U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics HIV in Prisons, 2001-2010 Laura M. Maruschak, BJS Statistician September 2012, NCJ 238877 ------------------------------------------ This report is one in series. More recent editions may be available. To view a list of all reports in the series go to http://www.bjs.gov/index.cfm?ty=pbse&sid=7 This file is text without graphics and many of the tables. A Zip archive of the tables in this report in Comma- delimited format (CSV) and the full report including tables and graphs in PDF format are available at: http://www.bjs.gov/index.cfm?ty=pbdetail&iid=4452 ---------------------------------------------- At yearend 2010, state and federal prisons held a reported 20,093 inmates who had HIV or AIDS, down from 20,880 at yearend 2009. As a result of this decline, the estimated rate of HIV/AIDS among prisoners in custody dropped from 151 HIV/AIDS cases per 10,000 inmates in 2009 to 146 per 10,000 in 2010. The number of state and federal inmates who died from AIDS-related causes also declined between 2009 and 2010, from 94 to 72 deaths. During the same period, the rate of AIDS-related deaths among inmates with HIV/AIDS declined from 47 deaths per 10,000 inmates to 38 per 10,000. The 1-year declines between 2009 and 2010 were consistent with trends over the past decade. Between 2001 and 2010, the estimated rate of HIV/AIDS among state and federal prison inmates declined an average of 3% each year (figure 1). During the same period, the AIDS-related death rate declined an average of 13% per year among inmates with HIV/ AIDS and 16% among all prison inmates. ************************************************** ***************** Highlights ***************** * The rate of HIV/AIDS among state and federal prison inmates declined from 194 cases per 10,000 inmates in 2001 to 146 per 10,000 at yearend 2010. * AIDS-related deaths among all state and federal prison inmates declined an average 16% per year between 2001 and 2010, from 24 deaths per 100,000 inmates in 2001 to 5 per 100,000 in 2010. * Among state and federal inmates with HIV/AIDS, the rate of AIDS-related deaths dropped significantly, from 134 deaths per 10,000 inmates with HIV/AIDS in 2001 to 38 per 10,000 in 2010. * When grouped by 2010 yearend custody populations, the declining rate of HIV/AIDS in small, medium, and large state prisons had the same 3% average annual decline in the national rate between 2001 and 2010. * Between 2001 and 2010, the average annual decline of 16% in the national AIDS mortality rate mirrored the decline in small (down 12%), medium (down 17%), and large (down 19%) state prison populations. * Rate of HIV/AIDS cases and AIDS-related deaths declined across all sizes of prison populations. * In 2009, the AIDS mortality rate among state prison inmates (6 per 100,000) fell below the rate for the U.S. general population (7 per 100,000). * AIDS-related deaths declined from 89 in 2009 to 69 in 2010 among males, from 70 to 43 among all black non- Hispanics, and from 87 to 60 among all inmates age 35 or older. **************************************************** Findings in this report are based on data collected from 50 states and Federal Bureau of Prisons through the Bureau of Justice Statistics’ (BJS) National Prisoner Statistics (NPS-1) (which annually collects data on prisoner counts, characteristics, admissions, and releases) and Deaths in Custody Reporting Program (DCRP) (which collects incident- level data on cause of inmate death and characteristics of those who died). This bulletin provides data on national trends in the estimated rates of HIV/AIDS cases and AIDS-related deaths among state and federal prison inmates for 2001 through 2010. It includes age-specific comparisons between state prison inmates and the U.S. general population from 2001 through 2009. It also provides jurisdiction-level data for 2009 and 2010 on the number of state and federal inmates who were infected with HIV/AIDS, HIV/AIDS cases by sex, confirmed AIDS cases, and AIDS- related deaths that occurred in state prison by demographics of inmates who died. ********************************************** Rates of HIV/AIDS and AIDS-related deaths in prisons have declined steadily since 2001 ********************************************** Between 2001 to 2010, the estimated number of inmates with HIV/AIDS declined by 16%, and the number of AIDS-related deaths in prison declined by 77% (not shown in table) resulting in declines in the rates of HIV/AIDS and AIDS- related deaths among inmates with HIV/AIDS. At yearend 2001, the estimated rate of HIV/AIDS among state and federal prison inmates was 194 HIV/AIDS cases per 10,000 inmates. By yearend 2010, the estimated rate was 146 cases per 10,000. Among the total inmate population, the rate of AIDS-related deaths declined from 24 per 100,000 inmates in 2001 to 5 per 100,000 in 2010. Among the inmate population with HIV/AIDS, the rate declined from 134 AIDS-related deaths per 10,000 inmates in 2001 to 38 per 10,000 in 2010. (See Methodology for discussion of calculation of rates.) ************************************************ Rates of HIV/AIDS cases and AIDS-related deaths declined across all sizes of prison populations ************************************************ Regardless of the size of the state prison population, as measured by the number of inmates in custody at yearend 2010, trends in the rates of HIV/AIDS cases and AIDS- related deaths in state prisons followed the overall decline of the national rate. States were grouped into three categories based on the size of their 2010 yearend custody populations. The smallest group included states with 24,999 or fewer inmates, the medium group included states with 25,000 to 49,999 inmates, and the largest group included states with 50,000 or more inmates. (See Methodology for a list of states included in each group.) From 2001 to 2010, the rate of HIV/AIDS declined from 275 to 211 HIV/AIDS cases per 10,000 inmates in the largest group, from 144 to 110 cases per 10,000 inmates in the medium group, and from 174 to 130 cases per 10,000 inmates in the smallest group (figure 2). Over the 9-year period, all groups, regardless of population size, saw an average annual decline of about 3% in the rate of HIV/AIDS cases. Between 2001 to 2010, the rate of AIDS-related deaths declined from 28 to 4 per 100,000 inmates in the large group of states, from 24 to 4 per 100,000 inmates in the medium group, and from 22 to 8 per 100,000 inmates in the small group. The smallest average annual decline in the rate of AIDS-related deaths was observed in small states (down 12%), followed by medium (down 17%) and large states (down 19%) (figure 3). ********************************** The rate of AIDS-related deaths for state prison inmates dropped below the rate for the U.S. general population ********************************** To make AIDS-related death rate comparisons to the general population, individual-level data from DCRP were used to calculate age-specific rates of AIDS-related deaths among those in state prison and AIDS-related deaths as a percentage of all deaths in state prison. (See Methodology for calculation of age-specific death rates and for sources of national mortality data.) The rate of AIDS-related deaths in state prisons among inmates ages 15 to 54 declined sharply between 2001 and 2009, compared to the more modest decline observed among the same age group in the U.S. general population. As a result, the AIDS-related death rate in state prisons fell below the rate in the U.S. general population in 2009. Between 2001 and 2009, the AIDS-related death rate among state prisoners ages 15 to 54 declined from 22 deaths per 100,000 inmates to 6 per 100,000, while the rate among that age group in the general population declined from 9 per 100,000 to 7 per 100,000 (figure 4). As a percentage of all inmate deaths, AIDS-related deaths among state prisoners ages 15 to 54 declined from 13.4% to 3.6% between 2001 and 2009. The AIDS death rate, as a percentage of all deaths among the general population ages 15 to 54, remained relatively stable at about 4% (figure 5). ******************************* Selected findings from 2009 and 2010 data ******************************* * A reported 20,093 inmates with HIV/AIDS were in custody in state or federal prison at yearend 2010, down from 20,880 at yearend 2009 (table 1). * The reported number of state inmates with HIV/AIDS decreased from 19,290 in 2009 to 18,515 in 2010. The number of federal inmates with HIV/AIDS also declined, from 1,590 in 2009 to 1,578 in 2010. * California, Florida, New York, and Texas each reported holding more than 1,000 inmates with HIV/AIDS at yearend 2010. These states held 51% (9,492) of all state prisoners with HIV/AIDS. * Among state and federal inmates with HIV/AIDS at yearend 2010, 18,337 were male and 1,756 were female, compared to 19,027 male and 1,853 female inmates at yearend 2009 (table 2). * Among jurisdictions reporting the number of inmates with confirmed AIDS in both 2009 and 2010 (37 states and the Bureau of Federal Prisons), 4,344 inmates had confirmed AIDS in 2009 and 3,913 in 2010—a decrease of 431 cases (table 3).***Footnote * BJS no longer estimates for jurisdictions not reporting confirmed AIDS due to insufficient data in recent years. See previous HIV reports for estimated confirmed AIDS cases and methodologies.*** * After an increase of five AIDS-related deaths in state prisons between 2008 and 2009 (89 to 94), the number fell to 72 deaths in 2010. Seven federal inmates died from AIDS- related causes, which was unchanged from 2009 (table 4). * AIDS-related deaths declined from 89 in 2009 to 69 in 2010 among males, from 70 to 43 among black non-Hispanics, and from 87 to 60 among inmates ages 35 or older (table 5). ************** Methodology ************** **************************** National Prisoner Statistics ***************************** The federal government began collecting National Prisoner Statistics (NPS-1), which primarily measures prison population movement, in 1926. The series comprises reports from the Departments of Corrections in the 50 states and the Federal Bureau of Prisons. NPS-1 includes yearend counts of prisoners by jurisdiction, sex, race, Hispanic/Latino origin, and admissions and releases during the year. In 1991, the Bureau of Justice Statistics (BJS) began using the NPS-1 to collect data on HIV and confirmed AIDS cases in prisons. BJS respondents indicated the circumstances under which inmates were tested for HIV and reported the number of inmates with HIV/AIDS in their custody. Until 2007, the respondents provided the number of deaths occurring among inmates under their jurisdiction, including those from AIDS-related causes. In this report, yearend and midyear custody counts as well as midyear jurisdiction counts were used when calculating the rates of HIV/AIDS cases and AIDS related death in prison. Jurisdiction refers to the legal authority of state or federal correctional officials over a prisoner regardless of where the prisoner is held. Jurisdiction counts include inmates in private facilities. To have custody of a prisoner, a state or federal prison system must hold that inmate in one of its facilities. Custody counts do not include inmates held in private facilities. *************************** Measuring deaths in prison *************************** From 1991 to 2006, BJS used NPS-1 to collect data on the number of deaths among inmates under the jurisdiction of each prison system by cause, including AIDS-related deaths. In 2001, BJS implemented the Deaths in Custody Reporting Program (DCRP), which collects information describing events and characteristics of each death excluding executions, occurring among inmates in custody of state prisons. Deaths reported through DCRP include those that occurred in private facilities. Prior to 2001, BJS relied solely on data in the NPS-1 to report on AIDS-related deaths. From 2001 to 2006, BJS estimated the number of AIDS-related deaths using both data sources. The greater of the two numbers of deaths reported was used to calculate the totals for those years. After 2006, BJS stopped collecting detailed data on deaths in NPS-1 and relied solely on the DCRP data. Whether using reported or estimated data from multiple data sources, a downward trend in the number of AIDS-related deaths has been observed (figure 6). *************************************************** Rate of HIV/AIDS and AIDS-related deaths in prison *************************************************** To calculate the estimated rate of HIV/AIDS among inmates in custody of state and federal prison, the estimated yearend count of HIV/AIDS cases was divided by the yearend custody population for each year. To calculate the rate of AIDS-related deaths in prison per inmate population for 2001 to 2006, the number of AIDS- related deaths estimated in each year from the NPS-1 and DCRP were divided by the midyear jurisdiction population for 2001 to 2006. The midyear jurisdiction population was the denominator because deaths in the NPS-1 were reported among those inmates under the jurisdiction of each prison system. For 2007 to 2010, the number of AIDS-related deaths reported in DCRP was divided by the midyear custody count (including inmates held in private facilities) in state prisons plus the midyear jurisdiction count in federal prisons. The midyear state custody population was the denominator because deaths reported in DCRP were among those inmates in custody of state prisons. The denominator included the midyear federal jurisdiction population because deaths reported in the federal prisons were among those under federal jurisdiction. To calculate the rate of AIDS-related deaths among those at risk for dying from HIV/AIDS (those estimated to have HIV/AIDS), the number of AIDS-related deaths reported in each year was divided by an estimate of the average daily population of inmates in prison with HIV/AIDS. The average daily population of those with HIV/AIDS for a given year was the average of the most recent year’s estimated yearend count of those with HIV/AIDS and the estimated yearend count from the previous year. For comparisons to the general population ages 15 to 54, data on deaths reported in the DCRP between 2001 and 2010 were used solely for the calculations of AIDS-related death rates among those 54 or younger in the state prison population. ************************************* Grouping of states by yearend custody ************************************* To assess the trends in the rates of HIV/AIDS and AIDS- related deaths among inmates, state prison populations were grouped into categories of small, medium, and large. Groups were based on the size of the 2010 yearend custody populations. Included in the small group were states with fewer than 25,000 inmates, specifically North Dakota, Vermont, Montana, Wyoming, Maine, New Hampshire, Rhode Island, Hawaii, South Dakota, New Mexico, Alaska, Nebraska, Idaho, West Virginia, Utah, Delaware, Kansas, Minnesota, Iowa, Massachusetts, Mississippi, Kentucky, Nevada, Oregon, Arkansas, Tennessee, Washington, Connecticut, Oklahoma, Colorado, Louisiana, New Jersey, Wisconsin, Maryland, South Carolina, and Indiana. Included in the medium group were states with 25,000 to 49,999 inmates, specifically Alabama, Virginia, Missouri, Arizona, North Carolina, Michigan, Pennsylvania, Illinois, Ohio, and Georgia. Included in the large group were states with 50,000 or more inmates, specifically New York, Florida, Texas, and California. ******************************** AIDS-related deaths in the U.S. general population ******************************** The number of AIDS-related deaths for persons ages 15 to 54 was based on the Center for Disease Control’s (CDC), HIV/AIDS Surveillance Report, yearend editions (Vols. 17- 22). Deaths in the U.S. population for persons ages 15 to 54 were taken from the CDC, National Center for Health Statistics’ (NCHS) Monthly Vital Statistics Report, Vol. 45, No. 11(S). Also, deaths were taken from the NCHS’s National Vital Statistics Report, Vol. 52, No. 3; Vol. 53, No. 5; Vol. 53, No.15; Vol. 54, No. 19; Vol. 57, No. 14; Vol. 58, No.19; Vol. 59, No.10; and Vol. 60, No.4. For 2005, U.S. general population deaths can be found in the NCHS’s Deaths: Preliminary Data for 2005. For 2001 to 2009, AIDS-related deaths were calculated as a percentage of all deaths among persons ages 15 to 54 in the U.S. general population. The national estimate of AIDS-related deaths of persons ages 15 to 54 minus AIDS-related deaths of persons ages 15 to 54 in state prisons was divided by the national mortality estimates of persons ages 15 to 54 minus total deaths of persons ages 15 to 54 in state prisons. For 2001 to 2009, the rates of AIDS-related deaths in the U.S. general population were calculated by taking the difference of the national estimate of AIDS-related deaths for persons ages 15 to 54 and the number of AIDS-related deaths for state inmates ages 15 to 54, and dividing it by the U.S. general population estimate of those 15 to 54 minus the state prison population ages 15 to 54. ******************************** Estimating for missing data on HIV/AIDS cases ******************************** The NPS-1 database for 2001 to 2010 was missing data on the number of inmates with HIV/AIDS in state and federal prison in six states that did not report the number of cases for 1 or more years. The missing data ranged from a state missing 1 year of data to another state where data for all 10 years (2001 through 2010) were not reported. Estimates do not include counts from the state missing data in all years and rates were calculated using a denominator that excluded this state prison population thereby assuming that the rate in this state was equal to the national rate. The remaining four states included two missing data for 2 years and two missing data for 5 years. The five states that reported missing data (excluding the one state that did not report data in any year) reported HIV/AIDS cases in 2001 and, in that year, comprised 3.4% of the total number of reported HIV/AIDS cases in state correctional facilities (759 out of 22,553 total cases). In 2008, the five states comprised a similar share of the total number of reported HIV/AIDS cases in state correctional facilities (3.3% or 665 out of 20,073). STATA, a statistical software package, was used to linearly interpolate the number of HIV/AIDS cases that were not reported. This method estimates the missing value by using observed data values that are closest in time to the missing data value. The missing value is assumed to fall on a straight line plotted between two known data points. The ipolate function in STATA interpolates the missing value in instances where observed values appear before and after the missing value to be estimated. The epolate function extrapolates the missing value where observed values appear only before or after the missing value that is to be estimated. For example, the interpolated value of HIV/AIDS cases (y) in a given state (x) is computed by using the closest known values for reported HIV/AIDS cases. If in 2001 (x1) there were 15 (y1) cases and in 2003 (x2) there were 10 (y2) known cases, the interpolated value for 2002 is computed using the following formula: y=y1+(x-x1)(y2-y1)/(x2-x1). In this example, the interpolated number of HIV/AIDS cases for that state in 2002 would be an estimated 12.5. In some cases, data were missing for multiple consecutive years. Data were interpolated using a multi-stage process. For example, in a state missing 2006 and 2007 data, the interpolated value for 2006 was derived by using observed values from 2005 and 2008. The interpolated value for 2007, in turn, was derived by using the interpolated value from 2006 and the observed value from 2008. In one state it was necessary to extrapolate the missing value as it occurred in the last year of the series (2010). This was computed by using a linear equation to predict values beyond the observed values using the epolate function in STATA 10. A linear equation is derived from the observed values for 2008 and 2009 (y=15x-2) and then used to predict the value for 2010 where x=3 to determine the third year value (2010) in the linear trend from 2008. The growth rate from 2009 to 2010 is assumed to be the same as from 2008 to 2009 (x=15) and the formula is solved as 43=(45)-2. The extrapolated value for 2010 is 43 HIV/AIDS cases. State-specific imputed values are not included in the report tables, but were used to calculate national rates of HIV/AIDS cases in prison and rates by size of state prison custody population. For each state with imputed values, both upper and lower estimates were produced around the imputed values. These upper and lower limits were calculated for each state’s imputed value or values by adding the largest 1-year increase in HIV/AIDS cases and subtracting the largest 1-year decrease from the imputed value or values. New national rates based on upper and lower limits did not vary more than plus or minus 1 HIV/AIDS case per 10,000 inmates for any given year (figure 7). Year-to-year declines were not affected by variations in reporting. ******************************** The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. James P. Lynch is director. This report was written by Laura M. Maruschak. Todd Minton verified the report. Fan Zhan and Kyle Harbecek, BJS interns, provided statistical assistance. The U.S. Census Bureau, U.S. Department of Commerce, carried out the data collection and processing. Catherine Bird and Jill Thomas edited the report, and Barbara Quinn produced the report under the supervision of Doris J. James. September 2012, NCJ 238877 ************************* ************************* Office of Justice Programs Innovation • Partnerships • Safer Neighborhoods www.ojp.usdoj.gov ************************** 8/30/2012/JER/11:45am