Medicaid Policies and Practices in US State Prison Systems

Abstract

Medicaid is an important source of health care coverage for prison-involved populations. From 2011 to 2012, we surveyed state prison system (SPS) policies affecting Medicaid enrollment during incarceration and upon release; 42 of 50 SPSs participated.

Upon incarceration, Medicaid benefits were suspended in 9 (21.4%) SPSs and terminated in 28 (66.7%); 27 (64.3%) SPSs screened prisoners for potential Medicaid eligibility.

Although many states supported Medicaid enrollment upon release, several did not. We have considered implications for Medicaid expansion.

Prisoners have a heavy burden of disease.^1–5 For qualifying individuals (e.g., disabled, impoverished adults with dependents and—starting in 2014 in states that choose to expand their Medicaid eligibility as part of health care reform—impoverished adults without dependents), Medicaid can provide health care coverage before and after incarceration.⁶ Although not assessed nationally, a study of former Texas and Ohio prisoners found that only 8% of men and 21% of women enrolled in Medicaid 8 to 10 months after release, whereas 68% and 58%, respectively, were without any health care coverage.⁷ Medicaid can also provide health care coverage for eligible prisoners during incarceration, when care is delivered in an inpatient setting, separate from the prison system.⁸

Despite the role of Medicaid in financing health care for prisoners and former prisoners, the current landscape of policies and practices affecting state prisoners’ enrollment in Medicaid has not been broadly assessed. To address this gap, we conducted a survey to understand Medicaid policies and practices employed in state prison systems (SPSs).

METHODS

From December 2011 through August 2012 we surveyed SPS personnel identified by top administrators as most knowledgeable about Medicaid policies employed within their SPS. Survey domains included (1) Medicaid termination or suspension upon incarceration, (2) assistance reenrolling in Medicaid, (3) challenges reenrolling in Medicaid, and (4) screening previously nonenrolled prisoners for potential Medicaid eligibility. We have described survey item responses using frequencies and medians.

We compared the geographic region and population size of participating and nonparticipating SPSs by using statistical tests, and conducted a sensitivity analysis to examine the influence of nonparticipating SPSs on our results.

RESULTS

Respondents representing 42 SPSs participated, with no statistically significant differences between represented SPSs and nonparticipating SPSs by size of prison population or geographic region. Respondents’ median time employed within their SPS was 13 years (Table 1).

TABLE 1—

Characteristics of Surveyed Personnel Knowledgeable About Medicaid Policies and Practices Implemented in Their State Prison System: United States, 2011–2012

Variable	No. (%) or Median (IQR)
Respondents’ division
Medical and mental health	23 (54.8)
Reentry	4 (9.5)
Administration	13 (31.0)
Other or not reported	2 (4.8)
Respondent time employed in current position, y	5 (2.0–10.0)
Respondent time employed in current prison system, y	13 (6.0–19.0)
Respondent time employed in any prison system, y	14 (9.0–20.0)
US region of state prison system
Northeast (n = 9)	9 (100.0)
Midwest (n = 12)	9 (75.0)
South (n = 16)	14 (88.0)
West (n = 13)	10 (77.0)
Prisoners incarcerated in respondents’ prison systema	1 137 748 (81.0)

Note. IQR = interquartile range. The sample size was n = 42.

^aDerived from estimates from the Bureau of Justice Statistics, December 31, 2009; total = 1 405 622 used in denominator of percentage calculation includes inmates incarcerated in nonresponse states.

About two thirds of SPSs employed policies of termination, and 21% employed suspension. In more than two thirds of SPSs with either policy, assistance was available to facilitate Medicaid reenrollment postrelease despite an array of challenges. More than one third of SPSs assessed whether prisoners requiring community inpatient care during their incarceration might be eligible for Medicaid coverage (Table 2).

TABLE 2—

Medicaid-Related Policies and Practices Among State Prison Systems: United States, 2011–2012

		Sensitivity Analysis, All States (n = 50)
Variable	Respondents (n = 42), No. (%)	Low %	High %
Prison has written Medicaid policy
Yes	10 (23.8)	(10)/50 = 20.0	(10+8)/50 = 36.0
No	19 (45.2)	38.0	54.0
Other	3 (7.1)	6.0	22.0
Don’t know or missing	10 (23.8)	20.0	36.0
Upon imprisonment, Medicaid is
Terminated	28 (66.7)	56.0	72.0
Suspended	9 (21.4)	18.0	34.0
Other	1 (2.4)	2.0	18.0
Don’t know or missing	4 (9.5)	8.0	24.0
Prison personnel assist Medicaid restoration
States with termination (n = 28)a	18 (64.3)	50.0	72.2
States with suspension (n = 9)b	7 (77.8)	41.2	88.2
Most common challenges restoring Medicaidcd
Unspecified release date or timing	8 (19.0)	16.0	32.0
Paperwork requirements	5 (11.9)	10.0	26.0
State interagency coordination	5 (11.9)	10.0	26.0
Prisoner or family engagement	4 (9.5)	8.0	24.0
Prison resources	3 (7.1)	6.0	22.0
Population targeted for Medicaid eligibility prescreening assessmentd
Any screening	27 (64.3)	54.0	70.0
Pregnant women or mothers	10 (23.8)	20.0	36.0
Prior Medicaid enrollment	9 (21.4)	18.0	34.0
Prior supplemental security income	12 (28.6)	24.0	40.0
HIV	15 (35.7)	30.0	46.0
Chronic mental health condition	18 (42.9)	36.0	52.0
Other chronic health condition	15 (35.7)	30.0	46.0
Hospitalization during incarceration	14 (33.3)	28.0	44.0
Prescreening assessment to enroll in Medicaid
During incarceration	15 (35.7)	30.0	46.0
After incarceration	11 (26.2)	22.0	38.0
Never or don’t know	16 (38.1)	32.0	48.0

Note. High % assumes that the 8 nonresponders would have endorsed. Low % assumes that the 8 nonresponders would not have endorsed.

^aSensitivity analysis denominator n = 36.

^bSensitivity analysis denominator n = 17.

^cOn the basis of open-end responses.

^dMultiple responses possible.

DISCUSSION

In 2000, nearly all states had policies terminating Medicaid enrollment upon incarceration.⁹ By contrast, we found that more than 20% of surveyed SPSs had policies suspending rather than terminating Medicaid enrollment. Notably, resumption of benefits in suspension states was not automatic but rather subject to a similar array of challenges experienced reactivating Medicaid in termination states. Nevertheless, in most suspension states, resumption reportedly occurred within a month of release, suggesting that suspension promotes timely reactivation of Medicaid benefits.

About two thirds of SPSs with a policy of either termination or suspension provided prisoners some assistance (e.g., help with social service program applications) resuming Medicaid services, indicating that efforts to support continuity of Medicaid benefits are relatively common and not necessarily contingent on the explicit policy of termination or suspension. At the same time, the lack of assistance in the remaining SPSs should be addressed, as this constitutes a basic function of discharge planning.

Most SPSs had policies to identify—and help prepare Medicaid applications for—prisoners who were potentially eligible for Medicaid but not enrolled previously. In 15 SPSs, Medicaid applications were submitted so that benefits could be used during incarceration to pay for inpatient care received in the community. (If the application is submitted within 90 days of care and enrollment is successful, Medicaid payments can be applied retroactively.¹⁰) Although the proportion of prisoners who require inpatient, community care is likely modest, their health care costs may be relatively high. Accordingly, use of Medicaid for these patients may substantively lower SPS medical expenditures.

Several states plan to expand Medicaid eligibility in 2014 to adults at 138% of the federal poverty threshold, regardless of disability or dependents.^11,12 Expanded Medicaid eligibility could dramatically increase the number of released prisoners with access to routine health care. Expanded Medicaid eligibility could also increase financial incentives for SPSs to provide Medicaid enrollment assistance to prisoners requiring community, inpatient care during their incarceration.

Our survey has a few limitations. First, it does not account for possible heterogeneity in the implementation of policies within SPSs. Second, we were unable to verify participants’ responses; however, their responses may reflect actual practice rather than official state policies. Finally, our responses represent only 42 SPSs. Our sensitivity analysis, however, suggests that inclusion of the other 8 SPSs would not have substantively changed our findings except for 1 item, the proportion of suspension states with Medicaid restoration assistance.

Despite the availability of services supporting Medicaid enrollment in many SPSs, a substantial proportion of SPSs had none. The expansion of Medicaid in many states will provide greater opportunities and incentives to facilitate Medicaid enrollment for disadvantaged prison-involved populations. Future research should evaluate SPSs’ success in facilitating prisoners’ Medicaid enrollment, characteristics of successful programs, and financial implications of enrollment for SPSs and for the Medicaid program.