Health Service Utilization by Adults Affected by the Justice System at a Large Federally Qualified Health Center Network

Abstract

This study examines health care service utilization of adults impacted by the justice system (n=3380) and general patient population (n=226,970) receiving care at one federally qualified health center (FQHC; n=62 service sites) in California between 2020-2022. A hybrid manual review of the electronic medical record was conducted. The top five departments used by adults impacted by the justice system, based on total number of encounters, were general medicine, case management, mental health services, specialty services, and substance use disorder services. Patients demonstrated vulnerabilities in education, employment, and housing and extensive use of clinical services. Adults impacted by the justice system can benefit from targeted and coordinated case management to support health care utilization. Findings may inform funding initiatives, staffing, and service delivery models to optimize care and support services for individuals who have experienced contact with the justice system.

The health of adults who have experienced incarceration or other contact with criminal legal systems is often compromised due to pervasive conditions such as hypertension, diabetes, obesity, or sexually transmitted infections.^1–3 In fact, about one-half of males incarcerated in state prisons report ever having a chronic condition (e.g., diabetes, hypertension, arthritis) as do 60% of females.² Adults in carceral settings may be more likely to have these conditions as compared to non-institutionalized adults.³ For example, the U.S. Centers for Disease Control and Prevention reports that HIV is approximately 3 times as prevalent among persons in carceral setting vs. the general population; similarly, hepatitis C infection is about 10 times higher among those who have been incarcerated vs. non-institutionalized persons.⁴ Another recent study finds that severe mental illness and depression are more prevalent among those who have been recently incarcerated and the authors identified disparities in management of chronic conditions using medications.⁵ Poor management of chronic conditions in community and correctional settings can adversely affect individuals’ health both during detention and in the community.

Importantly, most individuals detained in carceral settings will eventually return to the community and require stable health insurance and ties to the health care system to facilitate health treatment and monitoring. Incarceration has also been associated with early onset of aging-related conditions and disproportionate mortality as compared to persons without incarceration histories.⁶ In particular, the first 14 days post-release are associated with the greatest vulnerability: the leading causes of mortality are substance-use related overdose, cardiovascular disease, homicide, suicide, cancer, motor vehicle accidents and liver disease.⁷ A deeper understanding of how adults impacted by the justice system access and engage with community-based health services is needed since they often experience fragmented relationships with institutions (e.g., health insurance, health care providers, stable housing and employment, social support) that support health and well-being.^8,9

Safety-net non-profit health care clinics, including federally qualified health clinics (FQHC) which serve uninsured or low-income community members regardless of patients’ ability to pay, are an important component of health care services for disenfranchised patients in the United States.^10,11 Significantly, little is known about the characteristics, number, or types of services provided to adults impacted by the justice system receiving care at FQHC.

This descriptive retrospective study draws on FQHC administrative data to describe: 1) the characteristics of adults impacted by the justice system who obtained health care at a large federally qualified health care community clinic (one FQHC with 62 service sites) in a diverse southern California county; 2) the health care departments which provided services to adults impacted by the justice system and the general patient population between 2020-2022 and 3) patients’ median encounters across departments. Findings may inform policies to support adults impacted by the justice system as well as service planning efforts by policy makers and safety net providers serving this community.

Methods

Study setting: Region and clinic network description.

This study was conducted in San Diego, California, U.S. The region is home to ~3.3 million persons over 4,260 square miles and it is highly diverse: 33.4% of residents are Hispanic, 46.2% are White, 4.7% are Black, 11.9% are Asian American or Pacific Islander, 0.4% are American Indian or Alaska Native, and 3.5% are from Other racial and ethnic groups.¹²

Data were obtained from the Family Health Centers of San Diego (FHCSD) which is a Federally Qualified Health Care Center that is classified as a Patient-Centered Medical Home by the Joint Commission and the National Committee for Quality Assurance. FHCSD serves San Diego County with approximately 2,000 staff members.^12,13 At the time of the analysis, FHCSD had 62 service sites throughout the County, including 22 primary care clinics, 18 behavioral health service sites, eight dental clinics, four vision clinics, two mobile counseling centers, three mobile medical units, a pharmacy, physical therapy services, and an outpatient substance use treatment program; comprehensive sexual and reproductive health care services are also provided. Annually, the network serves >227,000 persons, most of whom are low-income; 29% of patients are medically uninsured.¹³

Electronic medical record chart review to identify patients impacted by the justice system.

Between February-March 2022, the clinic research staff undertook a systematic review of its electronic medical record (EMR) which included 226,590 adult patients served in 1.9 million unique encounters for the period spanning from January 1, 2020-February 16, 2022 across 62 clinical service sites. Screening for history of incarceration or other contacts with criminal legal systems (e.g., police, court) at the clinics was extremely limited at the time this study was conducted as it is not ascertained at registration nor systematically during clinical encounters. Thus, we implemented a hybrid automated/manual approach to identify patients with any direct personal contact with the criminal legal system. We developed a database query using a comprehensive keyword list matched against provider notes followed by a manual review of patient records. Keywords were chosen to identify records that mention justice system contacts (i.e., arrested, county jail, criminal justice system, diverted, homeless court, mental health court, drug court, incarcerat, jail, justice involved, justice-involved, parole, prison, probation, re entr, re-entr, sheriff), or local institutions such as prison or program names serving adults impacted by the justice system (i.e., Donovan, George Bailey, Las Colinas, Lighthouse, National Crossroads, MCC—Metropolitan Correction Center, Second Chance, Community Transitions Center). Levels of justice system contact varied as evidenced by the keywords, ranging from arrests and appearance in court to incarceration. We identified 6,289 patients by 15,161 notes in the keyword query as potentially having involvement in the criminal legal system. Some were eliminated immediately through language processing (e.g., mention of family members’ involvement, or MCC matching mention of “DMCCP” in the provider notes, which represents the clinics’ “Diabetes Management Care Coordination Program”); the remainder were reviewed manually). Similar approaches have been used to identify individuals directly or indirectly impacted by the justice system.¹⁴ Four research staff divided 4,635 patients and 14,797 provider notes evenly and performed the manual review over the course of 4-6 weeks through March 2022. Thus, in total, 3,380 adults ages 18+ were identified as having reported direct contact with the justice system and were served by the network between January 2, 2020 through February 16, 2022.

Variables.

All variables for this analysis are drawn from the EMR. Patients’ demographic data are self-reported on paper forms and data are entered by staff into the EMR (see Table 1). We included the following variables in the analysis and transformations are noted: age was aggregated into categories (18-26; 27-45; 46-54; 55-64; 65+); sex as assigned at birth (female; male); sexual orientation (heterosexual; homosexual/bisexual; unknown/other); primary language (English; Spanish; other/unknown); marital status was aggregated into categories (married/domestic partner; single/divorced/separated/widowed; unknown/other); employment status (employed; retired; student; unemployed; unknown); educational attainment (college/post-graduate; high school; less than high school; unknown); housing status (own/rent; doubling up; shelter/transitional housing; streets/canyon/migrant camp; other homeless; unknown); veteran status (non-veteran; unknown; veteran). Race and ethnicity are collected separately at patient intake. To create the “Ethnicity, Race” variable, we first differentiated between Latino/Hispanic and Non-Hispanic/Unknown individuals, and subsequently stratified responses based on racial identification. Specifically, individuals identifying as “Latino/Hispanic” of all races were combined, while “Non-Hispanic/Unknown” individuals were differentiated into subgroups based on their racial identification, including “Black”, “White”, and “Other/Unknown.”

Table 1.

Demographic Characteristics of Patients Receiving Care at a Large Federally Qualified Healthcare Network Stratified by General Population vs Justice System Impacted, Southern California, January 2, 2020 to February 16, 2022

Variables	FULL SAMPLE N = 226,9701	GENERAL PATIENT POPULATION N = 223,5901	ADULTS IMPACTED BY THE JUSTICE SYSTEM N = 3,3801	p-value2
Age Group				<0.001
18-26	38,502 (17%)	38,249 (17%)	253 (7.5%)
27-45	90,568 (40%)	88,862 (40%)	1,706 (50%)
46-54	35,375 (16%)	34,731 (16%)	644 (19%)
55-64	37,237 (16%)	36,613 (16%)	624 (18%)
65+	25,288 (11%)	25,135 (11%)	153 (4.5%)
Sex Assigned at Birth				<0.001
Female	123,578 (54%)	122,735 (55%)	843 (25%)
Male	103,392 (46%)	100,855 (45%)	2,537 (75%)
Sexual Orientation				<0.001
Heterosexual	161,657 (71%)	159,003 (71%)	2,654 (79%)
Homosexual/Bisexual	6,577 (2.9%)	6,429 (2.9%)	148 (4.4%)
Unknown/Other	58,736 (26%)	58,158 (26%)	578 (17%)
Ethnicity, Race				<0.001
Latino/Hispanic, all races	77,238 (34%)	76,463 (34%)	775 (23%)
Non-Hispanic/Unknown, Black	18,308 (8.1%)	17,756 (7.9%)	552 (16%)
Non-Hispanic/Unknown, White	87,159 (38%)	85,630 (38%)	1,529 (45%)
Non-Hispanic/Unknown, Other/Unknown	44,265 (20%)	43,741 (20%)	524 (16%)
Language				<0.001
English	145,124 (64%)	142,033 (64%)	3,091 (91%)
Spanish	69,222 (30%)	69,008 (31%)	214 (6.3%)
Other/Unknown	12,624 (5.6%)	12,549 (5.6%)	75 (2.2%)
Marital Status				<0.001
Married/Domestic Partner	80,260 (35%)	79,736 (36%)	524 (16%)
Single, Divorced/Separated, Widowed	123,617 (54%)	120,950 (54%)	2,667 (79%)
Other/Unknown	23,093 (10%)	22,904 (10%)	189 (5.6%)
Employment				<0.001
Employed	74,576 (33%)	74,111 (33%)	465 (14%)
Retired	9,993 (4.4%)	9,911 (4.4%)	82 (2.4%)
Student	15,894 (7.0%)	15,788 (7.1%)	106 (3.1%)
Unemployed	67,477 (30%)	65,311 (29%)	2,166 (64%)
Unknown	59,030 (26%)	58,469 (26%)	561 (17%)
Education				<0.001
College/Post-Graduate	51,764 (23%)	51,370 (23%)	394 (12%)
High School	96,848 (43%)	94,880 (42%)	1,968 (58%)
Less than High School	39,144 (17%)	38,435 (17%)	709 (21%)
Unknown	39,214 (17%)	38,905 (17%)	309 (9.1%)
Housing				<0.001
Own/Rent	177,428 (78%)	176,398 (79%)	1,030 (30%)
Doubling Up	14,199 (6.3%)	13,848 (6.2%)	351 (10%)
Shelter/Transitional Housing	10,739 (4.7%)	9,309 (4.2%)	1,430 (42%)
Streets/Canyon/Migrant Camp	3,391 (1.5%)	3,147 (1.4%)	244 (7.2%)
Other homeless	4,715 (2.1%)	4,508 (2.0%)	207 (6.1%)
Unknown	16,498 (7.3%)	16,380 (7.3%)	118 (3.5%)
Veteran Status				<0.001
Veteran	3,116 (1.4%)	3,000 (1.3%)	116 (3.4%)
Non-Veteran	222,580 (98%)	219,331 (98%)	3,249 (96%)
Unknown	1,274 (0.6%)	1,259 (0.6%)	15 (0.4%)

¹n (%)

²Pearson’s Chi-squared test

Health care service utilization was determined based on the categories of departments that provided care (see Table 2). To calculate utilization measures, we included encounter timestamps and services rendered by department in our analysis. No distinction was made between walk-in and scheduled appointments. General Medicine included primary care and primary care encounters included in Programs of All-Inclusive Care for the Elderly (PACE). Specialty services included a wide range of departments including physical therapy, cardiology, neurology, gastroenterology, endocrinology, other PACE care (occupational/physical therapy, home care), pediatric development services (PDS) and imaging. Other departments included in the analysis were Health Education, including nutrition, diabetes management, and smoking cessation; Case management, Family Planning, Vision services, Women’s Health, HIV, and Substance Use Disorder Services (SUDS) included relevant case management, medication-assisted treatment (MAT), and counseling. Mental Health included all mental health services not part of another category (such as SUDS) and psychiatry.

Table 2.

Healthcare Service Utilization by Department Patients Receiving Care at a Large Federally Qualified Healthcare Network, Stratified by General Population vs Justice System Impacted, Southern California, January 2, 2020 to February 16, 2022

	FULL SAMPLE N = 226,9701			GENERAL PATIENT POPULATION N = 223,5901			ADULTS IMPACTED BY THE JUSTICE SYSTEM N = 3,3801
Department	Number of Patients	Percent Engaged per Department	Number of Encounters by Department	Number of Patients	Percent Engaged per Department	Number of Encounters	Number of Adults	Percent Engaged per Department	Number of Healthcare Encounters
General Medicine	209,075	92.1	972,672	205,823	92.1	948,715	3,252	96.2	23,957
Specialty	31,967	14.1	186,365	31,182	13.9	181,965	785	23.2	4,400
Health Education	27,657	12.2	80,518	27,162	12.1	79,207	495	14.6	1,311
Case Management	26,910	11.9	101,696	25,778	11.5	96,498	1,132	33.5	5,198
Family Planning	26,079	11.5	49,241	25,817	11.5	48,798	262	7.8	443
Mental Health	25,044	11.0	250,114	24,113	10.8	242,414	931	27.5	7,700
Vision	18,767	8.3	30,786	18,208	8.1	29,917	559	16.5	869
Women’s Health	18,240	8.0	193,712	18,065	8.1	192,381	175	5.2	1,331
HIV	4,370	1.9	23,933	3,916	1.8	20,685	454	13.4	3,248
SUDS	4,139	1.8	43,574	3,496	1.6	33,357	643	19.0	10,217

Data analysis.

We conducted descriptive and statistical analyses using the statistical software package “R” Version 4.1.2.¹⁵ The study includes data for sociodemographic characteristics stratified by general patient population (i.e., no report of contact with the justice system) versus adults impacted by the justice system (Table 1). Pearson’s Chi-Squared test and Fishers exact test were used to assess statistical differences between groups. We also analyzed service utilization patterns by reporting the total number and percentage of patients in each group who had services in each department and the total number of encounters per department (Table 2). Figure 1 shows the median number of encounters to illustrate encounter frequency per department in each group, as the median is robust to data variability and influence of high utilizers.

Figure 1.

Median Healthcare Service Encounters Among Patients Receiving Services in Each Department, Stratified by General Population vs Justice System Impacted, Federally Qualified Healthcare Clinic System, January 2, 2020 to February 16, 2022

Ethics statement.

This study was approved by the Human Subjects Protection Program at the University of California San Diego. All patients aged 18 years or older must complete and sign Family Health Centers of San Diego’s Broad Consent form to receive treatment; it includes a specific authorization for the use of de-identified health information for population health and quality improvement studies. We have only included patients who had an up-to-date Broad Consent form within their EMR.

Results

Sociodemographic characteristics of the full sample.

Our network included a sample of 226,590 adults who received services between Jan 2, 2020 through Feb 16, 2022 (See Table 1). Overall, more than one-half of participants were 45 years or younger and 46% were male. A similar proportion were White (38%) and Latino/Hispanic (34%), while Non-Hispanic/Unknown, Black adults accounted for 8.1% of the full sample. English was the preferred language for services by about two-thirds of all patients followed by Spanish (31%). A similar proportion were employed (33%) or unemployed (30%); 43% of all patients reported having completed high school while 17% reported having less than a high school education and an equal proportion (17%) did not report their educational attainment. Overall, 23% reported a college or post-graduate education. Housing insecurity was high among patients: about two-thirds of patients (68.7%) reported doubling up, homelessness, or residing in a shelter, transitional housing, streets, canyons, or migrant camps.

Sociodemographic characteristics of adults impacted by the justice system vs general patient population.

One-half of patients impacted by the justice system were ages 27-45 years vs. 40% among the general patient population. The justice impacted patient population is predominantly male (75% vs. 45% of general population patients) and more likely to be White (45%) or Black (16%) than general population patients (i.e., 38% and 8%, respectively). Justice-impacted patients were more likely to speak English (91%), be single (79%), and unemployed (64%) than the general patient population (i.e., 64%, 54%, and 30%, respectively). Justice-impacted patients were significantly less likely to own or rent their housing (30%) and were more likely to be housed in shelters or transitional housing (42%) than the general patient population (i.e., 78% and 4.7%, respectively). Lastly, justice-impacted patients were more likely to be veterans (3.4% vs 1.4% among the general patient population).

Health care service engagement.

Table 2 presents health care service utilization encounters by major service departments among all adult patients and stratified by the general adult patient population and adults impacted by the justice system. Among the full sample, the General Medicine services were most commonly used (92.1% of all adults in the sample use these services), followed by Specialty care (14.1%), Education (12.2%), Case Management (11.9%), Family Planning (11.5%) and Mental Health (11%); less than 10% of patients used each of the remaining service departments. Similar patterns were observed among the general patient population. However, the proportion of adults impacted by the justice system using these services varied from proportions observed in the full sample or general patient population. For example among adults impacted by the justice system, General Medicine services were the most commonly used services (96.2%), followed by Case Management (33.5%), Mental Health (27.5%), Specialty care (23.2%), Substance Use Disorder Services (19%), Vision services (16.5%), Health Education (14.6%), HIV services (13.4%); less than 10% of patients used each of the remaining service departments (i.e., Women’s Health, Family Planning).

Median health care service encounters by patient subgroup.

We examined the median service utilization rates for each department by patients using those services; data for all patients, the general patient population and adults impacted by the justice system are shown (see Figure 1). Adults impacted by the justice system revealed higher median rates of service use for General Medicine (5 visits vs. a median of 2 visits for general population patients) and Substance Use Disorder Services (8 visits vs. 2 median visits for general population patients). Adults impacted by the justice system had a lower number of median mental health service visits than the general population (i.e., 4 vs. 6 respectively). Service use across patient groups was similar for the remaining departments.

Discussion

While research regarding health care utilization of adults impacted by the justice system is beginning to flourish, much remains unknown, especially regarding their use of safety-net services such as federally qualified health care clinics (FQHC) and as compared to the general patient population. This study leveraged administrative electronic medical record data to advance our understanding of health care use by adults impacted by the justice system, which in this study includes persons who reported diverse experiences such as arrests, appearances in court to incarceration. The large FQHC underlying this analysis is located in a geographically and demographically diverse region of Southern California.¹³ In brief, patients who reported justice system contact experienced vulnerabilities in education, employment and housing and they engaged widely with nearly all of the departments, with utilization of services being particularly high across all of the departments examined, as compared to the general adult patient population.

The demographic characteristics of adults impacted by the justice system mirror profiles of community-based samples of reentrants and those in carceral systems^9,16,17; for example, persons younger than 65 years of age, males, and African American/Black and Hispanic/Latino adults represented a disproportionate share of the study’s sample. A low educational attainment was pervasive among patients impacted by the justice system.¹⁶

Providers likely need to adopt multipronged approaches to ensure that patients impacted by the justice system have the information, skills, and tools needed to coordinate care for acute and chronic conditions, can use medications prescribed, and engage in other health-supporting activities. Programs that build health literacy can be an important service for individuals who have experienced incarceration or other contact with criminal legal systems. Additionally, health-related materials should be tailored to the needs of persons who may experience numeric, textual, or digital literacy challenges.^18–20 For example, providers can employ the “Teach-Back Method” [i.e., patients restate care instructions] during medical encounters to reinforce understanding of treatment regimens and correct misunderstandings.²¹ Other methods may also help individuals at risk of low health literacy access health content, including art-based strategies (e.g., drama/storytelling, drawing), interactive methods (e.g., role playing, peer-led activities), or technology based approaches (e.g., text messaging, informational or interactive websites to practice new skills).²² Offering activities that can build health literacy in pre and post-release settings that serve adults impacted by the justice system may aid skill development and confidence in applying those skills and knowledge. For example, areas that can be addressed include digital literacy (e.g., using health care patient portals, searching for health information online), self-advocacy in medical encounters, or medication management. These areas may be helpful particularly for persons who have longer experiences of incarceration and may be less familiar with technology or less comfortable with leading their health care in community settings after lengthy periods of institutionalization. Peer support from adults with lived experience can also be helpful in providing role-modeling and individualized coaching, mentorship and social support for adults exiting carceral settings.¹⁸ California’s Peer Support Specialist training program is designed to help augment the workforce of adults with lived experience²³ and this approach has been used successfully by providers seeking to engage adults exiting carceral settings.^10,24

Reducing administrative and structural barriers to health care use are critically needed. An estimated 26% of adults impacted by the justice system are uninsured¹⁶ which can contribute to delayed and forgone care or increased use of emergency care or hospitalization.^16,20,25 To address barriers to health care use, in 2023, California received a federal waiver to expand Medicaid health insurance (i.e., Medi-Cal in California) to persons in carceral settings starting 90 days pre-release. This approach seeks to expand access to selected services in correctional settings and facilitate access to, uptake of and continuity of care in community-based health care post-release.²⁶ This policy defines adults impacted by the justice system as a priority population and community organizations are being contracted to address this community’s unique needs through services that can address health and the social determinants of health (e.g., case management, housing navigation, personal care and homemaker services, medically tailored meals, sobering centers, home modifications, respite services, and others).²⁷ This approach is promising and important given that services such as case management are in high demand by this patient population due to the need for multiple health care services.

Data from this analysis showed that adults impacted by the justice system are engaging widely across nearly all service departments and at higher rates than the general patient population. Collectively, these data can inform policymakers’ plans to finance health care services and inform health care organizations’ staffing models to address the unique needs of adults impacted by the justice system. For example, in California, Medicaid can reimburse providers for services provided by Peer Support Specialists. This approach is important in helping to develop a trained workforce that can support primary care providers and offsetting the costs of peer-led services.²³ Additionally, organizations that hire individuals with lived experience such as trained Peer Support Specialists may be perceived as having high cultural competence and being welcoming of adults impacted by the justice system and peer staff may be able to help foster health care utilization among this patient subgroup as has been shown in other settings (e.g., mental health services).^28–30 Care coordination pre-release and post-release between carceral settings and community providers can also help adults returning to the community and may help reduce disruptions in care and medication management.

Limitations.

This study is novel in its approach to assess health care utilization of adults impacted by the justice system and who receive care in a large safety-net clinic. Results should be considered in light of the following factors. Results may not apply to other communities, yet the methods employed here may guide other health care systems seeking to better understand the experiences of patients impacted by the justice system. The data were culled from the network’s EMR using a hybrid method of key word search and manual extraction drawing from self-reported sociodemographic data and clinicians’ notes. The clinic network lacks systematic screening for criminal-legal system contact. Consequently, our analysis likely undercounts the number of adults directly impacted by the justice system since neither the health care system or providers systematically inquire about such contact or providers may not have documented it while some patients may not have disclosed it.

Data were available from 2003 through 2022 at the time that the database was queried, but patient identification for this analysis was restricted to January 2020 through February 16, 2022 due to server hardware capacity and staff resources available for a manual review of patient charts. Patients’ health literacy was not directly assessed ³¹; however, extant data underscore the challenges that persons with low educational attainment face in developing the skills and accessing resources to maintain or improve their health or access care consistently.^20,32,33 Despite these considerations, the study sheds light on health care service utilization by a diverse sample of adults impacted by the justice system and signals that targeted identification and tailored services and policy interventions are needed to support the health of this community.

Conclusions.

This study highlights the need for multipronged interventions which address the health and the social determinants of health of adults impacted by the justice system in carceral, community-based, and health care settings. Tailored services for this community are critically needed and investments in health care workforce development in key areas such as mental health, substance use, social work can aid communities in supporting persons exiting carceral systems or who are impacted by the criminal legal system. California has implemented a Peer Support Specialist training program that can be helpful to expanding the workforce of persons with lived experience working in health care settings. Peers may be able to help engage and connect adults impacted by the justice system and ensuring continuity of care once they are established as patients.²³ Such an approach is promising based on the experiences of the Transition Clinics Network²⁴ and mental health service providers.³⁰ Longitudinal research is needed to shed light on the long-term needs of adults impacted by the justice system and how novel approaches such as those mentioned here impact long-term well-being. Such data may inform the type and structure of services that are offered to support the long-term health and social outcomes of persons reintegrating into the community from carceral settings or who have experienced other contacts with legal systems.

A list of Abbreviations Used:

FQHC

Federally Qualified Health Clinic

FHCSD

Family Health Centers of San Diego

EMR

Electronic Medical Record

PACE

Programs of All-Inclusive Care for the Elderly

PDS

Pediatric Development Services

SUDS

Substance Use Disorder Services

MAT

Medication Assisted Treatment