Committee for a Responsible Federal Budget
Social Security_SSDI_Job Interview

Targeting Early Intervention Based on Health Care Utilization of SSDI Beneficiaries by State, with Emphasis on Mental Disorders and Substance Abuse

by Joyce Manchester

Read the full paper.

I. Introduction

Until now, little was known about the specific health care services used by beneficiaries of the Social Security Disability Insurance (SSDI) program. Understanding how SSDI beneficiaries use health care services in the years after eligibility can inform our understanding of their continued inability to work and the kinds of challenges they face. Such knowledge can also inform early intervention efforts for conditions that develop over time and might benefit from professional health care attention prior to limiting a person’s ability to work. More than 10 million individuals receive monthly benefits from the SSDI program (Social Security Administration 2018). In recent years, about three-quarters of them also have received Medicare benefits because they have been eligible for SSDI for at least 24 months (Centers for Medicare and Medicaid Services 2011).

Examining the types of health care services used, and how use of those services differs across states, offers insights into how the health care system interacts with the needs of SSDI beneficiaries in different parts of the country. Different states report varying shares of younger, middle-age, and older workers on the SSDI program, and awareness of the health care services used by the SSDI population across states could help explain why (Manchester 2016).

This paper complements analysis of early intervention efforts that address the medical and mental health needs of workers before they become unable to work and decide to apply for SSDI benefits. Recent research suggests that about half of SSDI applicants have strong, stable work histories prior to application, but the other half have struggled to work for several years prior to application (Contreary et al. 2017). Both types of applicants might benefit from early intervention efforts and more timely access to health care. A limited number of beneficiaries on the program might be able to work with additional support as well (Autor and Duggan 2003; Black, Daniel, and Sanders 2002; von Wachter, Song, and Manchester 2011).

Earlier work from the SSDI Solutions Initiative that points to strengthening supports for workers with health challenges, encouraging community-based health and work centers, and providing health coverage for workers with disabling conditions is especially relevant here (Stapleton et al. 2016, Christian et al. 2016, Perriello 2016). Identifying the types of services used by beneficiaries on the SSDI program across states points to services that, if available to workers prior to SSDI application, might keep them in the labor force, either full-time or part-time or on a transitional basis (Fichtner and Seligman 2016, Kerksick et al. 2016). Increasing access to those same services for workers with disabilities already on the SSDI program might allow more SSDI beneficiaries to work in part-time or transitional jobs.

Moreover, many younger adults qualify for the program through mental health disorders, and findings here show that states provide very different amounts of outpatient, non-pharmaceutical services for people with mental health issues. Policymakers might consider increasing services in the areas of mental health or specifically increasing access to treatment for substance abuse, especially for younger people, either before they apply for SSDI or once they are on the program.

Using the full set of Medicare claims data for 2012, the most recent year with complete data available to me, I examine the number of outpatient, non-pharmaceutical services by state and by primary diagnosis for the 5.7 million SSDI beneficiaries under age 65 who qualified for Medicare and were enrolled in Part B, fee-for-service (FFS) benefits (Table 1). After 24 months on the SSDI program, beneficiaries are eligible for Medicare Part A (hospital services) and Part B (office and outpatient services). Part A is free to beneficiaries, but those who choose to enroll in Part B must pay the Part B premium.

This paper will focus on the number of claims for office, laboratory, and other outpatient non-pharmaceutical services covered by Part B in the FFS program. The data files do not contain information on the cost of services, and I do not include the number of hospital stays because I have no information on the length of stay or intensity of treatment. The focus is on 2012 because data related to substance abuse services were redacted in subsequent years in the Dartmouth Atlas Project data used here, and looking at substance abuse services for the SSDI population is of special interest in this paper. Moreover, I look at SSDI beneficiaries with Medicare Part B FFS benefits who were under age 65 in 2012 to avoid conflating the results with Part B services for the general population who were eligible for Medicare benefits at age 65. About 99.3 percent of the 5.7 million FFS enrollees in Medicare Part B under age 65 in 2012 were SSDI beneficiaries.

The number of fee-for-service Medicare claims for the SSDI population under age 65 by state suggests that the quantity of services used by diagnosis type does not always match up with the reason for eligibility. With help from the Social Security Administration (SSA), I matched 3-digit ICD-9 medical codes from the Medicare claims data to the categories used by SSA to identify the primary reason for eligibility of SSDI beneficiaries. For example, the single largest reason for SSDI eligibility among all SSDI beneficiaries in 2010 (2 years before our Medicare data for 2012) was mental disorders, accounting for more than 36 percent of primary diagnoses for eligibility. The second largest reason for eligibility was conditions of the musculoskeletal system and connective tissue at about 25 percent. Across the country, the quantity of Medicare Part B FFS services shows the largest share of services going to musculoskeletal conditions at 14 percent of office and outpatient services; 10 percent of office and outpatient services were for mental disorders, including substance abuse services.

Of course, the primary diagnosis associated with each non-pharmaceutical service may hide other conditions. For example, many visits for chronic back pain would be coded in the musculoskeletal conditions category even if opioids were prescribed during the visit. If use of opioid prescriptions led to opioid dependence or abuse, however, services to treat those issues would be coded as opioid services.

Drilling down to services relating to substance abuse and opioids using 5-digit ICD-9 codes offers a window into services for SSDI beneficiaries across the states in light of the on-going opioid epidemic. In 2017 across the U.S. population, the age-adjusted rate of drug overdose deaths from various types of opioids was 3.6 times the rate in 1999 (Hedegaard et al. 2018). Rates were highest for adults ages 25-34, 35-44, and 45-54. Recent research suggests that SSDI beneficiaries on Medicare may be at elevated risk of opioid overdose; they accounted for 24.5 percent of hospitalizations for opioid overdose in 2013 (Peters et al. 2018).

Since 1996, individuals have not been able to qualify for SSDI benefits on the basis of substance abuse alone or if drug and/or alcohol addiction or abuse is a “material factor” in causing the individual’s disability (20 Code of Federal Regulations, § 404.1535). However, if the claimant has conditions that would be work-limiting if the substance abuse stopped, SSDI benefits may be awarded. Moreover, because people with mental health disorders are more likely than people without mental health disorders to experience an alcohol or substance use disorder, the high proportion of SSDI beneficiaries who become eligible based on a mental health disorder suggests an elevated risk of substance abuse (Center for Behavioral Health Statistics and Quality 2015).

After a discussion of overall health care services used by state and primary diagnosis in Section II, I look at services used for mental disorders in Section III and services related to substance abuse in Section IV. The term “mental disorders” is used by the Social Security Administration to include both intellectual disabilities as well as mental health conditions. I am able to focus on services related to substance abuse and opioids for SSDI beneficiaries under age 65 and also for SSDI beneficiaries under age 50 to shine a brighter light on the younger age group most affected by the opioid epidemic. Section V concludes the paper with some discussion of policy responses.

Read the full paper.