Introduction

In 2016, over 600,000 individuals were admitted to a state or federal prison in the United States (US) [1], and 10.6 million people were admitted to a US jail [2]. Nearly all of these individuals will return to the community following their incarceration [2, 3]. Previous research has shown that people in prison and jail have a high burden of chronic physical health conditions, infectious diseases, disability, mental illness, and substance use [4–9], which require ongoing health care during incarceration and upon release. People in prison or jail have higher odds of hypertension, asthma, arthritis, cervical cancer, and hepatitis compared to the general population [4]. In addition, it is estimated that 32% of individuals in prison and 40% of individuals in jail have at least one disability related to hearing, vision, cognition, mobility, self-care, or independent living [5]. Further, over half of all individuals in prison or jail have a mental health problem or meet criteria for drug dependence or abuse [6, 8]. In Hennepin County, MN, where this study was conducted, 52% of the jail population reported having a mental health problem, and 20% reported using or abusing opioids [10]. Addressing the health needs of individuals cycling in and out of prisons and jails requires coordination between community and correctional health care providers. However, this coordination is often inconsistent or nonexistent and complicated by barriers like sharing medical records between correctional and community health providers [11–13].

While some specialized programs have been developed to meet the needs of individuals leaving prison or jail, most people do not interact with these programs because they serve limited geographical areas [14–18]. The majority of individuals leaving prison or jail establish care with a community-based physician not affiliated with a re-entry program. Yet, little is known about community-based providers’ understanding of transitions related to incarceration or providers’ experiences with navigating care transitions for individuals who were recently incarcerated. Such information could identify opportunities to improve continuity of care for patients transitioning between the community and prison or jail.

The goal of this qualitative study was to address this gap in the literature by describing physicians’ views on, and experiences with, patient transitions into and out of incarceration and how these transitions impact care plans for patients.

Methods

Semi-structured interviews were conducted with physicians from three safety-net clinics in Hennepin County, Minnesota between October 2018 and May 2019. Hennepin County is Minnesota’s largest county with a population of 1.3 million people and contains the city of Minneapolis and several large, suburban cities. These clinics were selected because they serve a high number of patients who have histories of justice system involvement. Further, each clinic serves as a medical training institution that may be ideal sites for implementation of future physician training curriculum. This study was approved by the Institutional Review Boards at Hennepin Healthcare Research Institute and the University of Minnesota. Written informed consent was obtained from all participants.

Interview questions were developed by research team members LJ, TW, and RS, and were initially generated based on known gaps in physician training and education related to working with patients with justice system involvement. Given this gap, interview questions were created to assess physicians’ knowledge of the US criminal justice system and how it impacts their patients. These questions were reviewed and refined using an iterative process by both the research team and external stakeholders with backgrounds in public health and criminal justice. The interview questions were further refined after conducting two pilot interviews. The final interview guide covered five main topics: 1) knowledge of the criminal justice system; 2) medical education and training related to the criminal justice system and working with patients involved in the criminal justice system; 3) if and how physicians screen patients for criminal justice system involvement; 4) the physical, mental, and substance-related health needs of patients with criminal justice system involvement; and 5) recommendations on how the health care system should meet the needs of patients with histories of criminal justice involvement.

After receiving approval from the medical director at each institution from which we recruited, we extended invitations to physicians by email to participate in the study. We used purposive sampling to recruit physicians within the specialties of internal medicine, psychiatry, family medicine, and emergency medicine. In recruitment materials, participants were told that the purpose of the study was to better understand physicians’ knowledge and perceptions of the criminal justice system and its impact on patients to inform development of training curricula and programs that better meet the needs of providers and patients.

A total of 47 physicians were directly contacted by LJ. Of those 47 physicians, 27 (57%) responded to our inquiry, and nearly all of those who responded to our inquiry agreed to participate in a one-time interview (n = 23, 85%). Participants identified as predominantly White (n = 19, 83%), non-Hispanic (n = 21, 91%) and male (n = 16, 70%), with an average age of 48 years (range = 36–75).

All interviews were facilitated by LJ, a female project coordinator with a Master’s degree in Public Health and graduate-level experience and training in conducting qualitative research. A student notetaker was also present during interviews. Interviews lasted between 36 and 90 minutes, and no repeat interviews were conducted. Interviews were conducted in private offices at participants’ workplaces. One interview was conducted at a public coffee shop at the participant’s request. Following each interview, LJ compiled field notes to record observations. Field notes were used to record information regarding interview setting and length along with notes on interviewee responses and body language. These notes were used to support transcription of interview audio recordings. Data collection concluded upon agreement amongst the research team that no new findings were emerging and data saturation had been reached. Interviews were audio recorded and transcribed verbatim using a web-based transcription service. All transcripts were reviewed for accuracy by members of the study team. Transcripts were not returned to participants for comment or correction.

A thematic analysis approach was utilized to identify patterns within the data [19]. For this thematic analysis, both deductive and inductive coding was used [19, 20]. Codes are words or short phrases that are used to summarize key features of the data [19]. Deductive coding was conducted by applying a priori codes that were defined based on the interview guide. For example, “knowledge of the criminal justice system” was one a priori code used to organize data related to physicians’ ability to define criminal justice terminology (i.e., prison, jail, probation parole). Inductive coding was conducted by allowing data-driven codes that were not predefined to emerge [19, 20]. “Continuity of care,” the focus of this manuscript, was one such inductive code. The codes were organized into a codebook from which we identified patterns, or themes, that were present across the data [21].

Three transcripts were independently coded by LJ, RS, and CFB and a preliminary codebook containing both deductive and inductive codes was developed [21]. This preliminary codebook was independently applied to one transcript by LJ, RS, and CFB, and the team met to discuss codebook revisions. An iterative process was used to further refine the codebook, in which a subset of transcripts were independently coded by two of the coders and compared until the researchers were confident that a consensus on the meaning of all codes had been reached. The codebook was independently applied by LJ, RS, and CFB to the remaining transcripts. The codebook was ultimately refined to 18 codes (12 deductive and 6 inductive). From these codes, multiple themes were identified, and just one code, “continuity of care,” is the focus of this paper. LJ, RS, and CFB identified, “continuity of care,” as warranting further sub-analysis given the complexity of the data contained within this code. This sub-analysis revealed distinct patterns related to continuity of care, which was not one of the guiding questions of this research project. Those themes are described herein. NVivo (version 12) was used for all coding and analyses. Participants did not provide feedback on the findings.

Results

Four themes emerged from our interviews with physicians related to care transitions and factors that physicians perceived as contributing to discontinuity of care for patients entering and exiting incarceration: (1) disruptions in patient-physician relationships, (2) barriers to accessing prescription medications, (3) disruptions in insurance coverage, and (4) problems with sharing medical records. Physicians discussed how these factors impacted patients differently depending on whether the patient was transitioning from the community to prison or jail, or transitioning from prison or jail back to the community. In some instances, physicians’ knowledge of a patient’s past or upcoming incarceration played a significant role in informing the physician’s care plan for the patient.

Discussion

Physicians identified multiple ways in which incarceration disrupts continuity of care for patients transitioning between the community and prison or jail. Barriers to maintaining continuity of care included disruptions in patient-physician relationships during periods of incarceration, barriers to medication access (especially those for opioid use disorder), lapses in health insurance coverage, and difficulties obtaining medical records from correctional facilities. Thus, physicians were tasked with developing care plans for patients with histories of incarceration without complete information on the health care services patients did or did not receive during their incarceration. Further, physicians recognized that patients leaving prison or jail without health insurance might have difficulty affording medical visits and medications to address their complex health needs. Given these system failures identified by physicians, physicians implemented ad hoc solutions, for example, adapting care plans to reduce the financial burden of health care costs on uninsured patients.

The barriers to continuity of care identified by providers, including disruptions to patient-provider relationships, medication access and health insurance coverage, align well with findings from previous research that explored experiences of individuals recently released from incarceration, as well as physicians who work in jail or prison settings [22–25]. Such disruptions to care may contribute to poor health outcomes and an increased risk of recidivism. Previous research has found that maintaining a continuous relationship with a primary care provider is associated with a lower likelihood of return to prison or jail [26], suggesting that this relationship plays an important role in addressing health-related factors that contribute to recidivism. The current study adds the perspectives of community-based physicians to this existing body of literature, highlighting the barriers physicians face to providing care to patients recently released from prison or jail and their approaches to mitigating these barriers. In particular, this study finds, in lieu of functioning systems to support these transitions, physicians took it upon themselves to troubleshoot issues with little to no support for such complex transfers of care.

Physicians in our study were also concerned that many correctional facilities did not provide standard of care, particularly for patients with opioid use disorder. While access to medications for opioid use disorder has expanded, there are still many jails and prisons that do not provide this treatment to patients [27], and obtaining information on the particular policy a certain prison or jail has regarding treatment for opioid use disorder can be difficult. Further, physicians shared the difficulties their patients faced accessing community-based health care without health insurance, which suggests that linkage to health insurance is needed during transitions to the community as well.

Implications

Physicians’ ad hoc approaches to addressing care transitions for patients entering or leaving jail or prison, suggests that there is a lack of systems level coordination to support these transitions. Findings from this study indicate that there is a need to develop a system of care that ensures proper transitions of care between correctional facilities and community health care providers. Currently, specialized transition programs exist to facilitate transfer of care to community providers [14–18]. However, these specialized programs serve limited geographical areas and are not accessible to all individuals transitioning between prison or jail and the community.

In order to improve health outcomes and reduce recidivism, systems-level solutions are required to link care between community settings and prisons and jails. Physicians’ reports of their difficulties communicating with correctional health care providers and obtaining medical records from correctional facilities suggests that increased communication channels between the two settings are needed. This could include being more transparent about what services are offered in the correctional setting so that community providers know what patients have access to during their incarceration and more readily sharing medical records when consent is obtained from patients [28]. Because patients may not yet have an established relationship with a community provider at the time of release, they should also be provided access to free copies of their medical records from correctional health care providers to help reduce barriers to establishing care in the community.

Given the barriers patients face gaining access to health insurance coverage upon release from prison or jail, correctional health care agencies should consider partnering with health agencies to build systems to support patients upon release. This could include ensuring that patients leaving prison or jail are enrolled in health insurance coverage that will cover their medical expenses when they return to the community. This could be done through partnerships with community-based organizations that provide health insurance navigation services. Enrollment in health insurance, combined with improved medical record sharing, would allow patients to quickly establish care with a new provider in the community and receive medications they may need in a timely manner.

Medicaid and managed care organizations that administer Medicaid could play a key role in developing these system-level supports. Access to Medicaid for low-income adults was expanded under the Affordable Care Act, with most individuals released from prisons likely to be eligible for such coverage [29]. Medicaid expansion provides an opportunity to link individuals who are incarcerated to health insurance coverage prior to release and coordinate care during their transition back to the community. The barriers identified by physicians in this study suggest that coordination of care for this population through a Medicaid managed care organization would alleviate barriers to medication access, lapses in health insurance coverage, and difficulties obtaining medical records from correctional facilities. Such system-level support may lead to improved health outcomes, including a reduced risk of death, as well as reduced likelihood of recidivism for individuals with a history of incarceration.

The findings from this study highlight several important areas where further research is needed. Future research should implement and test system-level supports, such as those described herein, for patient transitions between incarceration and the community to examine their impact on health and well-being for individuals with a history of incarceration. Further, because this study was conducted in one urban, Midwest county, future research should explore rural physicians’ perceptions of the impact of incarceration on continuity of care for patients. Finally, additional research is also needed to better understand the impact of transitions within the criminal justice system, such as transitions from jail to prison, and how this impacts continuity of care and overall health.

Supporting information