Increased Likelihood of Psychiatric Readmission With Medicaid Expansion vs Legacy Coverage

, , , , ,
The American Journal of Managed Care, November 2021, Volume 27, Issue 11

Individuals who became eligible for Medicaid through Medicaid expansion have an increased likelihood of psychiatric readmission compared with their legacy-enrolled counterparts.

ABSTRACT

Objectives: To compare patterns of psychiatric hospitalization and readmission within 30 days for Medicaid expansion (expansion) vs previously insured (legacy) samples.

Study Design: Retrospective analysis using Medicaid behavioral health service claims.

Methods: We identified 24,044 individuals with hospitalizations in calendar years 2017 and 2018 within the network of a behavioral health managed care organization in Pennsylvania. Logistic regression was used to examine factors associated with readmission.

Results: Individuals covered under expansion (n = 7747) vs legacy (n = 16,297) were older and more likely to be male and European American, with higher rates of cooccurring mental health (MH) and substance use disorder (SUD) diagnoses, as well as lower rates of MH and SUD services in the 30 days prior and any prior MH hospitalization. A higher proportion of individuals with expansion vs legacy status were readmitted (11.3% vs 9.0%; P < .0001). Controlling for factors associated with readmission, regression showed an increased likelihood of readmission for expansion vs legacy status (adjusted odds ratio [AOR], 1.23; 95% CI, 1.12-1.35; P < .0001). Increased risk for readmission was also found across populations for male patients (AOR, 1.12; 95% CI, 1.02-1.22; P = .0124), those with prior MH hospitalizations (AOR, 1.65; 95% CI, 1.51-1.81; P < .0001) or other behavioral health services (AOR, 1.14; 95% CI, 1.03-1.26; P = .0142), those with longer hospitalization episodes (AOR, 1.01; 95% CI, 1.00-1.01; P < .0001), and those with cooccurring SUD (AOR, 1.58; 95% CI, 1.44-1.74; P < .0001).

Conclusions: Individuals with coverage through Medicaid expansion compared with legacy coverage have an increased risk of psychiatric readmission and may warrant targeted interventions that also address service utilization and cooccurring SUD.

Am J Manag Care. 2021;27(11):488-492. https://doi.org/10.37765/ajmc.2021.88776

_____

Takeaway Points

Six years after enactment of Medicaid expansion in Pennsylvania, little is known about psychiatric hospitalization and readmission under Medicaid expansion.

  • In 2017-2018, 24,044 individuals within the network of a behavioral health managed care organization were identified with a psychiatric hospitalization.
  • Individuals with hospitalization under expansion (n = 7747) had lower rates of prior hospitalization (P < .0001) and behavioral health services (P < .0001) and higher rates of cooccurring substance use disorder (SUD; P < .0001) compared with those with hospitalization under legacy coverage (n = 16,297).
  • Odds of readmission were 1.23 times higher for those with expansion vs legacy coverage (P < .0001).
  • Increased odds of readmission were found for male patients and those with prior behavioral health service use, longer hospitalizations, and/or SUD.

_____

Under the Patient Protection and Affordable Care Act (ACA),1 eligibility for insurance coverage through Medicaid, a federal and state program that subsidizes health care costs and facilitates services for individuals with limited income and resources, was expanded in some states beginning in 2014 to residents with incomes below 138% of the federal poverty level. A disproportionate number of the previously uninsured were expected to have chronic mental health (MH) and substance use disorders (SUDs)2 because these individuals may have had diminished earnings, could not afford commercial insurance, or had been denied coverage under plans based on preexisting conditions.3 Individuals with coverage under Medicaid expansion (expansion) compared with individuals previously insured (legacy) are less likely to be caregivers of young children and more likely to be male,4-7 older,4-5,7 and European American.4,5 In addition to providing needed benefits to individuals with MH conditions and SUDs,2 provisions from the ACA including Medicaid expansion have been associated with reduced racial and socioeconomic disparities in health care and improved health care quality.8-12

As evidenced by utilization data, access to physical4 and behavioral health services,6,7 which include MH and SUD services, has shown initial surges under expansion policy. Studies of the initial impact of the policy change consistently find larger increases in behavioral health service utilization rates in states with Medicaid expansion vs states without expansion.5,13,14 For example, greater increases in utilization rates of methadone,13,14 SUD outpatient services, and injectable naltrexone have been reported in Medicaid expansion states.13

A good continuum of care for psychiatric health disorders includes engagement in ambulatory, community-based care, which can address both mental illness and substance use through evidence-based practice without the high cost and restrictive settings of psychiatric inpatient services.15 Hospitalization is associated with increased rates of homelessness, suicide risk, poor quality of life, and hospital readmission.16,17 In the absence of linkage to community-based care, Medicaid expansion could be associated with high utilization of psychiatric hospitalization among individuals unfamiliar with behavioral health service options. Individuals funded through Medicaid programs have been shown to have higher readmissions compared with privately insured or uninsured individuals.18 A better understanding of utilization of community-based and acute levels of care in expansion vs legacy populations could help improve deployment of care management resources for these populations.

In Pennsylvania (the study state), Medicaid was expanded in 2015.19 Rates of uninsured have steadily decreased—9.6% in 2014 to 5.6% in 2016 and 5.5% in 2017 and 2018—due to increased Medicaid coverage.20 In Pennsylvania, behavioral health services for Medicaid-eligible adults in 41 of 67 counties have been facilitated through Community Care Behavioral Health Organization of the University of Pittsburgh Medical Center (UPMC) Insurance Services Division, a not-for-profit behavioral health managed care organization (BHMCO), since 1999. The BHMCO routinely monitors hospitalizations, rates of readmission, and timely follow-up to ambulatory care as part of its National Committee for Quality Assurance Healthcare Effectiveness Data and Information Set reporting.

The purpose of the current study was 2-fold: (1) to determine whether psychiatric readmission rates differed between individuals with expansion vs legacy status and (2) to examine factors associated with readmission under Medicaid expansion. It was hypothesized that individuals with coverage under Medicaid expansion with behavioral health disorders would utilize both inpatient and community-based care more often than individuals under legacy coverage and that those with community-based care would have lower inpatient readmission rates. These analyses were conducted to better understand utilization of acute care for newly enrolled individuals and to identify intervention targets for care management.

METHODS

Sample

The sample included Medicaid-enrolled individuals, aged 14 to 64 years, within the BHMCO network with MH hospitalization in calendar year (CY) 2017 or 2018. The first MH hospitalization was used for individuals with multiple hospitalizations in the study period and was considered the index hospitalization, with behavioral health service utilization examined prior to and 30 days following discharge. Evaluation activities were approved as quality improvement by the UPMC Quality Review Committee.

Study Design

A retrospective cohort design was utilized to examine the association of Medicaid eligibility status with readmission. Eligibility groups were defined from Medicaid administrative data provided by the Pennsylvania Department of Human Services (DHS). Individuals had Medicaid eligibility with status held constant for the 30 days prior and 30 days following hospitalization; however, Medicaid eligibility and eligibility group may have varied over the 3 years since Pennsylvania implemented the Medicaid expansion policy.

Measures

Sociodemographic characteristics. Sociodemographic variables included gender, race, ethnicity, and age, which were obtained from Medicaid administrative data provided by the Pennsylvania DHS.

Diagnoses. Diagnoses were obtained from the BHMCO’s paid service claims and categorized using Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition21 criteria. Each claim could include multiple diagnoses, and individuals could be included in multiple diagnostic categories.

Readmission. Individuals were identified as having a readmission if a paid claim for an MH hospitalization occurred within 30 days of discharge from the index hospitalization. The rate of readmission was chosen as the primary outcome variable.

Behavioral health services 30 days prior. Prior service utilization was obtained from the BHMCO’s paid service claims and was defined as at least 1 paid claim for an MH or SUD service not including inpatient MH or SUD service in the 30 days before the index hospitalization.

Prior MH hospitalization. Prior hospitalization was defined as at least 1 paid claim for an inpatient MH service at any time before the index hospitalization.

Statistical Analysis

Characteristics of individuals during the study period were compared using Wilcoxon tests for medians and Pearson χ2 tests. Logistic regression examined the association of Medicaid eligibility status and odds of 30-day readmission adjusted for factors found to be associated with readmission in univariate analyses. The Hosmer-Lemeshow test statistic was utilized as a goodness-of-fit test; a nonsignificant result provides evidence that the model fits the data appropriately, whereas significant likelihood ratio, score, and Wald tests indicate that the model with covariates is more appropriate than the null (intercept only) model.22,23 The Cox and Snell R2, Nagelkerke R2,24,25 and C statistic measure of association were assessed to determine the degree to which predicted probabilities agreed with outcomes. Results with P values less than .05 were considered significant. Analyses were performed using SAS 9.3 (SAS Institute).

RESULTS

We identified the first MH hospitalization during the study period for 24,044 individuals: 7747 expansion and 16,297 legacy status (Table 1). Expansion individuals with hospitalization were older compared with legacy individuals (median age, 35 vs 24 years, respectively; P < .0001) and were more likely to be male (57.7% vs 44.0%; P < .0001) and European American (79.1% vs 74.5%; P < .0001) but comparable in rates of non-Hispanic ethnicity (95.8% vs 95.7%). Expansion individuals had lower rates of behavioral health service 30 days prior (70.0% vs 76.7%; P < .0001) and prior MH hospitalization (42.2% vs 50.4%; P < .0001) and shorter lengths of stay for the index hospitalization (median, 6 vs 7 days; P < .0001) compared with legacy individuals. For the commonly reported diagnoses (> 10% of the population), expansion individuals, compared with legacy individuals, had significantly lower rates of anxiety (32.4% vs 39.3%; P < .0001), schizophrenia (17.2% vs 23.9%; P < .0001), conduct disorder (5.5% vs 21.0%; P < .0001), and attention-deficit/hyperactivity disorder (4.4% vs 18.6%; P < .0001); they had higher rates of major depressive disorder (64.0% vs 59.2%; P < .0001) and bipolar disorder (37.6% vs 33.8%; P < .0001). Rates of cooccurring MH and SUD diagnoses were higher in expansion vs legacy individuals (54.0% vs 30.2%; P < .0001).

Medicaid Eligibility and Readmission

The readmission rate for all individuals was 9.7%; the rate was higher in expansion vs legacy individuals (11.3% vs 9.0%; P < .0001). A higher proportion of those with readmission vs no readmission were male (53.3% vs 47.9%; P < .0001) and expansion status (37.5% vs 31.7%; P < .0001), and those with readmission had higher rates of cooccurring MH and SUD diagnoses (51.5% vs 36.4%; P < .0001). Individuals with a readmission vs no readmission had longer lengths of stay for the index hospitalization in the hospital (median [interquartile range], 8 [5-13] days vs 7 [4-10] days; P < .0001) and had higher rates of prior MH hospitalization (61.6% vs 46.2%; P < .0001) and behavioral health service 30 days prior (77.1% vs 74.3%; P = .0027). There were no differences between readmitted and nonreadmitted groups for age, race, or ethnicity (eAppendix Table [available at ajmc.com]).

After controlling for covariates, results from logistic regression in Table 2 show an increased likelihood of readmission for those with expansion status (adjusted odds ratio [AOR], 1.23; 95% CI, 1.12-1.35; P < .0001), as well as for those with prior MH hospitalization (AOR, 1.65; 95% CI, 1.51-1.81; P < .0001), behavioral health service 30 days prior (AOR, 1.14; 95% CI, 1.03-1.26; P = .0142), longer lengths of stay in the hospital (AOR, 1.01; 95% CI, 1.00-1.01; P < .0001); male patients (AOR, 1.12; 95% CI, 1.02-1.22; P = .0127); and patients with cooccurring MH and SUD diagnoses (AOR, 1.58; 95% CI, 1.44-1.74; P < .0001).

DISCUSSION

Our study found that individuals with expansion status had increased odds of MH readmission. It is important to note that this study focused on individuals with MH hospitalization to understand factors associated with inpatient hospitalization and readmission. In contrast to our hypotheses, fewer individuals in the expansion group utilized behavioral health services prior to the index hospitalization. Lower rates of prior MH hospitalizations in expansion vs legacy individuals could be due to less time in the BHMCO’s Medicaid history. As for ambulatory care immediately prior to a hospitalization, there may be several reasons for lower utilization of behavioral health services, including unmet health care need, poor continuum of care based on lack of knowledge of behavioral health service options, health care needs met through other services, or poor access to care for reasons other than insurance coverage. However, results from this study suggest that even 3 to 4 years beyond enactment of Medicaid expansion, newly insured individuals are not connected to appropriate community-based services. For example, individuals under Medicaid expansion had higher rates of cooccurring MH and SUD diagnoses, which may indicate higher illness complexity that would benefit from community-based intervention. Rates of engagement and retention in SUD services are commonly low, especially among Medicaid-enrolled individuals with psychiatric diagnoses,26 and this may have affected readmission.

The findings of the current study help to better understand the needs of individuals with Medicaid eligibility under expansion policy. Individuals eligible for coverage under Medicaid expansion policy showed less engagement in behavioral health services prior to hospitalization, yet differences between individuals with expansion vs legacy coverage highlight the need to engage newly insured individuals into behavioral health care. These results may help to better understand service need under other conditions of increased Medicaid coverage, such as times of economic downturns, which have been shown to be associated with increased rates of anxiety, depression, suicidality, and SUD.27 Similar to previously reported studies, the current study found that individuals with services under Medicaid expansion were more likely to be older, male, and European American,4-7 and these characteristics along with insurance history should be considered when generalizing results for service utilization for individuals with Medicaid coverage during times of rapid expansion. Results from the regression model suggest the need to work with SUD and acute service providers to improve quality and continuum of care for populations with high risk of inpatient utilization and readmission.

Utilization of behavioral health service 30 days prior to index hospitalization and prior MH hospitalization were associated with readmission in this study’s regression model. As with cooccurring MH and SUD diagnoses, utilization of ambulatory and acute care by some individuals prior to the index hospitalization may indicate severity of illness, which may have been associated with readmission. Study findings acknowledge the difficulty in being able to accurately identify health care need and other determinants that may affect an individual’s quality of health care beyond their insurance coverage. The BHMCO currently utilizes a care management intervention for individuals at higher risk for inpatient readmission, which has been found to be effective in populations with cooccurring MH and SUD diagnoses.28,29 Additional research is needed to understand how this care management intervention may be affected by Medicaid expansion.

Limitations

There are several limitations to consider in the interpretation of findings of the current study. The use of the 30-day readmission time frame to investigate readmission was selected to examine proximal outcomes. This time frame is limiting, and results may not be reflective of longer-term patterns of hospitalizations6 or generalize to populations outside of Pennsylvania, which has a robust continuum of MH and SUD services. Claims data do not reflect quality of care or accurate diagnosing. Increased odds of readmission, and thus higher utilization of MH hospitalization, could be due to regression to the mean because of lower rates of prior MH hospitalization in the expansion group. Improving insurance coverage may not improve access to best practices in clinical care. Claims data are limited, and so the authors were limited in factors and clinical information that could be investigated in the current model.

CONCLUSIONS

The current findings have implications for helping newly enrolled individuals with behavioral health conditions navigate the behavioral health system and point to the need to engage individuals into ambulatory, community-based MH and SUD services. Practice and policy may benefit from further study of the relationships among care access, an individual’s motivation to engage in care, and impact of interventions to improve service utilization for individuals at higher risk for hospitalization.

Acknowledgments

The authors wish to thank Amanda Maise, MSPH, of Community Care Behavioral Health Organization, Pittsburgh, PA, as well as staff from Advocates for Human Potential, Sunbury, MA, and the UPMC Center for High-Value Health Care, Pittsburgh, PA, for review of this manuscript.

Author Affiliations: Community Care Behavioral Health Organization, University of Pittsburgh Medical Center (UPMC) Insurance Services Division (SLH, IOK, ADH, DSW, MOH, JMS), Pittsburgh, PA.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (SLH, IOK, ADH, DSW); analysis and interpretation of data (SLH, IOK, ADH, MOH, JMS); drafting of the manuscript (SLH, ADH, JMS); critical revision of the manuscript for important intellectual content (SLH, ADH, DSW, MOH, JMS); statistical analysis (IOK); administrative, technical, or logistic support (SLH, ADH); and supervision (DSW, MOH).

Address Correspondence to: Shari L. Hutchison, MS, PMP, Community Care Behavioral Health Organization, UPMC Insurance Services Division, 339 Sixth Ave, Ste 1300, Pittsburgh, PA 15222. Email: hutchisons@ccbh.com.

REFERENCES

1. Patient Protection and Affordable Care Act, 42 USC § 18001 et seq (2010).

2. Mir MA, Mutter, R, Teich JL. State Participation in the Medicaid Expansion Provision of the Affordable Care Act: Implications for Uninsured Individuals With a Behavioral Health Condition. Substance Abuse and Mental Health Services Administration; 2015.

3. Beronio K, Glied S, Frank R. How the Affordable Care Act and Mental Health Parity and Addiction Equity Act greatly expand coverage of behavioral health care. J Behav Health Serv Res. 2014;41(4):410-428. doi:10.1007/s11414-014-9412-0

4. Fertig AR, Carlin CS, Ode S, Long SK. Evidence of pent-up demand for care after Medicaid expansion. Med Care Res Rev. 2018;75(4):516-524. doi:10.1177/1077558717697014

5. Hoopes M, Angier H, Gold R, et al. Utilization of community health centers in Medicaid expansion and nonexpansion states, 2013-2014. J Ambul Care Manage. 2016;39(4):290-298. doi:10.1097/JAC.0000000000000123

6. O’Malley JP, O’Keeffe-Rosetti M, Lowe RA, et al. Healthcare utilization rates after Oregon’s 2008 Medicaid expansion: within-and between-group differences over time among new, returning, and continuously insured enrollees. Med Care. 2016;54(11):984-991. doi:10.1097/MLR.0000000000000600

7. Springer R, Marino M, O’Malley JP, Lindner S, Huguet N, DeVoe JE. Oregon Medicaid expenditures after the 2014 Affordable Care Act Medicaid expansion: over-time differences among new, returning, and continuously insured enrollees. Med Care. 2018;56(5):394-402. doi:10.1097/MLR.0000000000000907

8. Andrews CM, Guerrero EG, Wooten NR, Lengnick-Hall R. The Medicaid expansion gap and racial and ethnic minorities with substance use disorders. Am J Public Health. 2015;105(suppl 3):S452-S454. doi:10.2105/AJPH.2015.302560

9. Cawley J, Soni A, Simon K. Third year of survey data shows continuing benefits of Medicaid expansions for low-income childless adults in the U.S. J Gen Intern Med. 2018;33(9):1495-1497. doi:10.1007/s11606-018-4537-0

10. Griffith K, Evans L, Bor J. The Affordable Care Act reduced socioeconomic disparities in health care access. Health Aff (Millwood). 2017;36(8):1503-1510. doi:10.1377/hlthaff.2017.0083

11. Mazurenko O, Balio CP, Agarwal R, Carroll AE, Menachemi N. The effects of Medicaid expansion under the ACA: a systematic review. Health Aff (Millwood). 2018;37(6):944-950. doi:10.1377/hlthaff.2017.1491

12. Obama B. United States health care reform: progress to date and next steps. JAMA. 2016;316(5):525-532. doi:10.1001/jama.2016.9797

13. Andrews CM, Pollack HA, Abraham AJ, et al. Medicaid coverage in substance use disorder treatment after the Affordable Care Act. J Subst Abuse Treat. 2019;102:1-7. doi:10.1016/j.jsat.2019.04.002

14. Mojtabai R, Mauro C, Wall MM, Barry CL, Olfson M. The Affordable Care Act and opioid agonist therapy for opioid use disorder. Psychiatr Serv. 2019;70(7):617-620. doi:10.1176/appi.ps.201900025

15. Hung P, Busch SH, Shih YW, McGregor AJ, Wang S. Changes in community mental health services availability and suicide mortality in the US: a retrospective study. BMC Psychiatry. 2020;20(1):188. doi:10.1186/s12888-020-02607-y

16. Herman DB, Conover S, Gorroochurn P, Hinterland K, Hoepner L, Susser ES. Randomized trial of critical time intervention to prevent homelessness after hospital discharge. Psychiatr Serv. 2011;62(7):713-719. doi:10.1176/ps.62.7.pss6207_0713

17. Loch AA. Discharged from a mental health admission ward: is it safe to go home? a review on negative outcomes of psychiatric hospitalization. Psychol Res Behav Manag. 2014;7:137-145. doi:10.2147/PRBM.S35061

18. Fingar KR, Barrett ML, Jiang HJ. A comparison of all cause 7-day and 30-day readmissions, 2014. Healthcare Cost and Utilization Project statistical brief No. 230. October 2017. Accessed April 1, 2020. https://hcup-us.ahrq.gov/reports/statbriefs/sb230-7-Day-Versus-30-Day-Readmissions.jsp

19. Governor Tom Wolf launches Medicaid expansion in Pennsylvania. News release. Commonwealth of Pennsylvania; February 9, 2015. Accessed February 26, 2020. https://www.governor.pa.gov/newsroom/medicaid-expansion-in-pennsylvania/

20. HIC-4. health insurance coverage status and type of coverage by state—all persons: 2008-2019. United States Census Bureau. July 25, 2018. Accessed September 28, 2021. https://www.census.gov/library/publications/2016/demo/p60-257.html

21. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Association; 2013.

22. Ambrosius WT, ed. Topics in Biostatistics. Humana Press; 2007.

23. Hosmer DW Jr, Lemeshow S, Sturdivant RX. Applied Logistic Regression. 3rd ed. Wiley and Sons; 2013.

24. Cox DR, Snell EJ. The Analysis of Binary Data. 2nd ed. Chapman and Hall; 1989.

25. Nagelkerke NJD. A note on a general definition of the coefficient of determination. Biometrika. 1991;78(3):691-692. doi:10.1093/biomet/78.3.691

26. Lind BK, McCarty D, Gu Y, Baker R, McConnell KJ. Predictors of substance use treatment initiation and engagement among adult and adolescent Medicaid recipients. Subst Abuse. 2019;40(3):285-291. doi:10.1080/08897077.2018.1550467

27. Frasquilho D, Matos MG, Salonna F, et al. Mental health outcomes in times of economic recession: a systematic literature review. BMC Public Health. 2016;16:115. doi:10.1186/s12889-016-2720-y

28. Hutchison SL, Flanagan JV, Karpov I, et al. Care management intervention to decrease psychiatric and substance use disorder readmissions in Medicaid-eligible adults. J Behav Health Serv Res. 2019;46(3):533-543. doi:10.1007/s11414-018-9614-y

29. Taylor C, Holsinger B, Flanagan JV, Ayers AM, Hutchison SL, Terhorst L. Effectiveness of a brief care management intervention for reducing psychiatric hospitalization readmissions. J Behav Health Serv Res. 2016;43(2):262-271. doi:10.1007/s11414-014-9400-4