Measuring Correctional Experience to Inform Development of HIV, Sexually Transmitted Infection, and Substance Use Interventions for Incarcerated Black Men Who Have Sex with Men

The systematic review of corrections-based interventions for Black men who have sex with men, by Harawa et al. in this issue of AJPH (p. S304), illustrates the paucity, in terms of quality and quantity, of existing interventions to improve health status. Because of disproportional racial minority confinement, men who are Black are more likely to experience incarceration than are Whites. Data on sexual orientation are not routinely collected in correctional settings; no evidence proves that minority sexual orientation heightens or lessens this risk. Nonetheless, public health investigators have not focused on the common setting of a correctional facility when designing interventions to improve access to HIV, sexually transmitted infection, and substance use services for Black men. Furthermore, most of the correctional studies examined in the review were not tailored for racial minorities. Many only sought to increase health knowledge rather than measure actual change in behavior, with or without biological markers. Study design was often weak. The handful of randomized controlled trials to improve engagement in HIV care showed equivocal results. Only three of 56 correctional studies focused primarily on sexual minorities; only one concentrated on Black men who have sex with men. None made transgender individuals a chief concern. The authors conclude that perhaps more modeling exercises could fill these voids in the research agendas of the investigators in the United States. In our opinion, this shortcut around empirical research is problematic, given that lifetime prevalence of incarceration either in a jail (a short-term correctional facility) or a prison (which houses persons serving felony sentences) has reached epidemic levels among Black men.

As Harawa et al. point out, prison growth in the United States has skyrocketed over the three decades of the HIV epidemic. This leads to the question of how many Americans the epidemic of incarceration has touched in total. To date, when discussing incarceration over a lifetime, researchers concentrate on imprisonment. For example, Western and Pettit estimated that the lifetime prevalence of imprisonment among recent birth cohorts of Black men without a high school education was 68%.¹ However, in addition to prisons, jails can also disrupt access to health care. A more comprehensive measure of the extent to which incarceration disrupts the lives of Black men would be the lifetime prevalence of stays in both jail and prison.

The General Social Survey (GSS), a national household survey of public opinion whose methods have been described elsewhere,² may offer an answer regarding lifetime prevalence of incarceration. In 2012, the GSS, for the first and only time to our knowledge, questioned 1974 respondents on prior experience with incarceration using a very comprehensive definition: “Have you ever spent any time in prison or jail?” To obtain a representative sampling of Americans, this data set needs to be supplemented with data from other nationally representative sources to compensate for those missed in a household survey because of other living experiences, including incarceration. Sampling of US households in the GSS excludes homeless persons and those living in group quarters (e.g., college residence halls in addition to correctional facilities).² Household-based surveys underestimate the true prevalence of incarceration exposure because those currently incarcerated are missed. Furthermore, those recently incarcerated are also likely excluded from household-based estimates, given well-acknowledged barriers to securing stable housing after incarceration.³ A one-day count of incarcerated persons is the minimum number of persons to add to counts of community-dwelling adults; counting persons incarcerated over the span of one year more likely represents the true number of adults missed by GSS.

Thus, to account for those not capable of being surveyed in 2012, one can supplement the GSS statistics using data from surveys and reports that serve as primary sources of statistical information for their respective sampling frames. Nationally representative sources are outlined in the Appendix (available as a supplement to the online version of this article at http://www.ajph.org). As shown in Figure 1, we estimate lifetime prevalence of incarceration for all US adults, as well as for subgroups: men, Black men, and Black men with less than a high school education. Because persons in recent birth cohorts have not accrued much time in the crime-prone age years, we focused on Black men who were born between 1965 and 1974 who were not high school graduates. To characterize the cohort, we used weights provided by the GSS to reflect nonrespondent subsampling, the number of adults in each household, and differential nonresponse across areas.² We calculated point estimates and confidence intervals for proportions by using GSS data. The lifetime prevalence of incarceration among housed US men in 2012 was 20.8% (95% confidence interval = 17.8%, 24.2%; Figure 1).

FIGURE 1—

Lifetime Prevalence of Incarceration Experience, 2012, Among US Adults (a) Living in households only, (b) Including Persons Incarcerated, Single-Day Count, and (c) Including Persons Ever Incarcerated in 2012, Year-Long Count

Source. NORC General Social Survey (GSS), Bureau of Justice Statistics (BJS), and the Current Population Survey

Note. LTHS = Less than high school educational attainment. The population estimates include persons living in households alone and combined estimates for persons living in households and those incarcerated in 2012. The 95% confidence intervals are shown for GSS household estimates only; combined household/incarceration estimates are supplemented with BJS enumerations. In panel b, GSS household incarceration estimates are supplemented by BJS enumerations of the number of persons incarcerated on a single day in 2012. These figures likely underestimate the lifetime prevalence of incarceration. In panel c, GSS household incarceration estimates are supplemented by BJS enumerations of the number of unique persons ever incarcerated in 2012 to reflect that persons recently released from correctional facilities may not have established a stable household prone to be targeted for GSS sampling.

^aRounded approximations for US adult population based on American Community Survey (2012) estimates and provide rough approximations for the base population count of each proportion: all women and men = 240 000 000; all men = 114 000 000; All Black men = 13 000 000; Black men LTHS all ages = 3 000 000; and Black men LTHS born 1965–1975 = 500 000.

While one quarter of men born before 1974 were ever incarcerated, only 16% of men born in 1975 or after had yet experienced confinement (P = .06). More than one third of men identifying as Black had ever been incarcerated, significantly higher than the rate among men identifying as White or other race (18.9%; P = .01; Figure 1). We estimate that as of 2012, a total of 42 million US adults, or one in six, had ever been incarcerated. Among subpopulations facing higher incarceration rates (men, Blacks, low education, recent birth cohort), the proportion ever incarcerated progressively increased with added intersectionality, such that in 2012, a Black man born between 1965 and 1974 without a high school diploma would have had an 80.9% chance of ever being confined (Figure 1; see the Appendix for sensitivity analysis).

Incarceration has reached epidemic proportions, a fact not apparent if examining only one-day counts of persons in prison or jail. Because of the stigmatizing nature of incarceration, a patient might not share this history during clinical encounters. Many lives, especially Black lives, have been affected by practices such as heightened surveillance in poorer neighborhoods. In the same 2012 GSS, the likelihood of marriage for men was 69.6%, on par with studies by the US Census Bureau.⁴ Comparing the 80.9% likelihood of incarceration to that of marriage, either to women or other men, indicates how widespread the experience of incarceration has become for Black men with less than a high school education.

Not only is incarceration common, but also the postrelease period is associated with poor health status. For example, nationwide 86% of persons with HIV in clinical care have viral suppression.⁵ While viral suppression during long incarcerations is the norm, among persons leaving jail, a recent meta-analysis showed that only 21% maintain viral suppression.⁶ The systematic review by Harawa and et al. illustrates that little effort has been made to study how to conduct public health interventions with persons going in and out of correctional facilities. Funding has been sparse. Until we move toward decarceration—a recent trend that has stalled in the current political climate—we will need better interventions to improve the settings in which our society often places Black men who have sex with men.