Incarceration of people living with HIV/AIDS: Implications for Treatment-As-Prevention

Abstract

Contact with the criminal justice system, including incarceration, is a common experience for many people living with HIV/AIDS. Optimism has recently been expressed that correctional facilities could be important locations for Treatment-as-Prevention (TasP)-based initiatives. We review recent findings regarding the effect of incarceration on patterns of HIV transmission, testing, treatment initiation and retention. We found that the prevalence of HIV infection among incarcerated individuals remains higher than analogous non-incarcerated populations. Recent studies have shown that voluntary HIV/AIDS testing is feasible in many correctional facilities, although the number of previously undiagnosed individuals identified has been modest. Studies have implied enhanced linkage to HIV/AIDS treatment and care in jails in the United States was associated with improvements in the HIV cascade of care. However, for many individuals living with HIV/AIDS, exposure to the correctional system remains an important barrier to retention in HIV/AIDS treatment and care. Future research should evaluate structural interventions to address these barriers and facilitate the scale-up of TasP-based efforts among individuals living in correctional settings.

INTRODUCTION

Since its first detection three decades ago, the HIV/AIDS pandemic has been profoundly shaped by criminal justice systems. Surveillance studies have commonly observed significantly higher prevalences of HIV infection among incarcerated populations as compared to analogous non-incarcerated groups [1–3]. The elevated seroprevalences among individuals held in correctional facilities are, for the most part, a result of the political and criminal justice structures that penalize groups who already face multiple HIV/AIDS-related vulnerabilities, including individuals addicted to illicit drugs, sex workers, and sexual/gender minorities [4]. However, evidence from a variety of studies, including molecular contact-tracing investigations and entry/exit antibody testing, have also identified within-prison transmission, facilitated by the dearth of evidence-based HIV prevention tools [5–7]. In addition, exposure to correctional facilities has been strongly associated with elevated levels of co-morbidities (i.e., hepatitis C, tuberculosis, untreated mental illness including unmanaged substance use disorders) and social/structural factors (stigma, discrimination, political disempowerment, socio-economic marginalization and exclusion from non-correctional healthcare systems) linked to increased risk of HIV transmission and pathogenesis [8–10].

Coincident with the emergence and spread of the HIV/AIDS pandemic has been the growth in the global population of individuals held in penal facilities [9,11,12]. According to the most recent estimates, approximately 10.1 million people were held in penal institutions, including police cells, pre-trial detention, so-called mandatory treatment facilities, jails, prisons and penitentiaries in 2011, equal to a rate of incarceration of 146 per 100,000 individuals [11]. Driven in many settings by increased numbers of criminalized drug users, global correctional populations have nearly doubled since 2005 [9,11]. While unequalled in size or scope, the United States’ correctional systems exemplify these trends [12]. In the 1980s, the United States enacted harshly punitive laws against illicit drug users, including mandatory minimum sentencing requirements, as a reaction to the emergence of crack cocaine, especially among the country’s urban African-American population. As a result, the number of individuals incarcerated in state and federal prisons increased by almost 500% in a generation [12,13]. More likely than the general population to be non-white [14], poor [12], not engaged in regular medical care [15], and suffering from untreated substance abuse [16], prisoners in the United States exhibit an HIV prevalence of 1.5% [17], and approximately 15% of all people living with HIV/AIDS are in contact with the criminal justice system each year [18].

Recent advances in our understanding of HIV/AIDS transmission dynamics, in particular the central role played by HIV-1 RNA viral loads (VL), have revealed the substantial impact of effective HIV treatment on the incidence of new infections [19,20]. In the wake of the HPTN-052 multicenter randomized controlled trial that demonstrated a >95% decrease in the likelihood of HIV transmission as a result of earlier highly-active antiretroviral treatment (HAART) [19], as well as other evidence from observational studies and mathematical models [21], renewed HIV/AIDS prevention campaigns with HIV treatment as a cornerstone have been launched. These seek to reduce the incidence of new infections by lowering the community-level HIV RNA viral load. Some key aspects of Treatment-as-Prevention (TasP) efforts are identifying undiagnosed individuals and engaging and retaining them in medical care including HAART, with the goal of achieving non-detectable plasma HIV-1 RNA viral loads (VL) [20,22].

Given the high proportion of individuals living with HIV/AIDS held in correctional facilities, optimism has been expressed that criminal justice systems could be an ideal focus of HIV testing and treatment efforts [23]. However, studies of people living with HIV/AIDS in community settings have consistently observed that incarceration is a substantial barrier to consistent engagement in HIV treatment and care, including antiretroviral therapy [24–26]. Thus, the criminal justice-based approach to individuals at high risk for HIV infection, especially people who use illicit drugs (PWID) and sex workers, and the high rates of incarceration borne by these groups, might compromise the impact of TasP-based campaigns [27]. In order to contribute to efforts to optimize TasP-efforts as well as reduce HIV/AIDS morbidity and mortality among incarcerated individuals, we performed this review of incarceration among people living with HIV/AIDS, focusing on HIV prevalence and transmission, HIV testing, and HIV treatment among incarcerated populations.

HIV PREVALENCE/RISK

As a result of the ethical and logistical challenges often faced by researchers working with incarcerated populations, the majority of scientific investigations into HIV/AIDS-related transmission and pathogenesis risks among prisoners have been cross-sectional serosurveys of infection with HIV or related conditions, including tuberculosis (TB), often anonymized and rarely longitudinal (i.e., before, during and after incarceration) or linked to external data sources. These surveys (for example, in the United States [28], Brazil [29] and Scotland [7]) have consistently identified elevated prevalence of infection compared to non-incarcerated populations. Meanwhile, observational studies conducted in non-correctional settings among groups at higher risk of infection with HIV, including injection drug users, sex workers, and members of sexual/gender minorities, have commonly associated incarceration with elevated risk of HIV infection in both crude and adjusted analyses [30,31].

Recent analyses of the prevalence of HIV and related infections in correctional populations have confirmed this pattern [32–36]. Of particular note are the results from an HIV/TB screening program established in six prisons in Zambia [34]. Among 2323 individuals entering, residing in or exiting the facility during the study period, 27.4% were living with HIV/AIDS, approximately twice the national level, and the prevalence of bacteriologically-confirmed TB, at 3900 per 100,000, was 4.5 times the national average [34]. The first published survey of HIV and TB in an African prison, the results suggest important linkages between HIV and TB infection in the prison and, through prison workers, prisoners and ex-prisoners, the local community [34]. Similarly, a representative nationwide survey of 402 soon-to-be-released prisoners in Ukraine found high seroprevalences of both infectious disease, including HIV (19%) and untreated substance use disorders, with over half screened positive for alcohol use disorders and over one-third reporting use of opioids in the 30 days prior to arrest [33]. Notably, more than half of the HIV-positive prisoners were unaware of their serostatus [33] and only five individuals (6.4%) were receiving ART. If TasP efforts are to have a beneficial impact on HIV incidence, the authors suggest, the under-diagnosis and under-treatment of people living with HIV/AIDS in prison systems must be addressed [33]. The studies from Ukraine and Zambia underline the important burden of co-infections described in many recent studies from prison settings, including high seroprevalences of hepatitis B and C [32,33,35–38] and TB [33,34,38–42].

Many studies among injection drug users in varied settings, including Canada [43], Thailand [44] and the United States [45] have identified a heightened risk of HIV infection associated with incarceration, possibly a result of higher rates of used syringe sharing within correctional facilities. A number of recent studies explored the effect of incarceration on sexual risk among African-Americans [46–48], who have not only borne among the heaviest burden of the HIV/AIDS epidemic in the United States, but make up approximately 40% of those currently incarcerated in the United States [49]. Among 1553 black men who have sex with men in six US cities, 60% reported a lifetime history of incarceration [48]. Although some have hypothesized the incarceration contributes to the high levels of HIV infection among African-Americans [50], and incarceration was common over the course of the study [47], there was no association between incarceration during follow-up and incident HIV infection [47]. In two separate studies, incarceration, in combination with marginal housing conditions, was associated with sexual risk behaviours among African-Americans [46] and in a multi-racial sample of young gay men in New York City [51].

HIV TESTING

Voluntary counseling and testing (VCT) for HIV infection has long been recognized as an important component of HIV prevention and treatment programmes [52]. In TasP efforts, which seek in part to lower community-level HIV viral loads through engagement in antiretroviral therapy, high levels of testing are prioritized in order to seek out undiagnosed individuals and link them to care [53]. Mathematical models have suggested that universal voluntary HIV testing followed by immediate ART initiation regardless of CD4+ cell count could lead to sharp declines in HIV incidence [54]. Since large proportions of HIV-positive individuals are unaware of their status, even in developed settings with advanced HIV/AIDS treatment and care systems where HIV tests are mandated to be a regular part of medical care [55,56], there is a clear need to increase testing rates, especially among populations with a high prevalence of HIV infection, such as incarcerated individuals. [17].

Due to the special conditions imposed by incarceration, in particular that they are coercive environments that constrain the freedom and privacy of individuals, there have long been concerns around the legal and ethical implications of HIV testing within penal settings [9,57,58]. As respect for patient autonomy, confidentiality and voluntary consent are fundamental aspects of healthcare and human rights, many national and international organizations oppose mandatory or compulsory testing of individuals in correctional settings, including the World Health Organization and UNAIDS [57,59]. In many correctional systems, early policies of compulsory HIV testing among all prisoners have been abandoned following legal challenges, the cost of providing HIV/AIDS care, and the finding that they were ineffective [59]. However, mandatory or compulsory testing continues in many state prison systems in the United States, Russia and some countries in Asia and the Pacific [59,60].

In their 2006 guidance document on HIV testing in prisons and other closed settings, the World Health Organization, UNAIDS and the United Nations Office on Drugs and Crime recommend that all prisoners have access to voluntary and confidential HIV testing, including pre- and post-test counseling [61]. In the United States, despite the wide variety of correctional settings and laws governing their operation [60], the United States Centers for Disease Control and Prevention recommends universal voluntary opt-out HIV testing for all adults in health-care settings, including correctional systems, unless prevalence of undiagnosed infection has been documented to be less than 0.1% [52]. Unfortunately, data on HIV testing within correctional systems globally is unavailable, and the distribution of different testing regimes — including mandatory, opt-in, opt-out, testing on demand, and none at all — is not known [59]. Further, only limited data exists on the actual practice of HIV testing within correctional settings, almost all from federal, state and local prison systems in the United States [61].

New findings from a multi-centre observational study conducted in jails in the United States offered some evidence of the possible results of jail-based voluntary testing and linkage to care initiatives [62]. Based on the recognition that 9 million individuals are admitted to a jail each year in the United States, the EnhanceLink initiative collected client and program-level data on patterns of HIV testing and linkage to care at 20 jails in 10 communities [62]. For HIV testing, the initiative’s objective was to demonstrate its feasibility within jail settings and identify new and previously-detected infections [62]. From 2007 to 2011, more than 877,000 admissions to the twenty institutions resulted in 212,464 (24.2%) individuals accepting an HIV test. In total, 822 (0.09% of all admissions, or 0.4% of all completed tests) newly-diagnosed individuals were detected [62]. In a study of a subset of these HIV-positive individuals, newly-diagnosed individuals were younger (34 years versus 41 years), exhibited far higher levels of HIV risk behaviours and were much less likely to report possessing healthcare benefits than previously-diagnosed individuals [63]. Newly-diagnosed individuals had a median CD4+ cell count of 432 cell/mL and less than one-quarter initiated ART prior to release from custody [63].

These findings add to evidence that jail-based opt-in or opt-out screening can identify previously-undiagnosed individuals, often earlier in their disease course [64,65]. For example, an evaluation of a programme of rapid HIV testing in three urban jails in settings with high HIV prevalence (Baltimore, Maryland; Philadelphia, PA; and the District of Columbia), reported that of more than 129,000 admissions over the study period, 41,612 (32%) completed voluntary rapid testing, resulting in 142 newly-identified infections (0.11% of all admissions, or 0.34% of all tests) [64]. These results come as a survey of prison and jails in the United States found that only a small proportion follow HIV testing and linkage guidelines [66]. Most notably, a study using blood sera collected during a compulsory screening programme for syphilis infection at intake into the state prison system in North Carolina found that the HIV prevalence among incarcerated individuals was approximately 1.45%. However, only 20 of 22,134 tests, or 0.09%, were HIV positive and previously undiagnosed, although the authors reported they did not have access to each individual’s entire testing history [67]. Thus, although voluntary HIV testing may be feasible even in busy correctional environments with high rates of inmate turnover, and detection of HIV among previously undiagnosed individuals is of obvious medical benefit to them, this evidence suggests expansion of jail- and prison-based HIV testing might not substantially reduce the number of individuals unaware of being HIV-positive [67].

IN-PRISON HIV TREATMENT AND LINKAGE TO POST-RELEASE CASE

In theory, correctional environments offer many advantages for the delivery of ART to HIV-positive individuals, including very low loss to follow-up, assured dosing schedules, and lower levels of availability of substances of misuse. Thus, controlled trials of directly-observed therapy among prisoners have demonstrated very high levels of optimal treatment outcomes [68,69]. Most notably, a recent retrospective cohort study of 882 HIV-infected and ART-initiated individuals incarcerated in the Connecticut correctional system found that while approximately 30% entered custody with plasma VL levels < 400 copies/mL, at exit, 70% of individuals were virally suppressed [70]. Although studies on HIV treatment patterns among incarcerated individuals in low-income settings are rare, a recent analysis of 148 individuals incarcerated in South Africa reported that almost three-quarters had plasma VL < 400 copies/mL after 96 weeks on ART [71].

Unfortunately, the clinical benefits of in-prison treatment for many are lost as individuals discontinue treatment upon release from custody. Most notably, only 5% of released prisoners filled a prescription for ART in the first 10 days following release from custody, thus avoiding treatment interruption, in one study from the prison system in Texas [72]. Observational studies of individuals transitioning from secure custody to community settings have identified a variety of barriers to retaining engagement in HIV/AIDS care, including poor or non-existent pre-release planning [66]; systematic removal of healthcare benefits from incarcerated individuals [73]; high-intensity drug and alcohol use in the immediate post-release period [74]; and poor housing status [75]. These barriers, along with sub-optimal HIV/AIDS treatment capacity in some correctional settings and a desire by some prisoners to conceal their serostatus, likely account for much of the observed association between incarceration and discontinuation or non-adherence to prescribed ART [76–78] and plasma VL rebound [24,79].

Recent studies among community-recruited groups of individuals engaged in HIV/AIDS treatment and care with substantial rates of incarceration confirm these dynamics. Studies among women living with HIV as well as illicit drug users identified associations between exposure to correctional settings and sub-optimal HIV/AIDS outcomes, including shorter time to AIDS [80], loss of virologic control [81] and dropping out of HIV care [82]. Although these studies were from settings in the United States, a similar dynamic was observed in a notable study from Ukraine [83]. Among 97 newly-released HIV-positive individuals, short-term detention by police was common and contributed directly to inability to access ART and prescribed opioid substitution therapy [83].

This study is one of a number to highlight the important role played by evidence-based treatment for substance use disorders for HIV-positive individuals engaged in the criminal justice system [84,85]. In one of the first studies on HIV-infected opioid-dependent individuals in Asia transitioning from correctional settings, optimal dosing with methadone (>80mg/day) was strongly associated with greater retention in treatment at 12 months [85]. Similarly, retention on treatment with buprenorphine/naloxone treatment was associated with maximal viral suppression among 94 opioid-dependent individuals released from custody in New England, United States [84]. An controlled trial of extended-release naltrexone for alcohol-dependent HIV-positive individuals released from custody currently underway may also provide evidence tom improve HIV treatment outcomes post-release [86].

As with HIV testing, the EnhanceLink study delivered some important insights into the factors associated with linking incarcerated individuals to care [75,87–96]. Conducted in 20 penal institutions in 10 cities in the United States between 2008 and 2011, the study enrolled 867 incarcerated individuals and followed them longitudinally in the first six months after release from custody. Activities associated with enhanced HIV linkage included case management programs, counseling, education and social support [87]. Of the 867 individuals, 34% were lost to care. Successful linkage to HIV/AIDS care within the first 30 days was associated with HIV/medication education and stable housing [75]. Services associated with staying in sustained care over six months were in-prison discharge planning and disease management sessions (both p < 0.05); and, after release, HIV education, a needs assessment, and transportation [88]. Although engagement in the HIV cascade of care improved for individuals at six months post-release (Figure 1), women were less likely to experience these benefits [87]. For example, while women and men did not differ in levels of viral suppression at baseline (26 vs. 25%; p = 0.91), women were significantly less likely to exhibit viral suppression after six months (18 vs. 30%, p < 0.001). Among all individuals, 26% had plasma VL < 400 copies/mL at six months, associated with attending a meeting with an HIV care provider within 30 days of release, after adjustment for achieving viral suppression in jail [97].

Figure 1.

Prevalences of various measures of engagement in HIV/AIDS treatment and care among individuals enrolled in the EnhanceLink study, stratified by gender, during incarceration (a) and six-months after release from incarceration (b), as reported by Meyer et al., [87]

The ultimate step on the HIV cascade of care is achieving non-detectable plasma VL through effective engagement in HIV treatment and care [98]. Unfortunately, few studies have investigated important virologic outcomes among criminal justice-involved populations, including community-level plasma VL, molecular cluster analysis or viral resistance to antiretrovirals. One study from Vancouver, Canada, using data from a long-running cohort of HIV-positive illicit drug users recruited from community settings found that incarceration was associated with detectable VL and, during periods of detectable VL, incarceration was also associated with sharing contaminated syringes [99], pointing to the need for effective ART among incarcerated individuals. In a notable study from Brazil, a survey of 1167 incarcerated individuals found that 110 (6.6%) were HIV-positive. A molecular phylogeny inferred from the analysis of viral RNA sequences from 40 individuals did not find any evidence of between-prisoner transmission; however, analysis of sequences from the pol gene evidenced high rates of HIV primary and secondary resistance [100]. In contrast, a survey of 367 viral sequences from individuals incarcerated in the state prison system of North Carolina found moderate levels of drug resistance, comparable to non-incarcerated populations [101].

CONCLUSION

In this review, we summarized the recent evidence regarding the effect of incarceration and correctional systems on patterns of HIV testing and treatment. Studies from a variety of settings confirmed that HIV prevalence remains higher among incarcerated populations as compared to analogous non-incarcerated groups [33–36,39] and often co-occurs with other infectious diseases, especially viral hepatitis and tuberculosis [32,34–36,39]. Recent evidence from the United States indicates that voluntary HIV rapid testing within jail and prison settings is feasible and effective [62,63]. However, the number of undiagnosed individuals identified in some surveys was low, indicating that voluntary testing in correctional populations might not yield enough new cases to have a substantial impact on the HIV cascade of care in the United States [64,67]. A large observational study of testing and linkage to care in 10 cities in the United States demonstrated improved HIV treatment outcomes at six-months post release [87]. Evidence on viral resistance to antiretrovirals among incarcerated populations was mixed [100,101], however incarceration was associated with elevated VL and associated risk factors among illicit drug users in Vancouver, Canada [99].

As many countries look to scale-up TasP campaigns to engage HIV-positive individuals in care, lower community-level plasma VL and curb the incidence of new infections, correctional populations are attractive targets. Unfortunately, there remain unanswered questions about whether the entrenched barriers to HIV/AIDS treatment and care in penal settings will compromise TasP campaigns, especially among people who use illicit drugs, sex workers, members of sexual/gender minorities and those criminalized for HIV non-disclosure [27]. Fortunately, there is emerging evidence that individuals from these traditionally hard-to-reach groups can be successfully engaged in HIV/AIDS treatment and care through TasP-based campaigns in community settings [102,103]. Although observational studies of recent programmatic interventions among prisoners have identified modest gains in HIV/AIDS testing and linkage to care, future research should evaluate the effect of structural changes to correctional systems. These reforms might include transferring responsibility for in-prison HIV/AIDS treatment to public health authorities and diverting people living with HIV/AIDS from custodial settings [9].

Four decades into the HIV/AIDS pandemic, all available evidence continues to describe persistent deficits and iniquities in HIV/AIDS prevention and treatment among incarcerated individuals. Through incarceration, many individuals with elevated risks of HIV disease progression continue to be exposed to environments with sub-optimal HIV/AIDS treatment and care. In light of the fundamental right of incarcerated individuals living with HIV/AIDS to receive healthcare equivalent to that delivered in analogous non-penal settings [104], the increased global adoption of universal access to ART underscores the urgent need to guarantee the benefits of HIV/AIDS treatment for all people, including those living in custody.