Background

Criminal justice (CJ)-involved persons, defined as persons with a history of incarceration in prison or jail and persons who have been under community corrections supervision (probation, parole), are disproportionately impacted by HIV infection. A high prevalence of substance use disorders, mental health disorders and racial and ethnic health disparities contribute to this increased burden [1, 2]. CJ populations experience high rates of unstable housing, unemployment, and poverty, which negatively influence retention in longitudinal HIV care [3]. Incarceration and subsequent release from a correctional facility have been associated with interruptions in antiretroviral treatment (ART) [4], loss of viral control [5] and failure to maintain adherence to ART [4, 6, 7]. Approximately one in seven HIV-infected persons in the United States (U.S.) passes through the CJ system annually [8], providing an opportunity to address the HIV care continuum [3] and to deliver interventions designed to improve access to HIV care and ART.

The majority of persons involved in the U.S. criminal justice system are men, with women representing only 7% of the prison population and up to 25% of persons on probation in the community [9, 10]. CJ-involved women and transgender persons have unique health challenges, yet these populations remain understudied. Incarcerated women have higher rates of substance use and addiction [11, 12], HIV infection [13, 14],hepatitis C virus infection [15], and mental health disorders [16, 17] compared to incarcerated males [18]. Transgender persons, particularly male-to-female (MTF) transgender women, experience high rates of HIV infection [19, 20], clinical depression [21], unstable housing [22], and gender violence and economic discrimination [23], which can contribute to risky sexual behaviors [24, 25].

Persons who cycle in and out of jails and prisons are at risk of transmitting HIV to sexual and drug-using partners; therefore, developing and implementing interventions to improve the delivery of HIV treatment to CJ-involved persons in order to achieve viral suppression is a public health priority. The Seek, Test, Treat, and Retain (STTR) strategy to control the HIV epidemic has been proposed as a public health approach to improve HIV outcomes through greater access to ART [26–28]. The STTR strategy is multi-pronged and includes increasing the identification of HIV-infected persons through expanded testing, linkage to HIV providers, improving access and adherence to ART, and maintaining long-term retention in care. If achieved, this, in turn, decreases community transmission of HIV by creating a population of HIV-infected persons who are less infectious by effective ART [29].

In 2010, the National Institutes of Health/National Institute on Drug Abuse funded a research initiative to investigate the application of the STTR strategy among CJ-involved persons [25]. As part of this initiative, we launched the CARE+ Corrections Study conducted in Washington, D.C., a city with an HIV prevalence estimated to be at 2.5% of all residents and 5.8% among Black males [30]. The goal of the parent study was to develop and evaluate the efficacy of a technology-based intervention designed to improve linkage to community HIV providers and ART adherence following release from a correctional facility. With this intervention, our study aimed to address the Treat and Retain aspects of the STTR strategy using an intervention to leverage information and communication technology (ICT) tools. ICTs have the potential to be scalable among correctional facilities and community partners who assist persons released from prison or jail, also referred to as “returning citizens”. Given the gender diversity of the CARE+ Corrections Study participants, we herein describe their baseline characteristics and compare risk behaviors and HIV care indicators across men, women, and transgender persons.

Methods

Results

A total of 497 persons were pre-screened and 219 were assessed for study eligibility. Of the 112 persons who provided informed consent and were enrolled in the study, 110 completed baseline data collection and were included in this analysis (see CONSORT flowchart in Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT Diagram.

https://doi.org/10.1371/journal.pone.0169078.g001

Table 1 displays the demographic characteristics and site of enrollment for these participants; 70% were enrolled from the community and 30% were enrolled within the DC DOC. Mean age was 40 (SD = 10.5), and most were Black or African American (85%). The majority of participants were male (58%), 24% were female, and 18% were transgender, all of whom were MTF transgender women. There were no differences across gender groups in race and lifetime number of incarcerations (mean 10.8, SD = 9.9). The length of last incarceration was shortest for women, at 4.8 months (SD = 11.5) compared to 19.8 months (SD = 47.4) for men and 8.9 months (SD = 21.4) for transgender women (p<0.0001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic Characteristics of CARE+ Corrections Study participants, stratified by gender (N = 110).

https://doi.org/10.1371/journal.pone.0169078.t001

Table 2 displays sexual behaviors, substance use and mental health characteristics. Nearly half (47%) of the study sample had engaged in condomless sex in the three months prior to their last incarceration, but this did not differ across genders. However, engaging in exchange sex was significantly higher for women and transgender women (54% and 65%, respectively) compared to men (12%, p<0.001). Among men and transgender women, 27% of men and all transgender women reported having had sex with a man (p<0.001), and 17% of men and 79% of transgender women reported having anal sex with a man in the three months before their last incarceration (p<0.001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sexual risk behaviors, substance use, and mental health disorders among CARE+ Corrections Study participants, stratified by gender (N = 110).

https://doi.org/10.1371/journal.pone.0169078.t002

Drug dependence (57%), lifetime history of injection drug use (15%), recent non-injection drug use (65%), and hazardous alcohol consumption (45%) were all highly prevalent in this sample but did not differ across genders. However, transgender women were significantly more likely to have participated in drug treatment compared to men and women (90% versus 61% and 50%, respectively, p = 0.0117). In terms of mental health, depressive symptoms (52%) and previous diagnosis of schizophrenia (23%), personality (17%) and bipolar disorders (59%), and post-traumatic stress disorder (44%) were all highly prevalent but did not differ across genders.

Table 3 displays baseline HIV treatment and care indicators. Overall, more than 80% of study participants had a regular HIV healthcare provider prior to their last incarceration, and nearly all (91%) had ever taken HIV treatment. Only half (51%) indicated optimal HIV treatment adherence in the 3 months prior to their last incarceration, and this did not differ by gender. However, being on HIV treatment during their last incarceration was significantly lower for women compared to men and transgender women (67% among women versus 96% and 95% among men and transgender women, respectively, p = 0.001). In addition, all transgender women reported currently having a regular HIV provider in the community, versus 75% among men and 77% among women (p = 0.02). Ninety-three (88%) study participants completed phlebotomy and 13 had their medical charts reviewed for baseline HIV plasma viral load and CD4 counts. At baseline, 80% of transgender women, 69% of men, and 48% of women achieved HIV viral suppression (<200 copies/mL), although these differences were not statistically significant. Women had the highest proportion of having a CD4 count ≤200 compared to men and transgender women (5% for transgender women versus 8% for men and 17% for women), but this also was not statistically significant.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. HIV treatment and care indicators among CARE+ Corrections Study participants, stratified by gender (N = 110).

https://doi.org/10.1371/journal.pone.0169078.t003

Discussion

HIV-infected persons with a history of incarceration in Washington, DC represent a vulnerable population, experiencing a high frequency of lifetime arrests as well as high prevalence of drug use, problem drinking, mental health disorders, and sexual risk behaviors. In our sample, engagement with HIV care before and after the most recent incarceration was high among the sample, but pre-incarceration ART adherence was sub-optimal and females had less access to ART during their incarceration. Baseline CD4 counts were lower and viral suppression was less frequent for women compared to males and transgender persons but these findings were not statistically significant. ART adherence remains a challenge for persons recently involved in the CJ system and needs to be the focus of public health efforts to reduce community viral load.

The majority of the study population reported having access to a regular healthcare provider and HIV treatment prior to their most recent incarceration (83% and 91%, respectively); however, only 51% reported optimal HIV treatment adherence on the VAS scale during the three months prior to their most recent incarceration. Nearly all (89%) reported receiving HIV treatment during their recent incarceration and 86% reported being on ART at the time the baseline assessment was completed in the community after release; 62% had a suppressed viral load based on objective laboratory data. This proportion of participants with viral suppression is higher than the estimate that only 30% of HIV-infected persons in the U.S. have achieved viral suppression; and more similar to recent studies that estimated viral suppression among persons in HIV care [40–42]. Collectively, these data suggest that this population in Washington, DC has reliable access to HIV providers prior to, during, and after incarceration, yet ART adherence could be improved and this would lead to higher rates of viral suppression. A recent study demonstrated that self-report of being engaged in HIV care was not always reliable when compared to clinic visit data [43]. Therefore, further research is needed to determine the causes of poor ART adherence in this population and the relative contribution of missed clinic visits. Regardless, there is a need to improve ART adherence among persons who are viremic and have concurrent risk behaviors given that they may transmit HIV to sex and drug using partners, supporting the premise that this is a high priority population for intervention.

The relatively large proportions of transgender women and men who have sex with men (MSM) that were recruited into the study was not anticipated. These groups, particularly HIV-infected MSM and transgender women, remain largely hidden within correctional populations and few studies have estimated their prevalence. The multi-site Enhance Link Initiative enrolled over 1200 HIV-infected persons with a history of jail incarceration and among these, 20% identified as MSM or bisexual [44], which was consistent with the proportion observed in our study. These findings and our results suggest that HIV-infected MSM, bisexuals, and transgender women may represent a larger proportion of HIV-infected persons behind bars than previously recognized. More research is needed to understand the epidemiology, risk profiles, and unique barriers to health services that these persons are likely to encounter in the coercive and highly stigmatized environment of correctional facilities.

The study revealed some important gender differences among our study population. Women exhibited worse outcomes compared to both men and transgender women, which is consistent with the literature [16, 17, 45, 46]. Women in our study were less likely to receive ART during their incarceration compared to men and transgender women and were least likely to have viral suppression at the baseline assessment. Being less likely to receive treatment during the recent incarceration may have been mediated by having shorter incarcerations. Women in our sample receiving ART had longer incarceration periods (3.38 months, SD = 2.14) compared to women who did not receive ART (mean = 0.74 months, SD = 2.14) (p = 0.003). Prior research has found that acceptance of ART in a correctional facility among women can be affected by stigma, perceived trust of a provider, and satisfaction of care [47], but there are few studies surrounding these issues and results are inconclusive. For example, contrary to our findings, another study found that even with shorter incarcerations, women were more likely to achieve viral suppression during incarceration compared to men [46]. More research is needed to understand acceptability of ART in a correctional setting and if shorter incarceration periods impose structural barriers (e.g., waiting to be seen by a provider) that reduce access to ART among women. This difference in access to ART is important because it is well documented that women, on average, serve shorter sentences compared to men [48].

Transgender women in our sample were more likely to have an HIV care provider prior to incarceration and at the baseline visit compared to both men and women. The reasons for these findings remain unclear, yet this is encouraging given other studies have shown HIV-infected transgender women to have poorer access to care including ART and a greater number of social and structural barriers to care [49–55]. Our findings suggest qualitative research is needed to identify both facilitators and barriers to HIV care and ART among transgender persons who have a history of incarceration.

Our study had several limitations. The generalizability of our findings to other correctional populations is limited by several factors: the sample size was small and therefore lowered the power to detect significant differences between gender groups that may have existed; the study population was relatively homogeneous in terms of race (85% were Black/ African-American); and one-half of study participants received a phone which may have prevented some persons from enrolling given there was a 50% chance of not receiving a phone. We opted to recruit study participants from the DC DOC facilities and from the community following release due to difficulty in launching recruitment inside of the DC DOC. This resulted in a more heterogeneous population that may differ with respect to HIV care and risk behaviors. However, there were few significant differences in risk behaviors by enrollment site, and no significant differences in demographic characteristics or HIV care (data not shown). Community recruitment largely occurred through social support agencies that supported returning citizens. This may have introduced some selection bias favoring HIV-infected persons who were engaged with these agencies; however, we also recruited through street-outreach in an effort to enroll a more representative population of returning citizens. Our analysis of baseline viral suppression included both laboratory testing conducted by the study and PVL data obtained from medical chart review, which increased the variability of length of time between study entry and assessment of viral suppression. Despite these limitations, this analysis represents a unique examination of HIV care and risk behaviors among recently incarcerated persons, including transgender women.

Conclusions

Improving HIV treatment outcomes among CJ-involved persons is an important component to the broader STTR strategy given the vulnerability of this population and the fact that the vast majority of incarcerated HIV-infected persons eventually return to their communities. Our findings emphasize that improving ART adherence among CJ-involved persons remains a high priority for intervention, particularly among CJ-involved women who are at greatest risk of poor HIV outcomes. Differences in HIV care indicators and risk behaviors among men, women, and transgender women further support the need for additional research to develop gender-specific interventions to address the unique needs of each group.

Supporting Information

Acknowledgments

We would like to acknowledge funding from the National Institutes of Health, National Institute on Drug Abuse (R01DA030747) and institutional support from the Lifespan/Tufts/Brown Center for AIDS Research P30AI42853 and the District of Columbia Center for AIDS Research P30AI117970. We would also like to acknowledge our District of Columbia Department of Corrections partners (Drs. Beth Mynett and Reena Chakraborty) and community-based partners for their support and assistance in conducting this work. Lastly, we would like to thank the study participants without whom we could not do this work.