group (DRG) code, a standardized diagnosis classification system for reimbursement rates for hospitalization. We kept admissions related to mental and behavioral disorders, and removed admissions related to infant delivery. The IPU exclusion criteria defined by the DRG code was not applied because our claims data did not include it.

Ambulatory visits to PCPs

First, we screened claims that included 1 of the 70 CPT codes, described in the HEDIS specification for “ambulatory outpatient visit (AMB).” Although the HEDIS AMB specification calls for the use of the revenue code and the aforementioned ED exclusion criteria, we did not use them for the reasons previously mentioned. Among the claims screened for AMB, we only kept claims of services provided by a physician, whose self-reported specialty was family practice, general practice, geriatrics, internal medicine, or pediatrics. Based on our interviews with caseworkers and nurses from pediatric specialty hospitals, conducted as part of the comprehensive ICP evaluation project (Heller et al., 2015), we included pediatricians because some young adults with disabilities had difficulty finding an adult provider and continued to see a pediatrician.

State health service expenditures

For each member of the comparison group, we calculated the state’s monthly expenditures by adding the amount of state reimbursement to providers for each FFS health service claims in a given month. For the treatment group members, we calculated the state’s monthly expenditures in the same manner for the pre-ICP period when the group was under FFS. For the post-ICP period, when they were under MMC, we used the sum of the state’s monthly capitation payment, a pre-determined set amount paid to MCOs for each subject in the group irrespective of his or her actual service utilization. Furthermore, we reduced the payment by US$62.20 per subject per month to account for the Medical Loss Ratio (MLR) payment. The state contract with MCOs specifies that each MCO spends at least 88% of the capitation revenues it collects each year on enrollee “benefit expenses.” If they fail to do so, they have to refund the difference to the state; it is called the MLR requirement. Thus, the MLR refund decreases the state’s expenditures for the treatment group. Although the 88% rule applies to the MCO as a whole, not at the level of individual enrollees, we calculated the per subject per month refund amount for the calendar year 2012 by dividing the total MLR payment evenly across all subjects in the treatment group. We subtracted it from the amount of the monthly capitated payment for each subject (Heller et al., 2015).

Design

To assess the causal effect of MMC on our study population, we used an inverse propensity weighted difference-in-difference (DID) design, a quasi-experimental research design that is robust to confounding from a broad class of threats to validity (Wing, Simon, & Bello-Gomez, 2018; Wong, Wing, Steiner, Wong, & Cook, 2013). At a broad level, we use inverse propensity score weights (IPW) to construct analytic city of Chicago samples that closely matched the Chicago suburbs samples. The outcome variables in our analysis were measures of each individual’s health service utilization (i.e., ED visits, inpatient admissions, and outpatient PCP visits) and the state’s total cost associated with each individual’s coverage as described above. We followed the treatment group members in the Chicago suburbs and the matched comparison group members in the city of Chicago across a 9-month pre-intervention period that extended from July 2010 to March 2011. Implementation of ICP in May 2011 in the Chicago suburbs marked the start of the intervention. We excluded the dates between May 2011 and December 2011, the first 8 months of ICP implementation, from the post-intervention period because of the progressive transition of enrollees to MMC and poor FFS data quality during these months (Heller et al., 2015). For the analysis of health service utilization, we set a 24-month post-intervention period, from January 2012 to December 2013, due to the availability of MCO data. For the analysis of the state health service expenditures, we set a 13-month post-intervention period from January 2012 to January 2013 because the costs for long-term services and supports (LTSS) were included in the capitation payment to the MCOs starting in February 2013.

Analysis

Our analysis consisted of two main stages. First, we used propensity score analysis to construct a matched comparison group using the ICP-eligible enrollees in the city of Chicago who closely resembled the treatment group enrollees in the Chicago suburbs with respect to a large set of covariates measured prior to the implementation of ICP. Second, we estimated the effect of MMC on utilization and cost measures using a DID regression framework. The DID model was designed to account for possible confounding and threats to validity that may have escaped the first stage matching procedure.

IPW

We fit logistic regressions of membership in the suburban sample on a large set of pre-ICP covariates including measures of the subject’s demographics, diagnoses related to disability, Home and Community-Based Services (HCBS) waiver enrollment, preexisting health conditions, historical health service utilization, and historical health service expenditures. The propensity scores were predicted values from the logistic regression model, and they represent the estimated probability that a comparison group member to represent the treatment group member given these covariates (Pattanayak, Rubin, & Zell, 2011). We used the propensity scores to construct IPW for each member of the city of Chicago sample. These weights served to down-weight Chicago members who had covariates that were very uncommon in the suburban sample and up-weight Chicago members who had covariates that were very common in the suburban sample. We imposed an overlap condition on our analysis by trimming observations from the Chicago sample who have estimated propensity scores that are smaller than the smallest propensity score in the suburban sample, and observations with estimated propensity score that are larger than the largest propensity score in the Chicago sample. We re-estimated the propensity scores after applying the trimming rule (Caliendo & Kopeinig, 2008).

DID regressions

The IPW procedure helps ensure that the suburban and Chicago samples were comparable with respect to measured pre-ICP characteristics. However, it is possible that the two groups differed with respect to some unmeasured covariates that may have affected health care utilization and health care costs. In addition, it was possible that utilization and health care costs may have changed over time in our sample for reasons that were not directly related to the ICP program. To account for these additional threats to validity, we estimated the effects of ICP by fitting panel data DID regression models to the IPW analytic sample.

The DID estimate of the treatment effect amounts to the difference between the average change in outcomes from pre- and post-ICP period in the suburban treatment group subjects and the average change in outcomes from pre- and post-ICP period in the matched Chicago comparison group subjects. Appendix B describes this DID design in a mathematical formula. These estimates represent the causal effect of the ICP intervention under two broad assumptions. First, any unmeasured time varying factor that affect costs and utilization (e.g., outbreak of a certain disease, changes in the health care market, and state’s administrative rules and regulations) must have affected both groups in the same way. Second, any unmeasured factor that differ across the suburban and Chicago samples must have remained the same before and after the ICP intervention. Together, these two restrictions require a common trends assumption, which implies that—in the absence of the ICP intervention—health care utilization and cost outcomes would have changed over time in the same way in the two groups (Dimick & Ryan, 2014; Wing et al., 2018).

One deviation from the basic DID design arose because Illinois implemented the Save Medicaid Access and Resource Together (SMART) Act (Public Act 097-0689) in July 2012, 7 months into the study’s post-ICP period. The SMART Act either eliminated or restricted funding for some Medicaid health services such as group psychotherapy, chiropractic services, nonemergency dental care, podiatry services for non-diabetics, and eyeglass provision (Heller et al., 2015; Illinois Department of Healthcare and Family Services, 2012a). The Act primarily impacted subjects in the comparison group who remained under FFS, as MCOs chose to maintain services at the pre-SMART Act levels in most cases (Heller et al., 2014). We controlled for the adoption of the SMART Act in our DID regressions.

To estimate the overall DID effect, we fit models with the following form:

\begin{array}{l} Model 1 : Y_{i t} = β_{0} + β_{1} (T r e a t m e n t_{i}) + β_{2} (T r e a t m e n t_{i} \times P o s t I C P_{t}) \\ + β_{3} (T r e a t m e n t_{i} \times P o s t S M A R T_{t}) + θ_{t} + ϵ_{i t} \end{array}

In Model 1, “Y_it” represented an outcome variable for the subject “i” in month “t.” Independent variables, Treatment_i, PostICP_t, and PostSMART_t were binary dummy variables. Treatment_i represented whether the subject_i was expected to transition to MMC. It was set to 0 for the comparison group member in the city of Chicago and 1 for the treatment group member in the suburb. PostICP_t noted if the observed month was before or after the start of ICP. It was set to 0 for the 9 months between July 2010 and March 2011, and set to 1 for the 24 months between January 2012 and December 2013. PostSMART_t referred to whether the observed month was before or after the implementation of the SMART Act. It was set to 0 for any months until June 2012, and 1 for months from July 2012 to December 2013. θ_t represents a full set of time period fixed effects, and є_it is an error term. β₂ is the DID estimate of the treatment effect of ICP prior to the adoption of the SMART Act, and β₃ is the estimate of the treatment effect of ICP after the adoption of the SMART Act. These coefficients represent measures of the average difference in the outcome variable between the two groups after accounting for baseline differences and time trends. Conceptually, the effect of ICP is defined relative to the health care cost and utilization outcomes that the treatment group members would have experienced if they had continued to receive care through FFS.

To estimate differential effects of the ICP intervention on older and younger enrollees, we fit an augmented DID model with the following form:

\begin{array}{l} Model 2 : Y_{i t} = β_{0} + β_{1} (T r e a t m e n t_{i} \times P o s t I C P_{t}) + \\ β_{2} (T r e a t m e n t_{i} \times P o s t I C P_{t} \times G_{i}) + \\ β_{3} (T r e a t m e n t_{i} \times P o s t S M A R T_{t}) + \\ β_{4} (T r e a t m e n t_{i} \times P o s t S M A R T_{t} \times G_{i}) + \\ β_{5} (G_{i}) + β_{6} (T r e a t m e n t_{i}) + \\ β_{7} (T r e a t m e n t_{i} \times G_{i}) + θ_{t} + ϵ_{i t} \end{array}

In Model 2, G_i is a binary variable set to 1 if the enrollee was aged 46 years or above during the pre-ICP period. In this framework, β₁ represents the DID effect of the pre-SMART effect of ICP on people below the age of 45 years, and β₃ represents the DID effect of ICP in the post-SMART period. β₂ and β₄ represent the differential effects on ICP among enrollees aged 46 years and above. Analyses were conducted using STATA ver. 15 (StataCorp, 2017), and the level of significance was set at .05. The Institutional Review Board of the first author approved this study protocol (reference # 2011–0363).

Results

Table 1 compares covariates in the pre-intervention period among the treatment group (Column 1), unmatched (Column 2), and IPW matched (Column 4) comparison group using within-group proportions and counts. Demographic characteristics between the two groups (i.e., Columns 1 and 2) were similar except for the racial breakdown. White (37.3%) was the largest racial group in the treatment group, whereas a majority of the comparison group identified as Black (65.9%). Across the two groups, almost all subjects had either MS, paralysis, CP, or hemiplegia, one third were 46-years-old or older with the oldest age being 89 and 88 years for the treatment and the comparison group, respectively, and one third accessed the Person with Disabilities waiver program. A profile of the comparison group after applying the IPW (Column 4) was more similar to the treatment group than the unweighted group represents.

Table 1.

Selected Baseline Covariates Comparison Between Treatment and Comparison Groups.

Covariates	Treatment (n = 881)	Comparison (n = 1,552)
		Unmatched		IPW matched
		Proportion	SMD^a	Proportion	SMD^b
Demographics
Female	0.530	0.446	0.169	0.527	0.006
White	0.373	0.130	0.593	0.405	−0.065
Black	0.394	0.659	−0.549	0.373	0.043
Hispanic	0.135	0.146	−0.032	0.149	−0.038
Ages ⩾46 (years)	0.327	0.382	−0.116	0.346	−0.041
Physical disability diagnosis^c
MS	0.160	0.150	0.027	0.155	0.015
MD	0.043	0.025	0.100	0.044	−0.002
Polio	0.012	0.012	0.008	0.015	−0.024
SCI	0.054	0.077	−0.090	0.054	0.002
SB	0.073	0.057	0.062	0.071	0.007
Paralysis	0.285	0.363	−0.168	0.285	−0.001
CP	0.396	0.265	0.281	0.389	0.016
Hemiplegia	0.141	0.224	−0.217	0.159	−0.050
OI	0.006	0.004	0.026	0.010	−0.049
HCBS
Person with disabilities	0.302	0.365	−0.135	0.305	−0.007
DD	0.195	0.076	0.360	0.174	0.055
No waiver	0.388	0.449	−0.124	0.393	−0.011
Health expenditure^d (US$)	14,401	17,756	−0.116	15,642	−0.047
Health condition^e	7.673	8.404	−0.097	8.274	−0.079
Health service utilization^f
Emergency department	0.877	0.992	−0.050	0.887	−0.004
Hospital admission	0.592	0.837	−0.161	0.611	−0.014
Primary care physician	2.318	1.947	0.101	2.377	−0.015
Dentist	0.297	0.245	0.073	0.305	−0.010
Prescription drug^g	19.965	17.079	0.157	21.680	−0.080

Note. IPW = inverse propensity score weights; SMD = standardized mean difference; MS = Multiple Sclerosis; MD = Muscular Dystrophy; SCI = Spinal Cord Injury; SB = Spina Bifida; CP = Cerebral Palsy; OI = Osteogenesis Imperfecta; HCBS = Home and Community-Based Service; DD = Developmental Disabilities.

Standardized mean difference between the treatment group and unmatched comparison group.

Standardized mean difference between the treatment group and IPW-matched comparison group.

Not mutually exclusive.

Total Medicaid health expenditures per person during the pre-intervention period.

Mean number of health condition categories of the Clinical Classification Software per person.

Mean number of accessed services per person during the pre-intervention period.

Mean number of medication claims per person during the pre-intervention period.

We used the standardized mean difference (SMD) to examine if the application of IPW has resulted in better balance of covariates between the two groups (Austin, 2009). SMDs between the treatment and comparison group before and after the application of IPW are shown in Columns 3 and 5, respectively. SMD for each covariate after IPW-matching was smaller than SMD before applying IPW. After matching, SMD is less than .08 standard deviations for each covariate which is within the range of tolerance that was recommended by literature on propensity score matching (Austin, 2009; Stuart, Lee, & Leacy, 2013). These findings suggest that the application of IPW successfully made the two groups similar enough to reduce the risk of biases due to preexisting group differences.

Table 2 summarizes the estimated impact of the MMC and the SMART Act on the four outcome variables using coefficients of the DID regression (Model 1) with the IPW-matched comparison group. In Column 1, an average monthly proportion of subjects who visited ED was estimated as 8.3% (β₀ + β₁) for the treatment group during the pre-ICP period. Implementation of ICP, then, lowered the proportion significantly by 3.2% (β₂), t = −3.81, p < .001, to 5.2% (β₀ + β₁ + β₂). The implementation of the SMART Act further decreased the proportion by 0.3% (β₃) although it did not reach statistical significance. When combining the impact of both the MMC and the SMART Act, the net change of ED visits between the pre- and post-ICP periods for the treatment group was estimated to be −3.5% (β₂ + β₃), F = 29.3, p < .001. There was no observable MMC impact on the other three outcome variables. The SMART Act did not have a significant impact on any of the outcome variables.

Table 2.

Effects of the MMC and the SMART Act on the Health Service Utilization and the State Health Service Expenditures Among Enrollees With Physical Disabilities.

	Health service utilization (%)			Health expenditures (US$)^a
	ED visit	Inpatient hospital admission	Ambulatory PCP visit	Health expenditures (US$)^a
Suburban effect (β₁)	−0.00172 (−0.23)	−0.00316 (−0.52)	−0.00225 (−0.16)	−137.9 (−0.87)
ICP effect (β₂)	−0.0318*** (−3.81)	−0.00343 (−0.55)	0.0000584 (0.00)	−91.87 (−0.51)
SMART Act effect (β₃)	−0.00339 (−0.49)	0.00315 (0.52)	0.0111 (1.12)	21.66 (0.14)
Intercept (β₀)	0.0852*** (9.25)	0.0532*** (8.81)	0.198*** (13.07)	1,776.5*** (12.75)
Person-fixed effect	Yes	Yes	Yes	Yes
Month-fixed effect	Yes	Yes	Yes	Yes
Inverse propensity score weight	Yes	Yes	Yes	Yes
N	80,289	80,289	80,289	53,526

Note. t statistics in parentheses. MMC = Medicaid managed care; SMART = Save Medicaid Access and Resource Together; ED = emergency department; PCP = primary care physicians; ICP = Integrated Care Program.

Post-ICP data cover until January 2013.

p < .05. **p < .01. ***p < .001.

Table 3 compares the health service utilization by and the state expenditures on the younger (ages ⩽45 years) and older (ages ⩾46 years) subjects in the treatment group as well as the unmatched comparison group across the pre- and post-ICP implementation. The monthly average percentage of older subjects who accessed the three health services was significantly higher than those of their younger counterparts in both the treatment and comparison groups for both time periods. The mean monthly state health expenditures per person for the older subjects were significantly higher than that of the younger subjects in the pre-ICP period for both treatment and comparison groups.

Table 3.

Health Service Utilization and State Expenditures Among Adults With Early-Acquired Physical Disabilities by Age Group, Before and After ICP Implementation.

	Utilization^a (%)/expenditures^b (U$)
	Age ⩽45 years	Age ⩾46 years	χ² or t value^c
Pre-ICP
Treatment group
ED visit	6.7	9.4	17.44***
Hospital admission	4.1	8.6	68.12***
Ambulatory visit to PCP	15.7	28.1	169.80***
Health service expenditures	1,308	2,200	−3.99***
Comparison group
ED visit	7.9	10.2	23.26***
Hospital admission	6.4	9.9	60.10***
Ambulatory visit to PCP	12.1	24.4	359.22***
Health service expenditures	1,787	2,272	−2.82**
Post-ICP
Treatment group
ED visit	3.2	4.5	21.39***
Hospital admission	3.6	5.7	52.08***
Ambulatory visit to PCP	16.0	27.5	383.25***
Health service expenditures^d	1,160	1,204	−1.49
Comparison group
ED visit	7.8	9.1	20.58***
Hospital admission	5.1	5.9	10.79**
Ambulatory visit to PCP	12.3	23.2	761.73***
Health service expenditure^d	1,442	1,609	−1.10

Note. ICP = Integrated Care Program; ED = emergency department; PCP = primary care physicians.

Mean monthly % of group members who accessed the service.

Mean monthly state expenditures in US$ amount.

χ² value for ED visit, hospital admissions, and outpatient visit to PCP. t value for state expenditures.

Post-ICP expenditures between January 2012 and January 2013.

p < .05. **p < .01. ***p < .001.

Table 4 summarizes the estimated MMC and SMART Act impact between the two age groups using coefficients of the DID regression (Model 2). In Row 1, the MMC reduced the ED visits for both age groups consistently. Relative to if they remained under FFS, it reduced the proportion of the older group members who visited an ED by 3.8% (β₁ + β₂), F = 12.98, p < .001, and that of the younger group members by 2.9% (β₁), t = −3.22, p < .01. The MMC impact varied between the two age groups on hospital admissions and state health expenditures. In Row 2, it was estimated that MMC decreased the proportion of the older group members who were admitted to the hospital by 2.57%, F = 7.23, p < .01, relative to if they remained under FFS. For the young group, there was no observable MMC impact. The cumulative difference of the MMC impact between the two age groups was −3.31%, t = −3.51, p < .001. In Row 4, similarly, the MMC decreased the average state health service expenditures for the older subjects by US$657/person/month, F = 6.10, p < .05, relative to if they remained under FFS. In contrast, the MMC impact on the young subjects was insignificant. The cumulative difference between the two groups was –US$839/person/month, t = −3.29, p < .01. There was no observable, differential impact of the SMART Act across the age groups.

Table 4.

Effects of the MMC and the SMART Act on Health Service Utilization and State Medicaid Health Service Expenditure on Enrollees With Physical Disabilities Between Two Age Groups.

	Impact^a (%)
	Age ⩽46 years(F value)^b	Age ⩾45 years(t value)	Difference(t value)
ICP	(β₁ + β₂)	(β₁)	(β₂)
Emergency department visit (%)	−0.0379*** (12.98)	−0.0288** (−3.22)	−0.0091 (−0.94)
Hospital admission (%)	−0.0257** (7.23)	0.0074 (1.14)	−0.0331*** (−3.51)
Outpatient visit (%)	−0.0168 (0.54)	0.0082 (0.54)	−0.0251 (–1.17)
Health expenditures^c (US$)	−657.00* (6.10)	182.60 (0.99)	−839.60** (−3.29)
SMART Act	(β₃ + β₄)	(β₃)	(β₄)
Emergency department visit (%)	−0.0076 (0.85)	−0.0013 (−0.19)	−0.0063 (–1.00)
Hospital admission (%)	0.0114 (2.16)	−0.0009 (−0.13)	0.0123 (1.80)
Outpatient visit (%)	0.0246 (2.44)	0.0044 (0.43)	0.0202 (1.28)
Health expenditures^c (US$)	66.3532 (0.18)	−0.0468 (−0.00)	66.4 (1.18)

Note. MMC = Medicaid managed care; SMART = Save Medicaid Access and Resource Together; ICP = Integrated Care Program.

Average impact on the treatment group member relative to the IPW-matched comparison group in the same age group.

F value from joint significance test.

Post-ICP data cover until January 2013.

p < .05. **p < .01. ***p < .001.

Discussion

In the present study, we examined the impact of MMC on the health service utilization of enrollees with early-acquired physical disabilities residing in the community as well as the corresponding state health service expenditures. DID analyses suggested that transitioning enrollees with early-acquired physical disabilities from FFS to MMC decreased their ED utilization. This was supported by the use of an IPW-matched comparison group that did not transition to MMC and remained under FFS in the post-ICP period. Furthermore, the MMC reduced hospital admissions and their state health service expenditures more significantly in the older group than it did on the younger group. This may be the first study examining the impact of the state’s transformation of its health service delivery model on Medicaid enrollees with early-acquired physical disabilities and its subgroups.

Although MMC was intended to provide better access to community-based health services for enrollees (Sparer, 2012), the present findings suggest that community-based health service utilization by enrollees with early-acquired physical disabilities did not change significantly under MMC. Previous findings on MMC and community services utilization showed mixed results. Consistent with the present findings, Beatty et al. (2003) reported that the access to PCPs by enrollees with physical disabilities didn’t differ between FFS and MMC. Other researchers reported that MMC was associated with reduced access to community services including both primary care and specialist services in enrollees with disabilities in general (Burns, 2009b) and specifically with enrollees with intellectual and developmental disabilities (IDD) (Yamaki et al., 2019). Some researchers observed increased utilization of community-based services only among MMC enrollees with disabilities in urban areas where the services were readily available (Coughlin, Long, & Graves, 2008). There are at least two potential reasons that we did not find MMC enrollment to have a significant impact on community-based health service utilization. First, non-physician providers, such as nurses and physician assistants, were not included in the present analysis. MCOs often use these providers as lower-cost substitutes for PCPs (Coughlin et al., 2008). Second, we did not include visits to specialty physicians. Complex health care needs of adults with physical disabilities could be beyond the expertise of PCPs and require the ongoing involvement of specialists (Dicianno & Wilson, 2010). Thus, access to specialists could constitute an important component of the community-based health services for this population (Beatty et al., 2003; Sawyer & Macnee, 2010; Young et al., 2007). Future studies should consider the inclusion of non-physician and specialty service provider use as an indicator of access to community-based health services.

Within the scope of the present study, we were unable to determine specific reasons for the reduced ED utilization under MMC among our subjects. It could be simply an indication of limited access to ED under MMC. MCO strategies implemented in the ICP (e.g., assignment of a PCP to each enrollees, referrals, and use of case management) might have effectively redirected our subjects who previously seeking ED services to alternative providers such as retail health and urgent care clinics (Kodjak, 2015; Puleo, 2014; Verdier et al., 2009; Weinick, Burns, & Mehrotra, 2010). It could be attributable, at least partially, to a decrease of ED visits due to conditions that were not emergencies and/or treatable in the community-based health services. A notable proportion of ED visits by adults with physical disabilities and IDD has been associated with such conditions (Mann, Royer, McDermott, et al., 2015; Mann, Royer, Turk, et al., 2015; Yamaki et al., 2019). Previous findings have suggested that MMC was effective in reducing their ED visits with these conditions (Verdier et al., 2009; Yamaki et al., 2019). Future studies should explore potential causes of the ED reduction to understand the MMC impact on the target population better.

The finding that MMC did not reduce state health service expenditures for enrollees with early-acquired physical disabilities is consistent with some previous studies of the ICP as a whole (Heller et al., 2015) and enrollees with IDD within the ICP (Yamaki et al., 2018). Other studies on MMC’s cost impact showed mixed results. The Lewin Group (2009) found that MMC reduced state expenditures over FFS on enrollees with disability, whereas others reported that MMC costed as much as FFS for the state (Burns, 2009a; Caswell & Long, 2015; Duggan & Hayford, 2013; Shern et al., 2007). Illinois’ low FFS reimbursement rate could explain the lack of MMC’s cost impact (Zuckerman & Goin, 2012). When FFS reimbursement rates for providers are low, the state’s health services spending on enrollees under FFS might be already very low. In this environment, MMC was less likely to result in additional cost savings for states because there had been little room for MCOs to bargain for lower prices with providers (Caswell & Long, 2015; Duggan & Hayford, 2013). Nonetheless, the state could stabilize its expenditures and increase its predictability for future spending with the monthly fixed capitation payment to MCOs, important benefits for the states with its limited cash flow (Yamaki et al., 2018).

In the general population, older adults typically use ED, inpatient hospitalization, and ambulatory care more frequently than young adults (Aminzadeh & Dalziel, 2002; Buie, Owings, DeFrances, & Golosinskiy, 2010; Schappert & Rechtsteiner, 2011). Our finding suggests that older enrollees with early-acquired physical disabilities are also higher users of health services than their younger counterparts. The present study did not support findings of a previous study that found there was little age variation in inpatient admission and ambulatory visits among adults with SB (Ouyang et al., 2008). This difference could be explained by the use of different data sources between these two studies. Ouyang et al. (2008) suggested the private insurance data they used might not include older individuals who had used up their lifetime insurance allowance and switched to public insurance. In addition, we suspect that those who were denied coverage in the private sector due to preexisting conditions might not be included in their data. Using Medicaid data, the present study could have captured more representative samples of older adults with early-acquired physical disabilities by including those who were heavy users of services. As the longevity of people with early-acquired physical disabilities has been improving, the numbers of older persons with physical disabilities will likely increase (Haak et al., 2009; Liptak, 2008; Webb, 2010). The current lack of information on potential causes of their higher health service utilization may need to be explored further to respond to their emerging health issues.

We found that MMC’s impact might not be equal among subgroups of adults with early-acquired physical disabilities. The impact of MMC was found to be stronger among older enrollees who use more health services. One interpretation of this finding could be that MCO’s strategies (e.g., care coordination, the assignment of a PCP, and referrals) were more efficient in reducing reliance on hospital-based services, which in turn reduced the state expenditures among high users. In fact, there are several reports indicating that case management and referrals were effective in reducing repeated use of the ED (Allen Conner, 2015; Bindman, Chattopadhyay, Osmond, Huen, & Bacchetti, 2005; Kodjak, 2015; Puleo, 2014; Verdier et al., 2009) and inpatient services (Ness & Kramer, 2013). Previous researchers reported that MMC impact might vary depending on the target population, type of MMC, and availability of health services in the target area (Baker & Afendulis, 2005; Freund et al., 1989; Garrett & Zuckerman, 2005; Pollack, Wheeler, Cowan, & Freed, 2007; Verdier et al., 2009). Our findings build upon these findings by empirically suggesting that the level of health service utilization might be another predictor of MMC impact.

Interpretation of the present findings warrants several limitations. First, the present findings may have limited generalizability. Given the variation of Medicaid programs and specific procedures of MMC implemented across states, our findings may be specific to individuals who were exposed to Illinois’ ICP. Second, the findings on state health service expenditures were based on observation during the early phases of MMC implementation. The observations excluded policy changes, implemented outside of the observation period, which might impact the state’s Medicaid expenditures (Heller et al., 2015; Heller et al., 2014). Third, we didn’t examine the fiscal impact of MMC on LTSS. Nationally, there are at least eight states which transferred LTSS for enrollees with physical disabilities to MMC in 2012 (Saucier, Kasten, Burwell, & Gold, 2012). Examining the financial and service implications of MMC on LTSS, particularly HCBS Persons with Disability waiver services, for these states may provide important MMC program information to state program planners and advocates.

Nationally, Medicaid spending has grown rapidly each year, reaching 20% of states’ total general fund spending in fiscal year 2017 (National Association of State Budget Officers, 2017). Because of budget constraints, many states have been transitioning enrollees with disabilities to MMC with a hope of both reducing or stabilizing the state’s costs and ensuring better access to health services for enrollees (Sparer, 2012). Yet, previous researchers have reported mixed findings on the impact of MMC (Burns, 2009b; Caswell & Long, 2015; Coughlin et al., 2008; Shern et al., 2007; Yamaki et al., 2019). For persons with early-acquired physical disabilities, Medicaid is a primary source of health services required to meet their complex health care needs, some of which may intensify as they age. Their reliance on Medicaid may make them particularly vulnerable to its changing health service delivery model (Heller, Owen, Bowers, & Gibbons, 2017). However, studies on the impact these changes have on persons with physical disabilities and their subgroups are scarce. Few researchers examined the health outcomes of these individuals, another goal of MMC along with those examined in the present study (Berwick, Nolan, & Whittington, 2008; Bowers et al., 2017; Heller et al., 2017). Monitoring changes in Medicaid programs at the state level and cumulating more empirical data in achieving intended goals are imperative to better understand and advocate for health care needs of this population.

Footnotes

Appendix A

Appendix B

Difference-in-Differences Design.

Group	Pre-intervention period	Post-intervention period	Difference
Treatment (Chicago suburbs)	Y₁₁ (MMC group baseline)	Y₁₂ (MMC group baseline + Trend + ICP effect)	ΔY_T = Y₁₂ – Y₁₁ (Trend + ICP effect)
Comparison (city of Chicago)	Y₂₁ (FFS group baseline)	Y₂₂ (FFS group baseline + Trend)	ΔY_C = Y₂₂ – Y₂₁ (Trend)
Difference in differences			ΔY_T – ΔY_C (ICP effect)

Note. MMC = Medicaid managed care; ICP = Integrated Care Program; FFS = fee-for-service.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The contents of this article were developed under grants from the United States Department of Health and Human Services (DHHS), Administration for Community Living (ACL), National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) Grant # 90RT5020-04, 90RT5023-03-00, and 90RT5026-01-03. However, these contents do not necessarily represent the policy of DHHS, and you should not assume endorsement by the Federal Government.

ORCID iD

Kiyoshi Yamaki

References

Allen Conner

. (2015). Grounding healthcare’s frequent fliers. Retrieved from https://www.healthify.us/healthify-insights/grounding-healthcares-frequent-fliers

Aminzadeh

Dalziel

W. B.

(2002). Older adults in the emergency department: A systematic review of patterns of use, adverse outcomes, and effectiveness of interventions. Annals of Emergency Medicine, 39, 238-247. doi:10.1067/mem.2002.121523

Asche

C. V.

Singer

M. E.

Jhaveri

Chung

Miller

(2010). All-cause health care utilization and costs associated with newly diagnosed multiple sclerosis in the United States. Journal of Managed Care Pharmacy, 16, 703-712. doi:10.18553/jmcp.2010.16.9.703

Austin

P. C.

(2009). Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Statistics in Medicine, 28, 3083-3107. doi:10.1002/sim.3697

Baker

L. C.

Afendulis

(2005). Medicaid managed care and health care for children. Health Services Research, 40(5 Pt 1), 1466-1488. doi:10.1111/j.1475-6773.2005.00427.x

Beatty

P. W.

Hagglund

K. J.

Neri

M. T.

Dhont

K. R.

Clark

M. J.

Hilton

S. A.

(2003). Access to health care services among people with chronic or disabling conditions: Patterns and predictors. Archives of Physical Medicine and Rehabilitation, 84, 1417-1425.

Berwick

D. M.

Nolan

T. W.

Whittington

(2008). The triple aim: Care, health, and cost. Health Affairs, 27, 759-769. doi:10.1377/hlthaff.27.3.759

Bindman

A. B.

Chattopadhyay

Osmond

D. H.

Huen

Bacchetti

(2005). The impact of Medicaid managed care on hospitalizations for ambulatory care sensitive conditions. Health Services Research, 40, 19-38. doi:10.1111/j.1475-6773.2005.00340.x

Binks

J. A.

Barden

W. S.

Burke

T. A.

Young

N. L.

(2007). What do we really know about the transition to adult-centered health care? A focus on cerebral palsy and spina bifida. Archives of Physical Medicine and Rehabilitation, 88, 1064-1073. doi:10.1016/j.apmr.2007.04.018

10.

Bowers

Owen

Heller

(2017). Care coordination experiences of people with disabilities enrolled in Medicaid managed care. Disability and Rehabilitation, 39, 2207-2214. doi:10.1080/09638288.2016.1219773

11.

Buie

V. C.

Owings

M. F.

DeFrances

C. J.

Golosinskiy

(2010). National Hospital Discharge Survey: 2006 summary. Washington, DC: U.S. Department of Health and Human Services. Retrieved from https://www.cdc.gov/nchs/data/series/sr_13/sr13_168.pdf

12.

Burns

M. E.

(2009a). Medicaid managed care and cost containment in the adult disabled population. Medical Care, 47, 1069-1076. doi:10.1097/MLR.0b013e3181a80fef

13.

Burns

M. E.

(2009b). Medicaid managed care and health care access for adult beneficiaries with disabilities. Health Services Research, 44(5 Pt 1), 1521-1541. doi:10.1111/j.1475-6773.2009.00991.x

14.

Caliendo

Kopeinig

(2008). Some practical guidance for the implementation of propensity score matching. Journal of Economic Surveys, 22, 31-72.

15.

Caswell

K. J.

Long

S. K.

(2015). The expanding role of managed care in the Medicaid program: Implications for health care access, use, and expenditures for nonelderly adults. INQUIRY: The Journal of Health Care Organization, Provision, and Financing, 52, 1-17. doi:10.1177/0046958015575524

16.

Charlifue

Jha

Lammertse

(2010). Aging with spinal cord injury. Physical Medicine and Rehabilitation Clinics of North America, 21, 383-402. doi:10.1016/j.pmr.2009.12.002

17.

Coughlin

T. A.

Long

S. K.

Graves

J. A.

(2008). Does managed care improve access to care for Medicaid beneficiaries with disabilities? A national study. INQUIRY: The Journal of Health Care Organization, Provision, and Financing, 45, 395-407.

18.

Dicianno

B. E.

Wilson

(2010). Hospitalizations of adults with spina bifida and congenital spinal cord anomalies. Archives of Physical Medicine and Rehabilitation, 91, 529-535. doi:10.1016/j.apmr.2009.11.023

19.

Dimick

J. B.

Ryan

A. M.

(2014). Methods for evaluating changes in health care policy: The difference-in-differences approach. Journal of the American Medical Association, 312, 2401-2402. doi:10.1001/jama.2014.16153

20.

Duggan

Hayford

(2013). Has the shift to managed care reduced Medicaid expenditures? Evidence from state and local-level mandates. Journal of Policy Analysis and Management, 32, 505-535. doi:10.1002/pam.21693

21.

Freund

D. A.

Rossiter

L. F.

Fox

P. D.

Meyer

J. A.

Hurley

R. E.

Carey

T. S.

Paul

J. E.

(1989). Evaluation of the Medicaid competition demonstrations. Health Care Financing Review, 11, 81-97.

22.

Garrett

Zuckerman

(2005). National estimates of the effects of mandatory Medicaid managed care programs on health care access and use, 1997-1999. Medical Care, 43, 649-657.

23.

Glew

G. M.

Bennett

F. C.

(2011). Cerebral palsy grown up. Journal of Developmental and Behavioral Pediatrics, 32, 469-475. doi:10.1097/DBP.0b013e3182202d42

24.

Gupta

S. K.

(2011). Intention-to-treat concept: A review. Perspectives in Clinical Research, 2, 109-112. doi:10.4103/2229-3485.83221

25.

Haak

Lenski

Hidecker

M. J.

Paneth

(2009). Cerebral palsy and aging. Developmental Medicine & Child Neurology, 51(Suppl. 4), 16-23. doi:10.1111/j.1469-8749.2009.03428.x

26.

Heller

Owen

Bowers

Gibbons

H. M.

(2017). Aging with a physical disability in Medicaid managed care. Research on Aging, 39, 778-798. doi:10.1177/0164027516681049

27.

Heller

Owen

Mitchell

Eisenberg

Wing

Bowers

. . . Viola

(2015). An independent evaluation of the integrated care program: Final report: Findings through the third year (FY14). Chicago, IL. Institute on Disability and Human Development, The University of Illinois at Chicago.

28.

Heller

Owen

Mitchell

Keys

Viola

Eisenberg

. . . Wing

(2014). An independent evaluation of the integrated care program: Findings from the baseline through year two (FY13). Chicago, IL: Institute on Disability and Human Development, The University of Illinois at Chicago.

29.

Illinois Department of Healthcare and Family Services. (2012a). Coverage changes resulting from the Save Medicaid Access and Resources Together (SMART) Act. Retrieved from https://www.illinois.gov/hfs/MedicalProviders/notices/Pages/prn120629d.aspx

30.

Illinois Department of Healthcare and Family Services. (2012b). Quality measures specifications for care coordination program. Springfield, IL.

31.

Illinois Department of Healthcare and Family Services. (2016). Integrated care program. Retrieved from http://www.illinois.gov/hfs/medicalproviders/cc/icp/Pages/default.aspx

32.

Jain

N. B.

Higgins

L. D.

Katz

J. N.

Garshick

(2010). Association of shoulder pain with the use of mobility devices in persons with chronic spinal cord injury. PM&R, 2, 896-900. doi:10.1016/j.pmrj.2010.05.004

33.

Jensen

M. P.

Truitt

A. R.

Schomer

K. G.

Yorkston

K. M.

Baylor

Molton

(2013). Frequency and age effects of secondary health conditions in individuals with spinal cord injury: A scoping review. Spinal Cord, 51, 882-892. doi:10.1038/sc.2013.112

34.

Kaiser Commission on Medicaid and the Uninsured. (2015). Medicaid at 50. Retrieved from http://files.kff.org/attachment/report-medicaid-at-50

35.

Kinne

Patrick

D. L.

Doyle

D. L.

(2004). Prevalence of secondary conditions among people with disabilities. American Journal of Public Health, 94, 443-445.

36.

Klingbeil

Baer

H. R.

Wilson

P. E.

(2004). Aging with a disability. Archives of Physical Medicine and Rehabilitation, 85(7 Suppl. 3), S68-S73; quiz S74-S65.

37.

Kodjak

(2015, October). A hospital reduces repeat ER visits by providing social workers. NPR. Retrieved from https://www.npr.org/sections/health-shots/2015/10/23/451154605/a-hospital-reduces-repeat-er-visits-by-providing-social-workers

38.

Lewin Group. (2009). Medicaid managed care cost savings—A synthesis of 24 studies. Retrieved from http://blogs.chicagotribune.com/files/lewinmedicaid.pdf

39.

Liptak

G. S.

(2008). Health and well being of adults with cerebral palsy. Current Opinion in Neurology, 21, 136-142. doi:10.1097/WCO.0b013e3282f6a499

40.

Mann

J. R.

Royer

J. A.

McDermott

Hardin

J. W.

Ozturk

Street

(2015). Hospitalizations and emergency room visits for adolescents and young adults with muscular dystrophy living in South Carolina. Muscle & Nerve, 52, 714-721. doi:10.1002/mus.24599

41.

Mann

J. R.

Royer

J. A.

Turk

M. A.

McDermott

Holland

M. M.

Ozturk

O. D.

. . . Thibadeau

J. K.

(2015). Inpatient and emergency room visits for adolescents and young adults with spina bifida living in South Carolina. PM&R, 7, 499-511. doi:10.1016/j.pmrj.2014.11.011

42.

Marrie

R. A.

Hanwell

(2013). General health issues in multiple sclerosis: Comorbidities, secondary conditions, and health behaviors. CONTINUUM: Lifelong Learning in Neurology, 19(4 Multiple Sclerosis), 1046-1057. doi:10.1212/01.CON.0000433284.07844.6b

43.

McNalley

T. E.

Yorkston

K. M.

Jensen

M. P.

Truitt

A. R.

Schomer

K. G.

Baylor

Molton

I. R.

(2015). A review of secondary health conditions in post-polio syndrome: Prevalence and effects of aging. American Journal of Physical Medicine & Rehabilitation, 94, 139-145. doi:10.1097/PHM.0000000000000166

44.

Molton

I. R.

Terrill

A. L.

Smith

A. E.

Yorkston

K. M.

Alschuler

K. N.

Ehde

D. M.

Jensen

M. P.

(2014). Modeling secondary health conditions in adults aging with physical disability. Journal of Aging and Health, 26, 335-359. doi:10.1177/0898264313516166

45.

Molton

I. R.

Yorkston

K. M.

(2017). Growing older with a physical disability: A special application of the successful aging paradigm. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 72, 290-299. doi:10.1093/geronb/gbw122

46.

National Association of State Budget Officers. (2017). State expenditure report. Retrieved from https://higherlogicdownload.s3.amazonaws.com/NASBO/9d2d2db1-c943-4f1b-b750-0fca152d64c2/UploadedImages/SER%20Archive/State_Expenditure_Report__Fiscal_2015-2017_-S.pdf

47.

National Committee for Quality Assurance. (2016). HEDIS® and quality compass. Retrieved from http://www.ncqa.org/hedis-quality-measurement/what-is-hedis

48.

National Council on Disability. (2013). Medicaid managed care for people with disabilities: Policy and implementation considerations for state and federal policymakers. Retrieved from https://ncd.gov/rawmedia_repository/20ca8222_42d6_45a5_9e85_6bd57788d726.pdf

49.

Ness

Kramer

(2013, August). Reducing hospital readmissions: It’s about improving patient care. Health Affairs Blog. Retrieved from http://www.healthaffairs.org/do/10.1377/hblog20130816.033808/full/

50.

Ouyang

Grosse

S. D.

Armour

B. S.

Waitzman

N. J.

(2007). Health care expenditures of children and adults with spina bifida in a privately insured U.S. population. Birth Defects Research Part A: Clinical and Molecular Teratology, 79, 552-558. doi:10.1002/bdra.20360

51.

Ouyang

Grosse

S. D.

Kenneson

(2008). Health care utilization and expenditures for children and young adults with muscular dystrophy in a privately insured population. Journal of Child Neurology, 23, 883-888. doi:10.1177/0883073808314962

52.

Pattanayak

C. W.

Rubin

D. B.

Zell

E. R.

(2011). [Propensity score methods for creating covariate balance in observational studies]. Revista Española de Cardiología, 64, 897-903. doi:10.1016/j.recesp.2011.06.008

53.

Pollack

H. A.

Wheeler

J. R.

Cowan

Freed

G. L.

(2007). The impact of managed care enrollment on emergency department use among children with special health care needs. Medical Care, 45, 139-145. doi:10.1097/01.mlr.0000250257.26093.f0

54.

Puleo

(2014). Coordinated care keeps frequent fliers from landing in ER. Retrieved from http://www.healthcarebusinesstech.com/coordinated-care-er/

55.

Saucier

Kasten

Burwell

Gold

(2012). The growth of Managed Long-Term Services and Supports (MLTSS) Programs: A 2012 update. Retrieved from https://www.medicaid.gov/medicaid-chip-program-information/by-topics/delivery-systems/downloads/mltssp_white_paper_combined.pdf

56.

Sawyer

S. M.

Macnee

(2010). Transition to adult health care for adolescents with spina bifida: Research issues. Developmental Disabilities Research Reviews, 16, 60-65. doi:10.1002/ddrr.98

57.

Schappert

S. M.

Rechtsteiner

E. A.

(2011). Ambulatory medical care utilization estimates for 2007. Washington, DC: U.S. Department of Health and Human Services. Retrieved from https://www.cdc.gov/nchs/data/series/sr_13/sr13_169.pdf

58.

Shern

D. L.

Jones

Chen

H. J.

Jordan

Ramoni-Perazzi

Boothroyd

R. A.

(2007). Medicaid managed care and the distribution of societal costs for persons with severe mental illness. American Journal of Psychiatry, 165, 245-260.

59.

Smith

V. K.

Gifford

Ellis

Edwards

Rudowitz

Hinton

. . . Valentine

(2016). Implementing coverage and payment initiatives: Results from a 50-state Medicaid budget survey for state fiscal years 2016 and 2017. Retrieved from http://files.kff.org/attachment/Report-Implementing-Coverage-and-Payment-Initiatives

60.

Sparer

(2012). Medicaid managed care: Costs, access, and quality of care. Retrieved from http://media.khi.org/news/documents/2013/01/14/managed-care-rwjf.pdf

61.

StataCorp. (2017). Stata statistical software: Release 15. College Station, TX: Author.

62.

State of Illinois. (2011). Governor Quinn signs landmark Medicaid reform legislation. Retrieved from https://www2.illinois.gov/pages/news-item.aspx?ReleaseID=9183

63.

Stuart

E. A.

Lee

B. K.

Leacy

F. P.

(2013). Prognostic score-based balance measures for propensity score methods in comparative effectiveness research. Journal of Clinical Epidemiology, 66(8), S84-S90.e1. doi:10.1016/j.jclinepi.2013.01.013

64.

Teoh

L. J.

Geelhoed

E. A.

Bayley

Leonard

Laing

N. G.

(2016). Health care utilization and costs for children and adults with duchenne muscular dystrophy. Muscle & Nerve, 53, 877-884. doi:10.1002/mus.24965

65.

Turk

M. A.

(2009). Health, mortality, and wellness issues in adults with cerebral palsy. Developmental Medicine & Child Neurology, 51, 24-29. doi:10.1111/j.1469-8749.2009.03429.x

66.

Verdier

Colby

Lipson

Simon

Stone

Bell

. . . Pérez

(2009). SoonerCare managed care history and performance: 1115 Waiver evaluation. Retrieved from https://www.mathematica-mpr.com/-/media/publications/pdfs/soonercare-summary.pdf

67.

Webb

T. S.

(2010). Optimizing health care for adults with spina bifida. Developmental Disabilities Research Reviews, 16, 76-81. doi:10.1002/ddrr.99

68.

Weinick

R. M.

Burns

R. M.

Mehrotra

(2010). Many emergency department visits could be managed at urgent care centers and retail clinics. Health Affairs, 29, 1630-1636. doi:10.1377/hlthaff.2009.0748

69.

Wing

Simon

Bello-Gomez

(2018). Designing difference in difference studies: Best practices for public health research. Annual Review of Public Health, 39, 453-469. doi:10.1146/annurev-publhealth-040617-013507

70.

Wong

V. C.

Wing

Steiner

P. M.

Wong

Cook

T. D.

(2013). Research designs for program evaluation. In Velicer

Schinka

(Eds.), Handbook of psychology: Research methods in psychology (2nd ed., pp. 316-341). Hoboken, NJ: John Wiley.

71.

Yamaki

Wing

Mitchell

Owen

Heller

(2018). Impact of Medicaid managed care on Illinois’s acute health services expenditures for adults with intellectual and developmental disabilities. Intellectual and Developmental Disabilities, 56, 133-146. doi:10.1352/1934-9556-56.2.133

72.

Yamaki

Wing

Mitchell

Owen

Heller

(2019). The impact of Medicaid managed care on health service utilization among adults with intellectual and developmental disabilities. Intellectual and Developmental Disabilities, 57, 289-306.

73.

Yorkston

K. M.

McMullan

K. A.

Molton

Jensen

M. P.

(2010). Pathways of change experienced by people aging with disability: A focus group study. Disability and Rehabilitation, 32, 1697-1704. doi:10.3109/09638281003678317

74.

Young

N. L.

Anselmo

L. A.

Burke

T. A.

McCormick

Mukherjee

(2014). Youth and young adults with spina bifida: Their utilization of physician and hospital services. Archives of Physical Medicine and Rehabilitation, 95, 466-471. doi:10.1016/j.apmr.2013.09.015

75.

Young

N. L.

Gilbert

T. K.

McCormick

Ayling-Campos

Boydell

Law

. . . Williams

J. I.

(2007). Youth and young adults with cerebral palsy: Their use of physician and hospital services. Archives of Physical Medicine and Rehabilitation, 88, 696-702. doi:10.1016/j.apmr.2007.03.005

76.

Young

N. L.

Steele

Fehlings

Jutai

Olmsted

Williams

J. I.

(2005). Use of health care among adults with chronic and complex physical disabilities of childhood. Disability and Rehabilitation, 27, 1455-1460.

77.

Zuckerman

Goin

(2012). How much will Medicaid physician fees for primary care rise in 2013? Evidence from a 2012 survey of Medicaid physician fees. Retrieved from http://www.kff.org/medicaid/issue-brief/how-much-will-medicaid-physician-fees-for/