## Abstract

Emergency room (ER) and hospital utilization among people with intellectual and developmental disabilities (IDD) are significant contributors to rising healthcare costs. This study identifies predictors of utilization among 597 adults with IDD. Using a retrospective survey of medical charts, descriptive statistics and logistic regressions were conducted. Individual-level risk factors for ER utilization included age, number of chronic health conditions, a diagnosis of cerebral palsy or neurological disorder, mental illness, and polypharmacy. Environmental predictors included community-based supported living. Hospitalization predictors included age and number of chronic illnesses. People residing in group homes were less likely to be admitted. This study found risk factors unique to individuals with IDD that should be addressed with tailored interventions as states transition to Medicaid managed care.

People with developmental disabilities (DD) experience lifelong cognitive and/or physical delays, with typical onset prior to age 22 (National Institutes of Health, 2010). While some developmental disabilities involve physical issues (e.g., cerebral palsy), many people with developmental disabilities experience concurrent intellectual disability. An intellectual disability (ID) is one type of developmental disability, characterized by significant limitations in cognitive functioning and decreased adaptive behavior in social, conceptual, and practical daily life skills (American Association on Intellectual and Developmental Disabilities [AAIDD], 2017). People with intellectual disability make up approximately 1% of the population worldwide (McKenzie, Milton, Smith, & Ouellette-Kuntz, 2016). Due to the high percentage of people diagnosed with both intellectual disability and developmental disabilities, the broad term of “intellectual and developmental disabilities”, or IDD, is commonly used to describe the population of people who receive publicly funded services and supports from local provider agencies (Braddock, Hemp, Tanis, Wu, & Haffer, 2017). For the purposes of this study, participants are described as people with IDD because of the strong relationship between DD and ID in the study sample (i.e., 99.5% of study participants were diagnosed with at least one developmental disability, with the majority (56%) having a co-occurring intellectual disability).

Given advances in technology and preventative health care, people with intellectual and developmental disabilities (IDD) are living longer (Heller, Stafford, Davis, Sedlezky, & Gaylord, 2010; Office of the Surgeon General, National Institute of Child Health and Human Development, Centers for Disease Control and Prevention, & U.S. Department of Health and Human Services, 2002). In addition, over the last 40 years, people with IDD in the United States (U.S.) have transitioned from living in institutions to living and working in more integrated community settings. Deinstitutionalization was hastened by the 1999 Supreme Court ruling, Olmstead v. L.C., which brought about rapid closures of congregate settings (ADA.gov, n.d.). Between 1997 and 2009, 36.2% of adults with disabilities moved from institutional settings to less restrictive, community settings like group homes and supported apartments (Healthy People 2010, 2010). The need for community-based housing options among people with IDD has rapidly increased and has compounded the demand for community-based healthcare and long-term services and supports (LTSS; Medicaid, 2013; Office for People with Developmental Disabilities [OPWDD], 2013).

People with IDD are often excluded from U.S. national health surveillance data and not always recognized as a disparate population (National Council on Disability, 2009). However, people with IDD have more complex medical, mental health, and behavioral health needs than those in the general population. These chronic health disparities require frequent medical care and monitoring (Havercamp, Scandlin, & Roth, 2004) and contribute to earlier mortality among people with IDD when compared to those without IDD (O'Leary, Cooper, & Hughes-McCormack, 2018). A study by Straetmans, van Schrojenstein Lantman-de Valk, Schellevis, & Dinant, (2007) compared people with intellectual disabilities to matched controls without IDD and found that those with IDD experienced a variety of health problems to a greater degree than the controls. These health disparities included higher prevalence rates of epilepsy, skin diseases, diabetes, and upper respiratory infections. People with cerebral palsy (CP) are also more likely to experience gastrointestinal issues, low bone density, fractures, decubiti, and chronic muscle tone abnormalities (Evenhuis, Henderson, Beange, Lennox, & Chicoine, 2000). Similarly, people with Down syndrome are at greater risk for chronic health conditions including congenital heart defects, epilepsy, gastrointestinal issues, and endocrine disorders. Early onset age-related disorders such as dementia and visual and hearing loss are also frequent among people with Down syndrome (Evenhuis et al., 2000).

Some chronic conditions experienced by people with IDD can be attributed to modifiable health behaviors and environmental risk factors (Haveman et al., 2010). Hsieh, Heller, Bershadasky, and Taub (2015) and Stancliffe et al. (2011) found that physical activity and obesity rates differed by living arrangement; people with IDD who lived in independent, community-based settings were more sedentary than those residing in institutional settings. High rates of obesity put people with IDD at greater risk for debility and obesity-related diseases such as coronary artery disease, hypertension, and diabetes (Evenhuis et al., 2000; Haveman et al., 2010). Furthermore, studies have indicated that 30-40% of people with IDD also have co-occurring psychiatric disabilities (Cooper, Smiley, Morrison, Williamson, & Allan, 2007; Einfeld, Ellis, & Emerson, 2011; Emerson, 2003; Morgan, Leonard, Bourke, Jablensky, 2008).

Long-term services and supports for people with IDD are largely funded through the Medicaid program. People with disabilities in the U.S. utilize more Medicaid dollars than any other population served. Since 2006, Medicaid expenditures have increased 2-3% annually (Kaiser Family Foundation, 2015) and reached $420 billion of all national health expenditures in 2012 (Centers for Medicare and Medicaid Services (CMS), 2011). Although people with disabilities comprised only 18% of Medicaid enrollees in 2010, they accounted for 44% of Medicaid's total expenditures (Centers for Medicare and Medicaid Services (CMS), 2010). The state of New York utilizes more Medicaid dollars than any other state in the nation (U.S. House of Representatives, 2013). A growing portion of this cost is due to avoidable emergency room visits and hospital admissions. One analysis of utilization among New York's Medicaid population found that 60% of ER visits and 16% of hospital admissions in 2011 were potentially avoidable. These avoidable utilization costs totaled$1.2 billion in expenditures (New York State Department of Health, n.d.).

New York and many other states are currently seeking transition of their IDD service systems to Medicaid managed care rather than remaining with the traditional fee-for-service payment structure. Although managed care has been widely used as a mechanism for controlling healthcare costs in programs for the general public, it has not been used extensively in systems serving people with IDD. This newer model of care utilizes needs-based funding, in which states determine a per member per month (PMPM) capitated rate to cover an individual's supports and services based on assessed risk levels (Engquist, Johnson, & Johnson, 2012). Some states have already demonstrated managed care cost savings through prevention and reduction of emergency room (ER) visits, hospital admissions, and hospital lengths of stay (The Lewin Group, 2004). Yamaki, Wing, Mitchell, Owen, and Heller (2017) analyzed the impact of Medicaid managed care on adults with IDD who received managed care (n = 1121) in the state of Illinois and those who did not (n = 1102). Preliminary findings demonstrated a significant decrease in utilization of the emergency room for those who received Medicaid managed care benefits, including care coordination, when compared to those who did not.

The Centers for Disease Control and Prevention (CDC), National Center on Birth Defects and Developmental Disabilities (NCBDDD) (2009), and Healthy People 2010 have all acknowledged the lack of information on the health status, health needs, and health disparities for people with IDD. There is great need for more health-related research on the IDD population (CDC & NCBDDD, 2009). This study addressed this gap by examining the health status and hospital utilization of adults with IDD.

## Specific Aims

People with IDD are consistently excluded from U.S. national health surveillance research (CDC & NCBDDD, 2009). As a result, little information exists on health services utilization in this population (World Health Organization [WHO], 2013). This study aimed to identify health trends and individual and environmental factors that predicted ER utilization and hospitalization for a sample of adults with IDD who resided in agency-supported settings throughout New York City, a high Medicaid recipient-count and cost region (New York State Department of Health, 2014). Specifically, the study examined: (a) the demographic characteristics of 597 people with IDD who resided in agency-supported settings throughout NYC; (b) associations between sample characteristics (e.g., age, mental health status, chronic disease, environmental factors) and ER and hospital utilization; and (c) individual and environmental factors that predicted ER and hospital utilization (for both medical and psychiatric reasons).

In an effort to understand the factors that lead people with IDD to utilize the ER/hospital and target those factors with interventions, this study aimed to contribute critical knowledge to the field. When predictors are known, they can better inform health models (for example, Accountable Care Organizations (ACOs) and Medicaid managed care models) for people with IDD. When programs provide interventions that target high-need, high-cost subgroups of people with IDD, cost savings can be achieved (The Lewin Group, 2004) and funds redirected to improving quality and access to a greater array of community-based LTSS options for people with IDD.

## Methods

### Participants

Participants were drawn from a group of 744 adults with IDD who received residential supports from a large nonprofit provider agency in New York City. The agency offers a wide range of supports and services to more than 20,000 people with IDD, learning disabilities, and physical disabilities across the lifespan and throughout the greater New York region. Services are provided to a diverse population of people with IDD from all five New York City boroughs, and three surrounding counties (Long Island, Suffolk/Nassau counties, and Westchester county). To meet criteria for this study, participants had diagnosed intellectual and developmental disabilities, were 21 years of age and above, and resided in some form of agency-supported housing for the entire 2011 calendar year. Based on these criteria, 597 residents were included and had their charts reviewed.

Thirty-eight agency nurses were trained to complete a structured survey of each resident's medical chart. The 597 individuals surveyed resided in one of 94 supported living arrangements including: (a) intermediate care facilities (ICFs)—congregate, institutional settings that house approximately 12-15 people with IDD who require 24-hour daily staff assistance; (b) individualized residential alternatives (IRAs)—a group home setting with 4-8 people with IDD receiving 24-hour daily staff support; or (c) supported apartments—a community-based setting of 1-4 adults with IDD who are typically more independent and receive a lower level of staff support (up to 20 hours/week).

### Procedures

A structured survey was designed by the research department of the agency with input from program administrators and the lead nurse. Four pages in length, the survey contained 22 items and multiple sub-items that covered individual and environmental characteristics. Many responses were multiple choice and/or frequency-based in nature. The following independent variables were collected using the survey and examined as potential predictors of ER utilization and hospitalization: age, sex, developmental disability (identified based on diagnostic information drawn from the Developmental Disabilities Profile-2 (OPWDD, n.d.), a New York State needs assessment completed by a team of healthcare professionals including a physician, pediatrician, psychiatrist and/or psychologist), level of intellectual disability, chronic health problems (identified by physician and/or specialist evaluations and documentation on the active treatment/monitoring of 20 different diagnosed health conditions), mental health diagnoses (determined by psychiatrist or psychologist evaluations and treatment notes for mental and/or behavioral health diagnoses), and polypharmacy (identified as the total number of prescription medications for medical/physical issues and psychotropic medications taken by mouth daily or PRN [as needed]). Five developmental disability types were provided on the survey—autism spectrum disorder, cerebral palsy, Down syndrome, neurological disorder, and other. Neurological disorder was identified based on a formal physician diagnosis contained within the medical chart (epilepsy was given as one example for nurses who completed the survey). More than one developmental disability could be indicated on the survey, if individuals had multiple diagnoses. A ‘chronic conditions' variable was created as a total count of the number of chronic health problems a person was being actively monitored/treated for. Prevalence of cardiovascular disease and diabetes were identified as separate independent variables, as they serve as Healthcare Effectiveness Data and Information Set (HEDIS) measures, which are regularly monitored as performance outcomes in Medicaid managed care models (National Committee for Quality Assurance [NCQA], 2017). Although no uniform definition of polypharmacy exists, the most commonly reported definition in health sciences literature is taking five or more medications daily (Masnoon, Shakib, Kalisch-Ellet, & Caughey, 2017). For this reason, polypharmacy was entered into the model as a dichotomous variable (0-4 medications vs. 5 or more medications).

Environmental characteristics were also collected and analyzed as independent variables including: supported living arrangement and region. Due to the racial, ethnic, and socioeconomic diversity of New York's geographic regions, six regions were included as environmental variables of interest. Although these regions are in close proximity of each other, socioeconomic status and access to health care services varies a great deal between them, with the Bronx being the lowest socioeconomic region, with a median household income of 34,299 USD and 30.3% of its population living in poverty, and Long Island being the wealthiest, with a median household income of 94,064 USD and only 6.95% of its population living in poverty (U.S. Census Bureau, 2016; United Hospital Fund, 2008). Region was included to serve as an indirect indicator of neighborhood deprivation and health services access. Cooper et al. (2011) found that adults with IDD were more likely to live in economically depressed regions. They also found that socioeconomic status and neighborhood deprivation were inversely related to emergency room utilization. This study's authors were interested in looking for similar trends within an urban U.S. population with IDD.

The study had four dependent variables: ER use for a medical/physical reason, hospitalization for a medical/physical reason, ER use for a behavioral/psychiatric reason, and hospitalization for a behavioral/psychiatric reason. Program nurses collected survey information on the number of ER visits and hospital admissions for medical and psychiatric reasons during the 2011 calendar year. Each ER and/or hospital admission was counted separately; for example, if an individual visited the ER and then was admitted to the hospital, one ER visit and one hospital admission were counted. Emergency room visits were classified based on the primary symptoms documented in the medical chart and on follow-up sub-questions on the survey (e.g., “The visit was medically necessary [e.g., critical injury/illness required immediate attention].” versus “Consumer needed to be evaluated, diagnosed, or stabilized”). ER visits and hospital admissions were defined as “being seen in an ER” and “being hospitalized” respectively, but the survey did not collect additional information on whom the person was seen by during these visits.

Prior to administration, the survey was pilot tested and minor revisions were made. In addition, program nurses attended an in-person training session at their monthly nurse's meeting and a follow-up training webinar on how to complete the survey. Program nurses reviewed hard copies of residential medical records to complete the survey. Upon completion of the survey, data were entered by two trained administrative assistants. A research associate reviewed the data and contacted program nurses when information was missing or unclear.

### Data Analysis

To explore the sample's distributions on the variables in this study, descriptive statistics are presented including frequencies and percentages for categorical predictors, along with means, standard deviations, skew, and kurtosis for continuous predictors. Bivariate correlations were explored to examine the relationships amongst the variables used in this study. To remediate potential collinearity problems, predicting variables exhibiting high correlations (.500 or above) with other variables were removed from the regression analyses. One exception, however, was that profound intellectual disability correlated at 0.522 with institutionalization, but both were retained as they conceptually represent separate individual and environmental factors.

Binomial logistic regression models were used to identify predictors of each ER utilization and hospitalization for either medical or behavioral/psychiatric reasons. These outcomes were respectively coded as: 0 for “No Hospital Admission” and 1 for “Hospital Admission”; and 0 for “No ER utilization” and 1 for “ER utilization”. In order to control for the effects of environmental factors when exploring individual characteristics, models were built hierarchically. Specifically, environmental factors (including type of living arrangement and the region in which participants resided) were entered in the first block, then individual characteristic (including age, sex, level of intellectual disability, type of developmental disability, number of chronic health conditions, mental health diagnosis, and polypharmacy) were added in a second block. The regions used in this study systematically differ from one another in terms of wealth and demographic composition. To account for variation due to clustering within region, regional indicators were entered as fixed-effects (Allison, 2009). The Queens region was used as the reference group as it had the largest share of the sample and a median household income that was closest to the middle. Regression coefficients (B) for each of the predicting variables were tested for statistical significance, and the corresponding odds ratios (OR [eB]) were used for interpreting effect size. Odds ratio values of 1.44, 2.48, and 4.27 correspond to Cohen's d values of 0.2, 0.5, and 0.8 respectively (Borenstein, Hedges, Higgins, & Rothstein, 2009). Within social sciences literature, these have come to be conventionally interpreted as representing small, medium, and large effects on the suggestion of Cohen (1988). However, it should be noted that Cohen's d is a treatment effect size and may not be appropriate when considering the prediction of odds from a continuous predictor. There were five missing responses to items on the survey. Listwise deletion was used when running the models; hence, these cases were not included in the analyses.

## Results

Table 1

Sample Demographics (n = 597)

Table 2

Bivariate Zero-Order Correlations Amongst Factors and Outcomes

Table 2

Extended

Logistic regressions were conducted to explore individual and environmental characteristics as independent predictors of: (a) ER visits for medical reasons; (b) hospital admission for medical reasons; (c) ER visits for behavioral/psychiatric reasons; and (d) hospital admission for behavioral/psychiatric reasons. Tables 3 and 4 highlight significant predictors of medical ER visits and hospital admissions (Table 3) and behavioral/psychiatric ER visits and hospital admissions (Table 4).

Table 3

Predictors of ER Use and Hospitalization for a Medical Reason

Table 4

Predictors of ER Use and Hospitalization for a Behavioral/Psychiatric Reason

A number of individual and environmental-level factors were found to be predictive of ER visits for medical reasons. Both region of residence and type of living arrangement had a significant impact on ER utilization for medical reasons. In the second poorest region, Brooklyn, residents appeared to have half the odds of visiting the ER for a medical issue (B = −.672, OR = .51, p = .03). This may be due to the fact that residents here had a lower average number of chronic conditions than those in the middle-income region (Queens), which served as the reference group. Once individual characteristics were entered, this effect fell to trend (B=−0.554, OR = 0.557, p = 0.100). Additionally, when individual characteristics were entered, the poorest region, the Bronx, exhibited a trend implying that residents in that region had nearly twice the odds of visiting the ER for medical reasons than their counterparts in a more middle-income region (Queens) (B = 0.642, OR = 1.901, p = 0.076). Residing in a supported apartment was also associated with increased odds of ER visits for a medical issue. When considering environmental factors alone, analysis shows that people residing in supported living had about 2.6 times the odds for visiting an ER than those who resided in a group home (B = .951, OR = 2.59, p = .03). Upon entering individual-level characteristics, this effect increased to nearly three times the odds (B = 1.046, OR = 2.85, p = .03), implying a possible suppression effect resulting from an association between residence type and individual-level characteristics.

The following individual-level characteristics were found to be predictive of ER utilization: each additional year of age (B = .019, OR = 1.02, p = .02) led to a higher odds of utilizing the ER; having cerebral palsy (B = .810, OR = 2.25, p = .01) or a neurological condition (B = .440, OR = 1.55, p = .05) (relative to those having no diagnosed developmental disability) had higher odds of ER utilization; ER visit odds increased with each additional chronic condition (B = .121, OR = 1.13, p = .04); those with a mental illness (B = .562, OR = 1.75, p = .01) had higher odds of ER utilization than those without a mental illness; and people who used five or more medications (polypharmacy; B = .787, OR = 2.20, p = .01) had higher odds for ER utilization than those who took 0-4 medications.

As age increased, the odds of visiting the ER for a medical reason increased by 2% for each additional year (B = .019, OR = 1.02, p = .02). Those with cerebral palsy (B = .810, OR = 2.25, p = .01) had over twice the odds of visiting the ER, while those with neurological disorders (B = .440, OR = 1.55, p = .05) had over one and a half higher odds of visiting the ER than those without a developmental disability.

Chronic illness and polypharmacy were significant predictors of ER visits for medical reasons, with a 13% increase in odds for each additional chronic illness (B = .121, OR = 1.13, p = .04) and those on 5 or more medications had over twice the odds of visiting the ER for medical reasons (B = .787, OR = 2.20, p = .01) than those who took 0 to 4 medications. Having a mental illness was also significantly associated with going to the ER for medical issues, with people diagnosed with a mental illness having 1.75 times higher odds (B = .562, OR = 1.75, p = .01) than those without a mental health diagnosis.

The results from the regression models examining predictors of hospitalization for a medical reason are also detailed in Table 3. When considering environmental risk factors alone, living in an institutional setting was the only significant predictor of hospitalization for a medical reason (B = .497, OR = 1.64, p = .04). When individual-level factors were added, the prediction fell to non-significance. It appears that the characteristics of individuals are explaining the association between institutional residence and medical hospitalizations. Namely, as mentioned before, having profound intellectual disabilities has a strong, positive relationship with institutionalization (r = 0.522).

Among individual-level predictors for medical hospitalization, age and number of chronic conditions were the only significant factors. The odds of medical hospitalization increased a factor of 4% for each additional year in age (B = .037, OR = 1.04, p = .001). For each additional chronic health issue, the odds of medical hospitalization increased by about 1.2 times (B = .176, OR = 1.19, p = .02).

Details pertaining to ER visits and hospitalization for behavioral/psychiatric reasons are presented in Table 4. There was an effect indicating that individuals residing in a wealthier region (Westchester) had approximately 75% times lower odds of visiting the ER for behavioral/psychiatric reasons than individuals living in a more middle-income region (Queens) (B = −1.353, OR = .26, p = .04). Once individual-level characteristics are entered, however, this effect falls to non-significance. Among individual level characteristics, those with mental illness diagnoses had nearly 19 times higher odds of visiting the ER for behavioral/psychiatric reasons (B = 2.931, OR = 18.75, p = .004) than those without a mental health diagnosis. No other factors were significant predictors of ER visits for behavioral/psychiatric reasons. In this study, no significant predictors of hospitalization for behavioral/psychiatric reasons were found.

## Discussion

Limited health surveillance and utilization data exist for people with intellectual and developmental disabilities in the U.S. This study was one of the first to examine predictors of ER and hospital utilization for medical and behavioral/psychiatric reasons. During the study period, 38% of the sample visited the ER for a medical issue, which was eight percentage points higher than the utilization rate (30%) reported in the Janicki et al. (2002) study of adults with IDD residing in upstate New York group homes. This ER utilization rate was also higher than the 2012 published rate for people in the general U.S. population (20%) (Gindi, Cohen & Kirzinger, 2012). Although additional research is needed, this finding suggests that people with IDD residing in the NYC area may have a unique set of characteristics that place them at greater risk for ER utilization.

Further, 15% and 3% of this study's sample were admitted to the hospital for medical and behavioral/psychiatric issues, respectively. Hospital admission for medical reasons was fairly consistent with Janicki et al.'s (2002) sample (16%) but lower than the general public's percentage (27%) (Gindi et al., 2012). Although speculative, this low rate of admission may represent the level of preventative care that residents receive by provider agencies who often have nurses assigned to group homes. These nurses are charged with ensuring residents' medical needs are attended to. The low rate of behavioral/psychiatric admissions may be due to misclassification error or unique features of the sample used in this study (e.g., a low incidence of hospitalization for behavioral/psychiatric reasons) and does not imply that this would be the case in a larger population.

Individual-level predictors of ER/hospital utilization were consistent between this sample, other samples of adults with IDD, the elderly U.S. population (e.g., polypharmacy) and general U.S. population (e.g., age, chronic illness) (Joynt, Gawande, Orav, & Jha, 2013; Lin et al., 2006; Wong, Marr, Kwan, Meiyappan, & Adcock, 2014). Individual-level health factors are often of focus in utilization studies (Hosking et al., 2017; Thomas et al., 2011). However, the impact of socioeconomic and environmental predictors on ER/hospital utilization rates have rarely been studied (Lin et al., 2006). Therefore, the environmental predictors of living arrangement and geographic region were of great interest in this study. When individual-level factors were controlled for, the risk of medical ER utilization intensified for supported apartment residents, placing them at three times greater odds for visiting the ER than those residing in group homes. Lunsky, Balogh and Cairney (2012) found a similar, yet non-significant, trend in which people with ID living in the community demonstrated a greater risk for visiting the ER (for psychiatric reasons) than group home residents.

Increased risk for ER utilization among adults with IDD who live in the community may be attributed to service access issues, low utilization of primary care and/or preventative wellness appointments, lack of understanding of IDD-specific needs among primary care physicians (PCP), less staff support in encouraging healthy eating, physical activity and regular health checks, and decreased or unreliable access to transportation (Friedman & Rizzolo, 2016; Hsieh et al., 2015; U.S. Public Health Service, 2005), all of which merit further investigation. People who live more independently or with family typically lack necessary access to medical and/or mental health services and receive “light touch” supports (Lunsky, Tint, Robinson, & Khodaverdian, 2011; Weiss, 2009). Improved access to LTSS for those living in supported apartments, independently, or with family is imperative to improving health-related quality of life and decreasing hospital utilization in this population.

Geographic region was also identified as a predictor of ER utilization in this study. The present study used geographic region as an indirect indicator of neighborhood deprivation and health services access. People who resided in the wealthier region of Westchester utilized the ER for psychiatric visits to a lesser degree than people in the middle-income region of Queens. Although only marginally significant, living in the Bronx, the lowest socioeconomic region, was a potential risk factor for both medical and behavioral/psychiatric ER visits. This is consistent with outcomes from other utilization studies of people with disabilities (Cooper et al., 2011; Rasch, Gulley, & Chan, 2013) and the broader New York City population (Robert Wood Johnson Foundation, 2013). Cooper et al.'s (2011) study of adults with IDD found that people with greater neighborhood deprivation and less access to ‘goods and services, resources and amenities, and of a physical environment which are customary in society' (p. 315) utilized less outpatient care and more emergency services. New York City health rankings within the general population have consistently found a greater number of Bronx residents with poor self-rated health, no insurance, decreased access to primary care, a higher prevalence of preventable hospital stays, and higher rates of psychiatric ER visits than any other region in New York City. (Robert Wood Johnson Foundation, 2013; Office of Mental Health [OMH], 2014).

### Clinical Implications

Programmatic and policy-level interventions have been piloted to improve quality outcomes and reduce emergency-related cost among people with IDD. These include patient-centered medical homes (PCMHs), Accountable Care Organizations (ACOs) and the IDD-specific behavioral health intervention, Systemic Therapeutic Assessment, Respite and Treatment (START; Center for START Services, 2019). U.S. provider agencies have also undertaken a variety of organizational level interventions including 24-hour nursing hotlines, telepsychiatry and telehealth systems that enable people with IDD and their staff to contact nurses/physicians after hours and receive support in deciphering between true emergencies and preventable or risk-aversive ER visits. Although some of these initiatives are targeted towards the general public or other patient populations, attention must be paid to their applicability for people with IDD.

Increased access to PCP and specialty providers, and enhanced training for these professionals in preventative care for people with IDD, have also proven effective in reducing ER utilization and hospitalization. Interventions such as salary and loan repayment incentives have proven effective in increasing access to PCPs and specialty outpatient providers in U.S. Health Professional Shortage Areas, like the Bronx (National Health Service Corps, 2014). In Canada, the Health Care Access Research and Developmental Disabilities (H-CARDD) Program has developed brief teaching videos involving actors with intellectual disabilities, a toolkit, and “About Me Passports” describing each person's needs in order to improve PCP and emergency room staff understanding of people with IDD and optimize the quality of care provided to them (Lunsky & Canso, n.d.). Training programs for physicians, such as the Leadership Education in Neurodevelopmental and Related Disabilities (LEND) program, prepare them to address the unique health needs of people with IDD (Ervin & Merrick, 2014).

Self-management programs for community-dwelling adults with IDD have also been piloted to improve health outcomes and decrease ER/hospital utilization rates. Bazzano et al. (2009) implemented a 7-month Healthy Lifestyle Change Program (HLCP) tailored to the needs of people with IDD. The program included peer mentorship and biweekly education aimed at improving knowledge, attitudes, behavior, diet and exercise. HLCP resulted in weight loss for 75% of its participants, a number of positive lifestyle changes, an increase in community capacity and overall improved quality of life for participants from pre- to post-intervention.

### Research Implications

Primary limitations of this study include its retrospective design and use of secondary survey data. Because the study utilized secondary data, important social/environmental factors and confounders were not accounted for in this analysis, as they were not collected using the survey tool (e.g., additional health conditions and chronic diseases, race and ethnicity data, socioeconomic status, level of family involvement, usual source of primary care, Emergency Severity Index (ESI) or triage scores, history of ER visits and hospitalizations, information on whom the individual was seen by in the ER). Human error is inherent in a survey based on medical record reviews. Misclassification and underreporting of behavioral/psychiatric emergencies may have also occurred. Further, ER visits and hospitalizations were modeled as being linearly related to age. Future research could benefit from exploring potential non-linear relations between age and these outcomes. In this study, the potential for interaction effects amongst the predictor variables was not addressed. Additional studies exploring potential interactions would serve as important follow-ups to the research presented here. Finally, the low incidence rate of hospitalizations due to behavioral/psychiatric reasons made it difficult to generate conclusions about predictors of this outcome.

Future utilization studies should more clearly define outcomes to aid in delineation between medical/physical and behavioral/psychiatric events and further explore specific characteristics unique to behavioral/psychiatric utilization. The impact of social determinants of health and environmental factors, specifically socioeconomic status and living arrangement (including independent living or living with family), warrants further investigation.

This study identifies utilization predictors for an urban sample of people with IDD. With a new administration, the U.S. is embarking on a critical period of Medicaid reform for people with IDD, their families, service providers and state health systems. The U.S. is challenged to not only contain healthcare costs, but also provide person-centered services aimed at closing health disparity gaps for people with IDD. As New York and other states across the nation transition to a Medicaid managed care model, the unique qualities of people with IDD must be considered when setting risk-adjusted capitation payments and determining per member per month rates. Further investigation is needed to determine the impact of new policy and practice-level interventions on utilization and health-related quality of life for one of its most underserved populations—people with intellectual and developmental disabilities.

## References

References
. (
n.d.
).
. .
Allison,
P. D.
(
2009
).
Fixed effects regression models (Vol. 160)
.
SAGE publications
.
American Association on Intellectual and Developmental Disabilities (AAIDD)
. (
2017
).
Definition of intellectual disability
.
Bazzano,
A. T.,
Zeldin,
A. S.,
Diab,
I. R.,
Garro,
N. M.,
Allevato,
N. A.,
&
Lehrer,
D.
(
2009
).
The healthy lifestyle change program: A pilot of a community-based health promotion intervention for adults with developmental disabilities
.
American Journal of Preventive Medicine
,
37
(
6 Suppl 1
),
S201
208
. doi:
Borenstein,
M.,
Hedges,
L. V.,
Higgins,
J. P.,
&
Rothstein,
H. R.
(
2009
).
Introduction to meta-analysis
.
John Wiley & Sons
.
D. L.,
Hemp,
R. E.,
Tanis,
E. S.,
Wu,
J.
&
Haffer,
L.
(
2017
).
The state of the states in intellectual and developmental disabilities: 2017
.
Boulder, CO
:
University of Colorado, Coleman Institute for Cognitive Disabilities, Department of Psychiatry
.
Center for START Services
. (
2019
).
. .
Centers for Disease Control and Prevention (CDC)
. (
2009
).
U.S. surveillance of health of people with intellectual disabilities: A white paper
.
CDC/NBCDDD Health Surveillance Work Group
. .
Centers for Medicaid & Medicare Services (CMS)
. (
2010
).
2010 actuarial report on the financial outlook for Medicaid
.
Centers for Medicaid & Medicare Services (CMS)
. (
2011
).
National health expenditure (NHE) fact sheet
.
Cohen,
J.
(
1988
).
Statistical power analysis for the behavioral sciences
.
Hillsdale, NJ
:
Lawrence Erlbaum Associates, 2nd edition
.
Cooper,
S.,
McConnachie,
A.,
Allan,
L. M.,
Melville,
C.,
Smiley,
E.,
&
Morrison,
J.
(
2011
).
Neighbourhood deprivation, health inequalities and service access by adults with intellectual disabilities: A cross sectional study
.
Journal of Intellectual Disability Research
,
55
,
313
323
. doi:
Cooper,
S.,
Smiley,
E.,
Morrison,
J.,
Williamson,
A.,
&
Allan,
L.
(
2007
).
Mental ill-health in adults with intellectual disabilities: prevalence and associated factors
.
The British Journal of Psychiatry
,
190
,
27
35
. doi:
Dunn,
K.,
Hughes-McCormack,
L.,
&
Cooper,
S. A.
(
2018
).
Hospital admissions for physical health conditions for people with intellectual disabilities: Systematic review
.
Journal of Applied Research in Intellectual Disabilities
,
31
(
1
),
1
10
. doi:
Einfeld,
S. L.,
Ellis,
L. A.,
&
Emerson,
E.
(
2011
).
Comorbidity of intellectual disability and mental disorder in children and adolescents: A systematic review
.
Journal of Intellectual & Developmental Disability
,
36
,
137
143
. doi:
Emerson,
E.
(
2003
).
Prevalence of psychiatric disorders in children and adolescents with and without intellectual disability
.
Journal of Intellectual Disability Research
,
47
,
51
58
.
Engquist,
G.,
Johnson,
C.,
&
Johnson,
W. C.
(
2012
).
Systems of care for individuals with intellectual and developmental disabilities: A survey of states
.
Center for Health Care Strategies Policy Brief
.
Ervin,
D. A.,
&
Merrick,
J.
(
2014
).
Intellectual and developmental disability: Healthcare financing. Frontiers in Public Health, 2, 160
.
Evenhuis,
H.,
Henderson,
C. M.,
Beange,
H.,
Lennox,
N.,
&
Chicoine,
B.
(
2000
).
Healthy ageing - Adults with intellectual disabilities: Physical health issues
.
Geneva, Switzerland
:
World Health Organization
.
Friedman,
C.,
&
Rizzolo,
M. C.
(
2016
).
The state of transportation for people with intellectual and developmental disabilities in Medicaid Home and Community-Based Services 1915(c) waivers
.
Journal of Disability Policy Studies
,
27
,
168
177
.
Gindi,
R. M.,
Cohen,
R. A.,
&
Kirzinger,
W. K.
(
2012
).
Emergency room use among adults aged 18-64: Early release of estimates from the National Health Interview Survey, January-June 2011
.
Hyattsville, MD
:
National Center for Health Statistics
.
Haveman,
M.,
Heller,
T.,
Lee,
L.,
M.,
Shooshtari,
S.,
&
Strydom,
A.
(
2010
).
Major health risks in aging persons with intellectual disabilities: An overview of recent studies
.
Journal of Policy and Practice in Intellectual Disabilities
,
7
,
59
69
.
Havercamp,
S. M.,
Scandlin,
D.,
&
Roth,
M.
(
2004
).
Health disparities among adults with developmental disabilities, adults with other disabilities, and adults not reporting disability in North Carolina
.
Public Health Reports
,
119
,
418
426
. doi:
Healthy People 2010
. (
2010
).
Disability and secondary conditions
.
Heller,
T.,
Stafford,
P.,
Davis,
L. A.,
Sedlezky,
L.,
&
Gaylord,
V.
(
Eds.
). (Winter
2010)
.
Impact: Feature Issue on Aging and People with Intellectual and Developmental Disabilities
,
23
(
1
).
[Minneapolis: University of Minnesota, Institute on Community Integration]
.
Hosking,
F. J.,
Carey,
I. M.,
DeWilde,
S.,
Harris,
T.,
Beighton,
C.,
&
Cook,
D. G.
(
2017
).
.
Annals of Family Medicine
,
15
,
462
470
. doi:
Hsieh,
K.,
Heller,
T.,
J.,
&
Taub,
S.
(
2015
).
Impact of adulthood stage and social-environmental context on body mass index and physical activity of individuals with intellectual disability
.
Intellectual and Developmental Disabilities
,
53
,
100
113
. doi:
Janicki,
M. P.,
Davidson,
P. W,
Henderson,
C. M.,
McCallion,
P.,
Taets,
J. D.,
Force,
L. T.,
Sulkes,
S. B.,
Frangenberg,
E.,
&
P. M.
(
2002
).
Health characteristics and health services utilization in older adults with intellectual disability living in community residences
.
Journal of Intellectual Disability Research
,
46
,
287
298
. doi:
Joynt,
K. E.,
Gawande,
A. A.,
Orav,
E. J.,
&
Jha,
A. K.
(
2013
).
Contribution of preventable acute care spending to total spending for high-cost Medicare patients
.
JAMA
,
309
,
2572
2578
.
doi.:10.1001/jama.2013.7103
Kaiser Family Foundation
. (
2015
).
Medicaid enrollment and spending growth: FY 2015 & 2016
. .
The Lewin Group
. (
2004
).
Medicaid managed care cost savings – A synthesis of fourteen studies: Final report
.
Falls Church, VA
:
Prepared by the Lewin Group for America's Health Insurance Plans
.
Lin,
J.,
Yen,
C.,
Loh,
C.,
Hsu,
S.,
Huang,
H.,
Tang,
C.,
Li,
C. W.,
&
Wu,
J.
(
2006
).
A cross-sectional study of the characteristics and determinants of emergency care utilization among people with intellectual disabilities in Taiwan
.
Research in Developmental Disabilities
,
27
(
6
),
657
667
. doi:
Lunsky,
Y.,
&
Canso,
D.
(
n.d.
).
H-CARDD 101
:
What is the Health Care Access Research & Developmental Disabilities Program doing that could be relevant to your family?
Lunsky,
Y.,
Balogh,
R.,
&
Cairney,
J.
(
2012
).
Predictors of emergency department visits by persons with intellectual disability experiencing a psychiatric crisis
.
Psychiatric Services
,
63
,
287
290
. doi:
Lunsky,
Y.,
Tint,
A.,
Robinson,
S.,
Khodaverdian,
A.,
&
C.
(
2011
).
Emergency psychiatric service use by individuals with intellectual disabilities living with family
.
Journal of Mental Health Research in Intellectual Disabilities
,
4
,
172
185
.
Masnoon,
N.,
Shakib,
S.,
Kalisch-Ellet,
L.,
&
Caughey,
G.E.
(
2017
).
What is polypharmacy? A systematic review of definitions
.
BMC Geriatrics
,
17
,
230
.
McKenzie,
K.,
Milton,
M.,
Smith,
G.,
&
Ouellette-Kuntz,
H.
(
2016
).
Systematic review of the prevalence and incidence of intellectual disabilities: Current trends and issues
.
Current Developmental Disorders Reports
,
3
,
104
115
.
Medicaid
. (
2013
).
Long-term services & supports
.
Morgan,
V. A.,
Leonard,
H.,
Bourke,
J.,
&
Jablensky,
A.
(
2008
).
Intellectual disability co-occurring with schizophrenia and other psychiatric illness: Population-based study
.
The British Journal of Psychiatry
,
193
,
364
372
. doi:
National Committee on Quality Assurance (NCQA)
. (
2017
).
HEDIS 2018
. .
National Council on Disability
. (
2009
).
The current state of health care for people with disabilities
.
Washington, DC
. .
National Health Service Corps
. (
2014
).
Loan repayment programs
. .
National Institutes of Health (NIH)
. (
2010
).
Intellectual and developmental disabilities
. .
New York State Department of Health (DOH)
. (
n.d.
).
New York's pathway to achieving the triple aim: Better care, better health, lower costs: Reducing avoidable hospital use through delivery system reform
.
New York State Department of Health (DOH)
. (
2014
).
Medicaid enrollees and expenditures by county
. .
Office of Mental Health (OMH)
. (
2014
).
Characteristics of recipients –High utilization criteria (High use psychiatric ER visits)
.
Office for People with Developmental Disabilities (OPWDD)
. (
2013
).
Services & supports
.
Office for People with Developmental Disabilities (OPWDD)
. (
n.d.
).
Developmental disabilities profile-2
. .
Office of the Surgeon General, National Institute of Child Health and Human Development, Centers for Disease Control and Prevention, & U.S. Department of Health and Human Services
. (
2002
).
Closing the gap: A national blueprint to improve the health of persons with mental retardation: Report of the surgeon general's conference on health disparities and mental retardation
.
O'Leary,
L.,
Cooper,
S. A.,
&
Hughes-McCormack,
L.
(
2018
).
Early death and causes of death of people with intellectual disabilities: A systematic review
.
Journal of Applied Research in Intellectual Disabilities
,
31
(
3
),
325
342
. doi:
Rasch,
E. K.,
Gulley,
S. P.,
&
Chan,
L.
(
2013
).
Use of emergency departments among working age adults with disabilities: A problem of access and service needs
.
Health Services Research
,
48
,
1334
1358
.
Robert Wood Johnson Foundation (RWJF)
. (
2013
).
.
Stancliffe,
R. J.,
Lakin,
C.,
Larson,
S.,
Engler,
J.,
J.,
Taub,
S.,
Fortune,
J.,
&
Ticha,
R.
(
2011
).
Overweight and obesity among adults with intellectual disabilities who use intellectual disability/developmental disability services in 20 U.S. States
.
American Journal on Intellectual and Developmental Disabilities
,
116
,
401
418
. doi:
Straetmans,
J. M.,
van Schrojenstein Lantman-de Valk,
H. M.,
Schellevis,
F. G.,
&
Dinant,
G.J.
(
2007
).
Health problems of people with intellectual disabilities: The impact for general practice
.
The British Journal of General Practice
,
57
,
64
66
.
Thomas,
K.,
Bourke,
J.,
Girdler,
S.,
Bebbington,
A.,
Jacoby,
P.,
&
Leonard,
H.
(
2011
).
Variation over time in medical conditions and health service utilization of children with Down syndrome
.
The Journal of Pediatrics
,
158
(
2
),
194
201
. doi:
U.S. Census Bureau
. (
2016
).
Income and poverty in the United States: 2015
.
United Hospital Fund
. (
2008
).
Patterns of emergency department utilization in New York City, 2008
.
U.S. Public Health Service (USPHS)
. (
2005
).
The Surgeon General's call to action to improve the health and wellness of people with disabilities
.
Washington, DC
:
U.S. Department of Health and Human Services, Office of the Surgeon General
.
U.S. House of Representatives
. (
2013
).
Billions of Federal tax dollars wasted annually by New York's Medicaid program
.
Walsh,
K.,
Kastner,
T.,
&
Criscione,
T.
(
1997
).
Characteristics of hospitalizations for people with developmental disabilities: Utilization, costs, and impact of care coordination
.
American Journal on Mental Retardation
,
101
,
505
520
.
Wong,
J.,
Marr,
P.,
Kwan,
D.,
Meiyappan,
S.,
&
L.
(
2014
).
Identification of inappropriate medication use in elderly patients with frequent emergency department visits
.
,
147
,
248
256
. doi:
Weiss,
J. A.,
Lunsky,
Y.,
Gracey,
C.,
Canrinus,
M.,
&
Morris,
S.
(
2009
).
Emergency psychiatric services for individuals with intellectual disabilities: Caregivers' perspectives
.
Journal of Applied Research in Intellectual Disabilities
,
22
(
4
),
354
362
. doi:
World Health Organization (WHO)
. (
2013
).
Disability and health
. .
Yamaki,
K.,
Wing,
C.,
Mitchell,
D.,
Owen,
R.,
Heller,
T.
(
2017
).
The impact of Medicaid managed care on emergency department and primary car utilization among adults with intellectual and developmental disabilities
[PowerPoint slides]
.