HIV and corrections: every statistic tells a story.

Everyone has probably heard the old saying that anyone can lie with statistics. Or, as it is often stated, there are "lies, damn lies, and statistics" (variably attributed to Benjamin Disraeli or Mark Twain). Anybody can try to lie with statistics, but he or she will find it difficult to lie to an informed audience. These days it is popular to refer to such an informed audience as "literate." There is media literacy, product literacy and health literacy, to name a few. This article will start with one of the fundamentals of research literacy to help readers accurately respond to questions about the scope of HIV found in correctional institutions and correctional populations. This will assist corrections professionals in explaining to various audiences why so much different information is circulating about HIV in corrections.

There are three primary ways of examining the prevalence of HIV within correctional systems in the published literature: administrative data, seroprevalence data and the HIV/AIDS Reporting System (HARS). While all three are potentially related to each other, they are often utilized independently of each other. This can create some confusion as different groups refer to different statistics--often the statistics that best make their point.

It is important to mention one quick note about terminology that sometimes confuses criminal justice and public health professionals when they talk about recorded incidence of disease or injury. When criminal justice personnel talk about reported crimes, incidents of prison rape, etc., the term "monitoring" is generally utilized. When public health personnel talk about the incidence of disease or injury, the term "surveillance" is used. While there are some differences between the two, they generally mean the same thing--a systematic recording of the incidence of a particular event or activity of interest over a period of time.

Readers of Corrections Today may be most familiar with the administrative approach taken by the Bureau of Justice Statistics to determining the prevalence of HIV within inmate populations. A count of HIV-infected inmates known to prison officials and deaths attributed to AIDS among inmates are provided in year-end reports from state prison systems, the Federal Bureau of Prisons and the District of Columbia. (1) The data are part of the BJS National Prisoner Statistics series. They represent a census of all administratively known HIV-infected inmates housed within a particular system on December 31 of a given year. Maruschak (2) provided the questions asked in the survey:

  On Dec. 31, 2000, how many inmates under your custody were:

  Asymptomatic HIV positive
  Male___Female___
  Infected with lesser forms of
  symptomatic HIV disease
  Male___Female___
  Confirmed to have AIDS
  Male___Female___

BJS asks these questions at the corrections system level, rather than of individual prisons (the organizational level). The testing data are consistent with those collected in the 1996-1997 Update: HIV, STD, and TB in Correctional Facilities (3) study and the recently completed survey by the National Institute of Justice (NIJ) and the Centers for Disease Control and Prevention (CDC) by Hammett and his colleagues at Abt Associates.

Perhaps the most comprehensive HIV surveillance information comes from seroprevalence studies. In such studies, blood taken from inmates is tested for the presence of HIV antibodies. The proportion of inmates testing positive for the virus is reported. Seroprevalence studies can be "linked" or "blinded." The majority of the seroprevalence studies in the published literature are blinded, meaning that the researchers or others were unable to identify the inmate from whom the samples were taken. In states with mandatory testing for HIV, the samples would be linked to the inmate's identity. Other systems use a sample of blood taken for various purposes, unfortunately, interpretation of their results are limited by the sampling procedure employed.

The HIV/AIDS Reporting System (HARS) reports new "officially" diagnosed cases of HIV/AIDS. HARS is a "named" diagnosis system, where the names of positive persons are reported to a central state-based data set with appropriate privacy provisions and, in turn, reported to the CDC. Such diagnoses may come from testing new intakes or periodic testing of incarcerated inmates. It should be kept in mind that HARS data count new, initial clinical diagnoses of HIV infection. Name-based reporting has not been accepted in all states, though most states will soon participate, depending on available funding. As a result, earlier studies using HARS have reported on either the 25 states participating at that time (4) or a single participating state. (5)

As the CDC guidance on interpreting HARS data cautions, individuals may have utilized home testing kits or anonymous testing previously and those diagnoses would not appear in HARS data. Thus, one could be aware that one was living with HIV, but only "officially" appear in HARS data during correctional medical intake tests (or from testing in another clinical setting). Since the existing HARS data form (6) does not include jail or prison settings as either a facility "setting" or "type," accurate reporting is also dependent on "jail" or "prison" being specified in the "other" facility type data field. Future electronic versions of the HARS will include specific corrections data.

Each of the information approaches outlined here gives a different impression of the scope of HIV/AIDS among correctional populations. For example, the HARS data reported by Dean, et al. (7) showed that in the 25 states participating between the years 1994-2000, six percent of all newly identified HIV/AIDS cases were reported by correctional systems. Utilizing seroprevalence methods, the Maryland Department of Corrections has documented the fluctuations in HIV prevalence in the prison system topping out at nearly eight percent (7.9 percent) among imprisoned males during the late 1980s and early 1990s, with female inmates showing a 15.3 percent infection rate (on intake). The BJS annual reports present a national average of known cases of HIV, with state-level reports also provided. In 2003, the national level of HIV in prisons was 1.9 percent (2.8 percent among female inmates). These percentages are still more than three times the rate of HIV estimated in the non-prison population.

The take-home message here is that when discussing HIV (and many other controversial topics) in correctional populations or responding to questions about this topic, one needs to be clear about what data are being used. It is a different thing to say that 6 percent of new HIV cases are among inmates, than to say that 1.9 percent of state and federal prisoners are living with HIV--but both are accurate. Those two statements also cover very different time periods (1994-2000 vs. 2003), and areas (25 states vs. state and federal prisons). The image provided is quite different and may be used by different groups to promote their individual causes or interests.

Where one looks determines what one sees. How one looks at that area further focuses what one sees. Understanding how data are collected, what the limitations of data are and how one properly presents data, can help corrections professionals provide a more honest picture of the situation. If one has doubts about data being presented or what to present, he or she may consult in-house researchers. If one does not have in-house research staff, then he or she should return to the advice offered in Corrections Today's July 2005 article on practitioner-researcher partnerships (8) for some guidance on how to get the information one needs.

People attempt to structure reality with statistics every day. Presenting and interpreting data accurately will add credibility to the arguments of corrections professionals, researchers or legislators. Yes, one can lie with statistics, but it is very difficult to lie to people who know the truth, especially about statistics.

ENDNOTES

(1) Maruschak, L. 2004. HIV in Prisons, 2003. Washington, D.C.: Bureau of Justice Statistics. (NCJ 210344).

(2) Maruschak, L. 2002. HIV in prisons, 2000. Washington, D.C.: Bureau of Justice Statistics. (NCJ 196023)

(3) Hammett, T., P. Harmon and L. Maruschak. 1999. 1996-1997 update: HIV/AIDS, STDs, and TB in correctional facilities. Washington, D.C.: National Institute of Justice.

(4) Dean, H.D., A. Lansky and P.L. Fleming. 2002. HIV surveillance methods for the incarcerated population. AIDS Education and Prevention, 14(Supplement B):65-74.

(5) Krebs, C.P. and M. Simmons. 2002. Intraprison HIV transmission: An assessment of whether it occurs, how it occurs, and who is at risk. AIDS Education and Prevention, 14(Supplement B):53-64.

(6) Adult HIV/AIDS Confidential Case Report (OMB No. 0920-0573). Atlanta: Centers for Disease Control and Prevention.

(7) Op cit Dean, et al., 2002.

(8) Smith, L., R. Tewksbury, and RH Potter. 2005. Practitioner-researcher partnerships: Partnering for productivity. Corrections Today, 67(4): 106-107.

Roberto Hugh Potter, Ph.D., has been working in and around corrections since 1977 as a therapist, researcher, trainer, educator and program person. Kera Moseley, DrPH, is currently president of The Prometheus Group, a New Orleans based nonprofit organization, with more than 10 years of involvement with correctional health systems.