Racial/Ethnic Disparities in Injection Drug Use in Large US Metropolitan Areas HANNAH COOPER, SCD, SAMUEL R. FRIEDMAN, PHD, BARBARA TEMPALSKI, MA, MPH, RISA FRIEDMAN, MPH, AND MARIE KEEM, MED
PURPOSE: Because blacks and Latinos bear a disproportionate burden of injection-related health problems compared with whites, we sought to describe black/white and Latino/white disparities in injecting drugs in 94 US metropolitan statistical areas (MSAs) in 1998. METHODS: Using US Census data and three databases documenting injectors’ use of different healthcare services (drug treatment, HIV counseling and testing, and AIDS diagnoses), we calculated database-specific black/white and Latino/white disparities in injecting in each MSA and created an index of black/white and Latino/white disparities by averaging data across the three databases. RESULTS: The median black/white injecting disparity in the MSAs ranged from 1.4 to 3.7 across the three databases; corresponding median Latino/white injecting disparities ranged from 1.0 to 1.1. Median black/white and Latino/white index disparity values were 2.6 and 1.0, respectively. CONCLUSIONS: Although whites were the majority of injectors in most MSAs, database-specific and index black/white disparity scores indicate that blacks were more likely to inject than whites. While database-specific and index disparity scores indicate that Latinos and whites had similar injecting rates, they also revealed considerable variation in disparities across MSAs. Future research should investigate these disparities’ causes, including racial/ethnic inequality and discrimination, and study their contributions to the disproportionate burden of injection-related health problems borne by blacks and Latinos. Ann Epidemiol 2005;15:326–334. Ó 2005 Elsevier Inc. All rights reserved. KEY WORDS: Blacks, Hispanics, Substance Abuse/Intravenous, Substance-related Disorders, HIV, US/ Epidemiology.
INTRODUCTION Public health research and surveillance systems in the US consistently find that blacks and Latinos bear disproportionate burdens of injection-related health problems compared with whites, including injection-related HIV/ AIDS and overdose morbidity and mortality(1–10). Racial/ ethnic disparities in injection-related AIDS, for example, Medical and Health Research Association of New York City, Inc. (H.C.); National Development and Research Institutes, Inc. (H.C., S.R.F., B.T., R.F., M.K.); Institute for AIDS Research, National Development and Research Institutes, Inc. (S.R.F.); Department of Epidemiology, Bloomberg School of Public Health, Johns Hopkins University (S.R.F); and Center for Urban Epidemiologic Studies, New York Academy of Medicine (S.R.F) New York, NY. Address correspondence to: Dr. Hannah Cooper, NDRI, 71 West 23rd Street, 8th Floor, New York, NY 10010. Tel.: (212) 845-4641; Fax: (917) 438-0894. E-mail:
[email protected] While conducting this research, the first author was supported by a Behavioral Science Training in Drug Abuse Research post-doctoral fellowship sponsored by the Mental and Health Research Association of New York City, Inc. and National Development and Research Institutes, Inc. with funding from the National Institute on Drug Abuse (5T32 DA07233). All other authors were supported by NIDA grant # R01 DA13336 (‘‘Community Vulnerability and Response to IDU-Related HIV’’). Points of view, opinions, and conclusions in this article do not necessarily represent the official position of the US Government, Medical Health and Research Association, or National Development and Research Institutes, Inc. Received June 7, 2004; accepted October 19, 2004. Ó 2005 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
are stark: in 2000, blacks and Latinos each comprised 12% of the adult population but 50% and 24%, respectively, of newly-diagnosed injection-related AIDS cases, while whites constituted 71% of adults but only 25% of newly-diagnosed injection-related AIDS cases (11, 12). Characterizing the racial/ethnic composition of injecting populations is vital to investigating whether racial/ethnic differences in rates of injecting contribute to excess injection-related morbidity and mortality among blacks and Latinos, as has been previously hypothesized (9, 13, 14), and thus in furthering progress toward the Healthy People 2010 goal of reducing racial/ethnic differences in HIV/AIDS rates (15). Efforts to characterize the composition of injecting populations have, however, been complicated by the stigmatized and illegal nature of injecting (16–20). The present analysis uses US Census data and three databases documenting injectors’ use of different healthcare services (drug treatment, HIV counseling and testing, and AIDS diagnoses) to calculate database-specific black/white and Latino/white disparities in injecting in each metropolitan statistical area (MSA) and also index black/white and Latino/white disparities, created by averaging data across the three databases. Racial/ethnic disparities, a major focus of Health People 2010 (15), can be seen as relative risks comparing the prevalence of a behavior or illness in one racial/ethnic group to its prevalence in another racial/ethnic 1047-2797/05/$–see front matter doi:10.1016/j.annepidem.2004.10.008
AEP Vol. 15, No. 5 May 2005: 326–334
Selected Abbreviations and Acronyms AIDS Z acquired immune deficiency syndrome APID Z AIDS Public Information Database CDC Z Centers for Disease Control and Prevention CTS Z Counseling and Testing System HIV Z human immunodeficiency virus MSA Z metropolitan statistical area SAMHSA Z Substance Abuse and Mental Health Services Administration TEDS Z Treatment Episode Data Set
group (21). We use pairwise comparisons (i.e., black vs. white and Latino vs. white) rather than a measure summarizing disparities across multiple racial/ethnic groups to enable future research to investigate the possibly distinct factors shaping the causes and consequences of injecting for blacks, Latinos, and whites (22, 23). MSAs are contiguous counties that include one or more central cities of at least 50,000 people that collectively form a single cohesive socioeconomic unit, defined by intercounty commuting patterns and socioeconomic integration (24, 25). MSAs were selected as the study’s unit of analysis because data were readily available at this geographic level and because we posited that MSAs were meaningful epidemiologic units with which to study injectors: drugrelated epidemics travel from central cities to their surrounding suburbs and injectors often live in suburbs but buy drugs and perhaps receive drug-related services in the central city (26, 27). Race/Ethnicity and Injection Drug Use Research regarding the racial/ethnic composition of injection drug using populations has reached disparate conclusions. The Substance Abuse and Mental Health Services Administration’s (SAMHSA’s) 1998 National Household Survey on Drug Abuse report, the most recent report describing injection drug use prevalence by race/ethnicity, indicates that the prevalences of lifetime injecting are similar among blacks, whites, and Latinos (1.5%, 1.3%, and 0.9%, respectively, with overlapping 95% confidence intervals) (28). SAMHSA statisticians, however, recognize that the survey substantially underestimates injection drug use prevalence (29, 30), suggesting that it may not be a reliable source of information on racial/ethnic injecting distributions. Other research suggests that racial/ethnic variations in drug addiction exist; given addiction’s links to injecting (31–33), these variations may shape injection drug use. Research on frequently-injected drugs such as cocaine and heroin suggests that while whites start using earlier and have a higher lifetime prevalence of drug use than blacks (34–37), blacks stay addicted for longer periods than whites and are more likely to relapse after leaving treatment (34, 38, 39). While Latinos as a group report patterns of lifetime use of injectable drugs that are similar to those of whites (40),
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
327
substantial intra-group variation exists, with individuals of Puerto Rican descent and more acculturated individuals reporting higher rates of substance use and abuse than other Latino subgroups (41–43). Both blacks and Latinos report that drug treatment programs are difficult to access and poorly designed to meet their needs (44–46), a circumstance that might prolong addiction among these groups. This research collectively suggests that the prevalence of injecting might be higher among blacks and some Latino subgroups than whites. METHODS Overview To calculate black/white and Latino/white disparities in injecting, we first estimated the percent of injectors in each MSA who were white, black, and Latino in each of three databases: SAMHSA’s Treatment Episode Data Set (TEDS); Centers for Disease Control and Prevention’s (CDC’s) HIV Counseling and Testing Service database (CTS) and CDC’s AIDS Public Information Database (APID). Using the resulting percents and US Census data on the percent of the total adult population in each MSA that was white, black, and Latino, we then calculated black/ white and Latino/white disparities in injecting in each MSA for each of the three databases. In addition to these databasespecific disparity estimates, we also created a single black/ white and Latino/white disparity estimate for each MSA by averaging data across the three databases. We report all four estimates of black/white and Latino/white disparities rather than just the index in the interests of transparency and to recognize that readers may have a preference for one estimate over the other, based on their needs. To validate our estimates, we correlated disparity scores with estimates of racial/ethnic disparities in injection-related health problems. While the focal year for all analyses was 1998, we combined 1997 to 1999 APID data to approximate 1998 because of small numbers of injection-related AIDS cases among blacks and Latinos in some MSAs. Sample To be included in the sample, MSAs had to have been home to more than 500,000 residents in 1993. Ninety-six MSAs met this criterion; 2 (San Juan-Bayamon, Puerto Rico and Richmond-Petersburg, Virginia), however, lacked needed data and were excluded, leaving a sample of 94. These 94 MSAs were located in 38 states and Washington, DC and had a median population of 1.2 million (range, 545,220– 9,519,338) in 2000. MSA boundaries remained constant between 1993 and 2000 (Personal Communication, M. Ratcliffe, Chief of Population Division, US Census Bureau, 2004).
328
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
AEP Vol. 15, No. 5 May 2005: 326–334
Database Descriptions TEDS, CTS, and APID document encounters with the healthcare system in all or almost all of the 94 MSAs in the study sample. TEDS records data on admissions to public and private drug treatment programs that receive state funds, certificates, or licenses (47); APID describes individuals diagnosed with AIDS in the US (48). CTS records data on individuals tested at the 11,640 HIV counseling and testing sites reporting to CDC (49). Inclusion of eligible cases in these databases, and in the 2000 Census, was high (85–98.8%) (47, 48, 50). Table 1 displays additional information on these databases. Identifying Eligible Individuals To be eligible for inclusion in the analysis as an injector, individuals in TEDS, CTS, and APID had to meet the following criteria: 1) be white, black, or Latino; 2) report injecting drugs since 1978; and 3) live in one of the 94
MSAs studied. Data on, and thus ascertainment of, injecting differed across databases: while TEDS contains information on current (but not past) injecting, APID and CTS contain information only on injecting at any point since 1978. Methods of linking individuals to MSAs also varied across databases: while APID records the MSA in which people resided at diagnosis, TEDS and CTS record the MSA in which individuals received services. 151,542 individuals met the study’s inclusion criteria in TEDS, 96,810 in CTS, and 22,975 in the 1997–1999 APID databases. These individuals were aggregated to form MSA-level variables within each database. Individuals in the 2000 US Census were included if they resided in one of the 94 MSAs studied, were 18 years old or older, and identified as Latino or reported that they were not Latino but recorded their race as white, black, or a combination of races that included white or black. Approximately 1% of all non-Latino black or white adults in the 94 MSAs identified themselves as having multiple
TABLE 1. Description of databases and variable ascertainment in the Treatment Episode Data System (1998), Counseling and Testing Service (1998), AIDS Public Information Database (1997–1999), and 2000 US Census Database characteristics Ascertainment of injection drug use
Ascertainment of race/ ethnicity
Ascertainment of metropolitan statistical area (MSA) Number of cases aggregated to create MSA-based data Number of sampled MSAs missing data
Treatment Episode Data System
Counseling and Testing Service
AIDS Public Information Database
Individual reports currently injecting his/her primary, secondary, or tertiary drug of abuse when admitted to treatment. Latino: Individual identifies Hispanic origin as Puerto Rican, Mexican, Cuban, other specific Hispanic origin, or unspecified Hispanic origin.
Individual reports injecting drugs at some point since 1978; includes men who had sex with men and also injected. Latino: Individual identifies as Hispanic.
Individual reports injecting drugs at some point since 1978; includes men who had sex with men and also injected. Latino: Individual identifies as Hispanic.
Black: Individual identifies race as black and is ‘Not of Hispanic origin’ or has missing data for this variable. White: Individual identifies race as white and is ‘Not of Hispanic origin’ or has missing data for this variable. Location of treatment program
Black: Individual identifies as black, not Hispanic.
Black: Individual identifies as black, not Hispanic.
White: Individual identifies as white, not Hispanic.
White: Individual identifies as white, not Hispanic.
White: Individual identifies as white and not Spanish/Hispanic/ Latino.**
Location of testing site
Location of residence
Location of residence
96,810 individuals receiving HIV counseling and testing at 11,640 sites 5
22,975 individuals diagnosed with AIDS.
130,598,079 adults (>18 years old) living in 94 MSAs 0
151,542 participants in 593 drug treatment programs 4
5yy
2000 US Census Not applicable.
Latino: Individual identifies as Mexican, MexicanAmerican, Chicano, Puerto Rican, Cuban, or as a member of another Spanish/Hispanic/ Latino group. Black: Individual identifies as black, African American, or Negro and not Spanish/ Hispanic/Latino.**
**The Deterministic Equal Fractions method was used to assign individuals to racial/ethnic groups when they marked multiple racial/ethnic identities. yy Because Counseling and Testing Services data were used to adjust the AIDS-based estimates, we could not calculate AIDS-based estimates in metropolitan areas where the Counseling and Testing Services database was missing data. The AIDS Public Information Database itself was not missing data for any of the metropolitan areas studied.
AEP Vol. 15, No. 5 May 2005: 326–334
races. In these cases, we used the Deterministic Equal Fractions method to assign individuals to racial/ethnic groups, a method that allocates equal fractions of a case to each of the multiple racial/ethnic identities selected (51). Sampled individuals were then aggregated to the MSA level. Analysis Analytic steps included 1) correcting APID-based estimates of the injecting population’s racial/ethnic composition for HIV seroprevalence, a step that was unnecessary for TEDS and CTS; 2) characterizing the racial/ethnic composition of the injecting population in each MSA, for each database and all three combined; 3) estimating and describing MSAlevel racial/ethnic disparities in injecting (both databasespecific and index); and 4) validating disparity estimates. 1) Calculating APID-based estimates of the injecting population’s racial/ethnic composition. The racial/ethnic composition of injectors in APID is produced by racial/ ethnic differences in 1) the injecting population and 2) HIV seroprevalence among injectors, which we assumed was linearly associated with AIDS rates. We therefore divided the number of injection-related AIDS cases in each racial/ ethnic group by the HIV seroprevalence of injectors in that same racial/ethnic group, a figure derived from the 1998 CTS database (see Formula 1). Formula 1. Calculating APID-based estimates of the injecting population’s racial/ethnic composition # InjectorsRace i; MSA1 / # Injection-Related AIDS Cases Racei; MSA1 = HIV Seroprevalence in Injectors Racei; MSA1 In 11 MSAs, 20 or fewer black injectors were tested at CTS sites; 20 or fewer Latinos were tested in 37 MSAs. Since small sample sizes lead to unstable seroprevalence rates, we used regression imputation methods to estimate the HIV seroprevalence for blacks and Latinos in these instances (see Formula 2). Formula 2. Regression imputation methods for estimating HIV seroprevalence among injectors HIV Seroprevalence in Injectors Racei; MSA1 ¼ aCb1 ðHIV Seroprevalence in Injectors Racej; MSA1 ÞC b2 ðHIV Seroprevalence in Injectors white; MSA1 Þ 2) Calculating racial/ethnic compositions of injecting populations: To calculate the percentages of injectors in the 94 MSAs who were white, black, and Latino for each of the TEDS, CTS, and (adjusted) APID databases (‘‘databasespecific percents’’), we divided the number of injectors in each racial/ethnic group by the total number of white, black,
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
329
and Latino injectors in each database. Single estimates of the percent of injectors in each racial/ethnic group in each MSA were calculated by averaging database-specific estimates together for each MSA and racial/ethnic group (‘‘index percent’’). Measures of central tendency and dispersion were tabulated for these index and databasespecific percents. 3) Estimating racial/ethnic disparities: To calculate database-specific black/white disparities in injecting for each MSA we first divided the database-specific percent of injectors who were black in each MSA by the percent of the total adult MSA population that was black. The ratio of injectors to the general population for blacks was then divided by the corresponding figure for whites to calculate black/white disparity scores for each database and MSA (‘‘database-specific disparities’’); database-specific Latino/ white disparities were calculated similarly (see Formula 3). Formula 3. Estimation of racial/ethnic disparities in injecting Racei =white Disparity in Injection Drug Use ¼ %InjectorsRacei; MSAj; Databasek = % Adult PopulationRacei; MSAj; Databasek % Injectorswhite; MSAj; Databasek = % Adult Populationwhite; MSAj; Databasek
We used identical methods to calculate a single black/ white and Latino/white disparity score (‘‘index disparity’’) for each MSA, except that we substituted the index percent of injectors who were white, black, and Latino for the database-specific values in Formula 3. Regression methods were used to impute index disparities in the 10 MSAs that lacked TEDS, CTS, or APID data. We then tabulated the central tendency and dispersion of both the databasespecific and index disparities. 4) Validating disparity estimates: To validate these estimates, we correlated racial/ethnic disparities in injecting with racial/ethnic disparities in 1) newly-diagnosed injection-related AIDS cases in 1998, calculated using APID, and 2) overdose fatalities, which we hypothesized might follow a similar pattern to disparities in injecting (52–54). Recognizing possible circularities inherent in correlating disparities in injection-related AIDS with the index injecting disparity (calculated using APID data), we ran two sets of correlations to validate the index black/ white and Latino/white injecting disparities, one using the original index disparity and another in which the index disparity was re-calculated using only CTS and TEDS. Overdose data were obtained from the CDC’s Multiple Cause of Death database; because of sample size and data limitations, we analyzed data from 1993 to 1998 and could not limit cases to those that were injection-related. Disparities in overdoses and injection-related AIDS were calculated using methods identical to those used to calculate racial/ethnic disparities in injecting.
330
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
AEP Vol. 15, No. 5 May 2005: 326–334
RESULTS Median database-specific and index values indicate that whites constituted the majority of injectors in the 94 MSAs studied (55.6–66.0%; see Table 2), followed by blacks (13.2–30.2%) and Latinos (8.4–10.0%). Census data, in contrast, indicate that whites comprised a median of 77.8% of the total adult population in the MSAs studied, while blacks comprised 9.1% and Latinos 6.2%. While calculations of the percent of injectors who were Latino were consistent across estimation methods, APID produced a higher percent of black injectors and a lower percent of white injectors than the other estimation methods. Measures of dispersion indicate substantial inter-MSA variation in the racial/ethnic compositions of injectors in the database-specific and index percents, particularly for Latinos. Median black/white disparity estimates in all four databases were 1.4 or above (see Table 3), suggesting that blacks were at greater risk of injecting than whites. The medians’ magnitude varied across databases, with APIDbased estimates considerably higher than TEDS-based estimates. Seventy-five percent of MSAs in all databases except TEDS had disparities >1.5. In contrast to findings regarding black/white disparities, median Latino/white disparities were approximately 1.0 in all four databases, suggesting that Latinos were as likely to inject as whites. The interquartile range around the median values and the percent of MSAs with disparity scores above 1.5 and below 0.67 indicate substantial variation in Latino/ white disparities across MSAs that is consistent across estimation methods. The ratio of white injectors to the adult population of whites (the denominator of the Latino/white disparity formula) was relatively constant across MSAs (and estimation methods), suggesting that variations in the ratio of Latino injectors to the adult population of Latinos drove much of the geographic variation found in these disparity scores. The validation analysis indicated that database-specific and index disparities (calculated both with and without the
APID-based estimates) correlated well with disparities in newly-diagnosed injection-related AIDS rates (see Table 4). Injecting disparities correlated modestly with racial/ethnic disparities in overdose fatalities, with the exception of the APID-based Latino/white disparity. One extreme value in the APID-based estimates for Latino/white disparity depressed the estimates’ correlation with overdose disparities; when this MSA (Buffalo-Niagara, NY) was removed, the correlation was 0.22 (p ! 0.04). DISCUSSION Though whites constituted the majority of the injecting population, all estimation procedures indicate that blacks were more likely to inject than whites in most of the 94 MSAs studied. In contrast, there was no evidence suggesting that the prevalence of injecting among Latinos was higher than that among whites, but Latino/white disparities were highly variable across the MSAs studied. The racial/ethnic disparities in injecting documented here are consistent with past research on racial/ethnic patterns of addiction duration and relapse (34, 38, 39) and lend credence to the proposition that disparities in injecting exist and contribute to pronounced and enduring disparities in injection-related health outcomes, including HIV/AIDS and overdoses. The moderate to high correlations between disparities in injecting and in injection-related AIDS rates also corroborate this hypothesis. Correlations with overdose fatality disparities might have been higher had we been able to restrict overdose deaths to those that were injection-related and perhaps remove possible racial/ethnic bias in medical examiners’ ascertainment of overdose as a cause of death. Future research is needed to explore possible relationships between disparities in injecting and in injection-related health problems. The causes of the racial/ethnic disparities in injecting found here also merit investigation. As has been suggested for black/white disparities in other health-related outcomes (55–63), it is possible that black/white injecting disparities stem in part from inequitable racial/ethnic relations. Past
TABLE 2. Measures of central tendency and dispersion for the racial/ethnic composition of the injection drug using and general adult populations in 94 US metropolitan statistical areas (MSAs), calculated for each database and for the average across all databases (‘‘index’’) MSA characteristic Racial/Ethnic composition of injecting population CTS-based TEDS-based APID-based Index-based Racial/Ethnic composition of general adult population
Median percent white (Interquartile range)
Median percent black (Interquartile range)
Median percent Latino (Interquartile range)
64.8 (53.0–74.4) 66.0 (54.4–81.7) 55.6 (37.4–69.2) 61.6 (51.2–73.4) 77.8 (66.9–86.8)
19.5 (11.1–30.5) 13.2 (6.0–27.1) 30.2 (15.8–45.7) 20.2 (10.3–34.1) 9.1 (6.0–17.0)
9.2 (2.1–24.1) 8.4 (1.8–21.3) 8.8 (1.9–19.1) 10.0 (2.2–25.0) 6.2 (2.9–17.4)
AEP Vol. 15, No. 5 May 2005: 326–334
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
331
TABLE 3. Measures of central tendency and dispersion for black/white and Latino/white disparitieszz in injection drug use in 94 US metropolitan statistical areas (MSAs), calculated for each database and for the index across all databases
Measures of central tendency and dispersion
CTS-based disparities (N Z 89)
TEDS-based disparities (N Z 90)
APID-based disparities (N Z 89)
Index disparities (N Z 94)
Black/white Latino/white Black/white Latino/white Black/white Latino/white Black/white Latino/white
Median (interquartile Range) 2.5 (1.5–3.6) 1.1 (0.6–1.7) 1.4 (0.8–3.0) 1.1 (0.5–1.8) 3.7 (2.3–6.6) 1.0 (0.6–2.7) 2.6 (1.7–3.5) 1.0 (0.6–1.8) Number (%) MSAs with disparity < 0.67 9 (10%) 33 (37%) 23 (26%) 34 (38%) 8 (9%) 33 (37%) 4 (4%) 26 (28%) Number (%) MSAs with disparity > 1.5 67 (75%) 28 (31%) 42 (47%) 27 (30%) 76 (85%) 32 (36%) 77 (82%) 28 (30%) zz
These racial/ethnic disparities can also be thought of as relative risks.
research delineates one possible pathway through which one manifestation of such inequitable relations, residential segregation, might increase injecting rates among blacks: predominately black neighborhoods of segregated cities tend to be characterized by high poverty levels, substandard schools, constrained employment opportunities, and limited access to municipal and health services; each of these neighborhood characteristics has been associated with adverse mental health outcomes that render individual residents vulnerable to using or abusing substances (64–69). Additional community characteristics and MSA policies, including differential policing of black and white communities and differential access to drugs in these communities (70–73), might be at work as well. Future research, ideally using multilevel methods, should investigate whether residential segregation and other aspects of inequitable racial/ethnic relations are associated with black/white disparities in injecting. TABLE 4. Validation of black/white and Latino/white injecting disparity estimates produced by the four estimation methods
Disparity in injecting (N) CTS-based (N Z 89) Black/white disparity Latino/white disparity TEDS-based (N Z 90) Black/white disparity Latino/white disparity APID-based (N Z 90) Black/white disparity Latino/white disparity Index, calculated with APID data (N Z 94) Black/white disparity Latino/white disparity Index, calculated without APID data (N Z 94) Black/white disparity Latino/white disparity
Correlation with corresponding injectionrelated AIDS disparity (p-value)
Correlation with corresponding overdose disparity (p-value)
0.61 (p!0.0001) 0.70 (p!0.0001)
0.32 (0.003) 0.48 (!0.0001)
0.61 (p!0.0001) 0.70 (p!0.0001)
0.35 (0.0007) 0.59 (!0.0001)
0.42 (p!0.0001) 0.80 (p!0.0001)
0.43 (!0.0001) 0.11 (0.31)
0.70 (p!0.0001) 0.80 (p!0.0001)
0.48 (!0.0001) 0.47 (!0.0001) N/A
0.64 (p!0.0001) 0.72 (p!0.0001)
Likewise, it is possible that the variation seen here in Latino/white disparity scores across MSAs partially mirrors differential patterns of immigration, acculturation, and discriminatory treatment of the many subgroups of the Latino populations in the 94 MSAs. A deeper investigation of our data suggests that migration patterns and residential segregation could conceivably account for some of the dispersion in Latino/white disparities seen here. We linked MSAs with US Census regions and found that MSAs with index Latino/white injecting disparities >1.5 tended to be located in the Northeast, where Puerto Ricans comprise the majority of Latino residents (74). Puerto Ricans, who report a higher prevalence of cocaine use than other Latino subgroups (43), are also the only other racial/ethnic group to endure a level of residential segregation comparable to blacks (66). One fruitful line of inquiry into the causes of Latino/white injecting disparities might thus lie in investigating the relationships among migration, residential segregation, and injection drug use. Our findings must be understood in light of their limitations. The fact that blacks and Latinos are more likely than whites to both underreport illicit drug use and be incarcerated might have led us to underestimate racial/ ethnic disparities in injecting (75), regardless of the database analyzed. Countering this possibility is the fact that the Census undercounts blacks and Latinos more than whites (50), an undercount that may have inflated our estimates of the relative percentages of blacks and Latinos who inject in each MSA in all databases. Finally, our use of 1998 CTS seroprevalence data to adjust APID-based data might have biased APID-based disparity estimates in unknown ways, a possibility somewhat supported by the validity analysis. We did not, however, wish to exclude APID-based disparity estimates because, as discussed below, they might capture a portion of the injecting population that is not usually linked to treatment and HIV testing services. All measures of Latino/white injecting disparities converged on similar results, suggesting that any one of these four estimation procedures could be selected as the ‘‘best estimate’’ with the exception of the APID-based
332
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
estimates that correlated poorly with disparities in overdose fatalities. In contrast, black/white disparities in injecting varied by estimation method, thus complicating the selection of a ‘‘best estimate.’’ This variation can perhaps be understood as reflecting the possibility that each of the three databases analyzed sampled different portions of the underlying population of black and white injectors in each MSA. TEDS might have captured a portion of the injecting population that was able to access relatively high-threshold services (i.e., inpatient and outpatient drug treatment programs). APID, in contrast, may have sampled injectors who had less contact with drug-related health services, a hypothesis that research on drug-related services and HIV infection and progression to AIDS substantiates (76–82). Given that blacks have poorer access to healthcare services than whites (44, 83), these inferences would produce precisely the pattern in disparities across databases found here: namely, that APID-based black/white disparity estimates would be substantially higher than those calculated using TEDS. The possibility that TEDS may have undersampled methamphetamine injectors, who tend to be white and underserved by treatment programs (84), may have attenuated this pattern. As a relatively low-threshold health service, CTS may have engaged a broader range of the injecting population in each MSA than TEDS. CTS, however, will distort the racial/ethnic composition of injectors to the extent that 1) injectors who have received a positive test have little reason to get re-tested and 2) the history of the local HIV epidemic varies by race/ethnicity (85). We are most comfortable using disparity index results as ‘‘best estimates’’ because they combine data sampled from different injecting populations that have varying access to health care services. This composite of samples might together represent the racial/ethnic demographics of the underlying injecting population in each MSA better than any single database-specific sample alone, though admittedly we have weighted each sample equally in our index disparity calculations and thus might have over-sampled one particular sub-group. The validation analysis supports the index as the ‘‘best estimate’’: index black/white injecting disparities correlated somewhat better with disparities in both injection-related AIDS and overdose fatalities than database-specific estimates. These findings suggest that black/white disparities in injecting exist in the majority of the 94 largest metropolitan areas in the USA. While Latinos and whites appeared to have similar rates of injecting, there was considerable variation in these disparities across MSAs. Future research that investigates the causes of these disparities, including racial/ethnic inequality and discrimination, and their consequences might help reduce enduring and pernicious racial/ethnic disparities in health.
AEP Vol. 15, No. 5 May 2005: 326–334
The authors thank the post-doctoral fellows in the Mental Health Research Association Behavioral Sciences Training Program, Drs. Greg Falkin, Terry Rosenberg, Holly Hagan, Sherry Deren, and Peter Flom, and Mr. Milton Mino for their invaluable comments on this article.
REFERENCES 1. Tardiff K, Gross E, Wu J. Analysis of cocaine-positive fatalities. J Forensic Sci. 1989;34:53–63. 2. Substance Abuse and Mental Health Services Administration. Emergency Department Trends from the Drug Abuse Warning Network, Final Estimates 1995–2002. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2003. 3. Harlow K. Patterns of rates of mortality from narcotics and cocaine overdose in Texas, 1976–1987. Public Health Rep. 1990;105:455–462. 4. Galea S, Ahern J, Tardiff K, Leon A, Coffin P, Derr K, et al. Racial/ethnic disparities in overdose mortality trends in New York City, 1990–1998. J Urban Health. 2003;80:201–211. 5. Friedman SR, Sotheran JL, Abdul-Quader A, Primm BJ, Des Jarlais DC, Kleinmann P, et al. The AIDS epidemic among Blacks and Hispanics. Milbank Q. 1987;65:455–499. 6. Friedman SR, Quimby E, Sufian M, Abdul-Quader A, Des Jarlais DC. Racial aspects of the AIDS epidemic. California Sociologist. 1988;11: 55–68. 7. Novick DM, Trigg HL, Des Jarlais DC, Friedman SR, Vlahov D, Kreek MJ. Cocaine injection and ethnicity in parenteral drug users during the early years of the human immunodeficiency virus (HIV) epidemic in New York City. J Med Virol. 1989;29:181–185. 8. Selik RM, Castro KG, Pappaioanou M, Buehler JW. Birthplace and the risk of AIDS among Hispanics in the United States. Am J Public Health. 1989;79:836–839. 9. Selik RM, Castro KG, Pappaioanou M. Racial/ethnic differences in the risk of AIDS in the United States. Am J Public Health. 1988;78:1539– 1545. 10. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2001. Atlanta, GA: Centers for Disease Control and Prevention; 2001. 11. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2000;12(2). Atlanta, GA: Centers for Disease Control and Prevention; 2000. 12. US Bureau of the Census. American Factfinder. Vol. 2004: US Bureau of the Census; 2003. 13. D’Aquila R, Peterson LR, Williams AB, Williams AE. Race/ethnicity as a risk factor for HIV-1 infection among Connecticut intravenous drug users. J Acquir Immune Defic Syndr. 1989;2:503–513. 14. Selik RM, Castro KG, Pappaioanou M. Distribution of AIDS cases by racial/ethnic group and exposure category, United States, June 1, 1981– July 4, 1988. MMWR Morb Mortal Wkly Rep. 1988;37:1–3. 15. US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. Washington, DC: US Department of Health and Human Services; 2000. 16. Cox S, Shipley M. Counting the uncatchable? An epidemiological method for counting drug misusers. Soc Psychiatry Psychiatr Epidemiol. 1997; 32:19–23. 17. Hickman M, Cox S, Harvey J, Howes S, Farrell M, Frischer M, et al. Estimating the prevalence of problem drug use in inner London: A discussion of three capture-recapture studies. Addiction. 1999;94:1653– 1662. 18. Kraus L, Augustin R, Frischer M, Kummler P, Uhl A, Wiessing L. Estimating prevalence of problem drug use at national level in countries of the European Union and Norway. Addiction. 2003;98:471–485.
AEP Vol. 15, No. 5 May 2005: 326–334
19. Larson A, Bammer G. Why? Who? How? Estimating numbers of illicit drug users: Lessons from a case study from the Australian Capital Territory Australian and New Zealand Journal of Public Health. 1996;20:493–499. 20. Larson A, Stevens A, Wardlaw G. Indirect estimates of ‘‘hidden’’ populations: Capture-recapture methods to estimate the numbers of heroin users in the Australian Capital Territory. Soc Sci Med. 1994; 39:823–831. 21. Carter-Pokras O, Baquet C. What is a ‘‘health disparity?’’ Public Health Rep. 2002;117:426–434. 22. Keppel KG, Pearcy JN, Klein RJ. Measuring Progress in Healthy People 2010. Healthy People 2010 Statistical Notes. 2004;25:1–16. 23. Pearcy JN, Keppel KG. A summary measure of health disparity. Public Health Rep. 2002;117:273–280. 24. Office of Management and Budget. Standards for defining metropolitan and micropolitan statistical areas. Federal Register. 2000;65:8228–8238. 25. US Bureau of the Census. State and Metropolitan Area Data Book, 1997– 1998. Washington, DC: US Bureau of the Census; 1998. 26. Pierce T. Gen-X junkie: Ethnographic research with young white heroin users in Washington, DC. Substance Use Misuse. 1999;34:2095–2114. 27. Wallace R, Wallace D. Socioeconomic determinants of health: Community marginalisation and the diffusion of disease and disorder in the United States. BMJ. 1997;314:1341–1345. 28. US Department of Health and Human Services. National Household Survey on Drug Abuse: Population Estimates 1998. Rockville, MD: US Department of Health and Human Services; 1999. 29. Wright D, Gfroerer J. Estimation of hardcore drug users. Journal of Official Statistics. 1997;13:401–416. 30. Wright D, Gfroerer J, Epstein J. The use of external data sources and ratio estimation to improve estimates of hardcore drug use from the NHSDA. NIDA Research Monograph. 1997;167:477–497. 31. Dinwiddie SH, Cottler L, Compton W, Abdallah AB. Psychopathology and HIV risk behaviors among injection drug users in and out of treatment. Drug and Alcohol Dependence. 1996;43:1–11. 32. Barrio G, De La Fuente L, Lew C, Royuela L, Bravo MJ, Torrens M. Differences in severity of heroin dependence by route of administration: The importance of length of heroin use. Drug and Alcohol Dependence. 2001;63:169–177. 33. Irwin KL, Edlin BR, Faruque S, McCoy HV, Word C, Serrano Y, et al. Crack cocaine smokers who turn to drug injection: Characteristics, factors associated with injection, and implications for HIV Transmission. Drug Alcohol Depend. 1996;42:85–92. 34. Warner LA, Kessler RC, Hughes M, Anthony JC, Nelson CB. Prevalence and correlates of drug use and dependence in the United States. Results from the National Comorbidity Study. Arch Gen Psychiatry. 1995; 52:219–229. 35. Wallace JM, Bachman JG, O’Malley PM, Johnston LD, Schulenberg JE, Cooper SM. Tobacco, alcohol, and illicit drug use: Racial and ethnic differences among US high school seniors, 1976–2000. Public Health Rep. 2002;117:S67–S75. 36. Johnston L, O’Malley PM, Bachman JG. Monitoring the Future: National Survey Results on Drug Use, 1975–2000. Volume 1: Secondary students. Rockville, MD: US Department of Health and Human Services; 2001. 37. US Department of Health and Human Services. Drug Use among Racial/ Ethnic Minorities, Revised. Bethesda, MD: US Department of Health and Human Services; 2003; 1–172.
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
333
Available at http://www.samhsa.gov/oas/nhsda/2k2nsduh/html/toc.htm. Accessed February 10, 2004. 41. Substance Abuse and Mental Health Services Administration. Results from the 2002 National Survey on Drug Use and Health: National Findings. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2003; 1–256. 42. Vega WA, Alderete E, Kolody B, Aguilar-Gaxiola S. Illicit drug use among Mexicans and Mexican Americans in California: The effects of gender and acculturation. Addiction. 1998;93:1839–1850. 43. Amaro H, Whitaker R, Coffman G, Heeren T. Acculturation and marijuana and cocaine use: Findings from HHANES 1982–1984. Am J Public Health. 1990;80:54–60. 44. Wells K, Klap R, Koike A, Sherbourne C. Ethnic disparities in unmet need for alcoholism, drug abuse, and mental health care. Am J Psychiatry. 2001;158:2027–2032. 45. Walton MA, Blow FC, Booth BM. Diversity in relapse prevention needs: Gender and race comparisons among substance abuse treatment patients. Am J Drug Alcohol Abuse. 2001;27:225–240. 46. Porter J. The street/treatment experiences of Puerto Rican injection drug users. Substance Use Misuse. 1999;34:1951–1975. 47. Substance Abuse and Mental Health Services Administration. Treatment Episode Data Set (TEDS). Vol. 2004: Substance Abuse and Mental Health Services Administration; 2004. 48. US Department of Health and Human Services. AIDS Public Information Data Set. Washington, DC: US Department of Health and Human Services; 2000; 1–52. 49. Centers for Disease Control and Prevention. HIV Counseling And Testing in Publicly Funded Sites. Annual Report 1997–1998. Atlanta, GA: Centers for Disease Control and Prevention; 1999. 50. Executive Steering Committee for A.C.E. Policy (ESCAP). Report of the Executive Steering Committee for Accuracy and Coverage Evaluation Policy. Washington, DC: US Bureau of the Census; 2001:1–28. 51. Office of Management and Budget. Provisional Guidance of the Implementation of the 1997 Standards for Federal Data on Race and Ethnicity. Vol. 2003: Office of Management and Budget; 2000. 52. Stewart D, Gossop M, Marsen J. Reductions in non-fatal overdose after drug misuse treatment: Results from the National Treatment Outcome Research Study (NTORS). J Subst Abuse Treat. 2002;22:1–9. 53. Brugal MT, Barrio G, De LF, Regidor E, Royuela L, Suelves JM. Factors associated with non-fatal heroin overdose: Assessing the effect of frequency and route of heroin administration. Addiction. 2002;97:319– 327. 54. Darke S, Zador D. Fatal heroin ‘overdose’: A review. Addiction. 1996; 91:1765–1772. 55. Hart KD, Kunitz SJ, Sell RR, Mukamel DB. Metropolitan governance, residential segregation, and mortality among African Americans. Am J Public Health. 1998;88:434–438. 56. Jackson S, Anderson R, Johnson N, Sorlie P. The relation of residential segregation to all-cause mortality: A study in black and white. Am J Public Health. 2000;90:614–617. 57. James S, LaCroix AZ, Kleinbaum DG, Strogatz DS. John Henryism and blood pressure differences among lack men. II. The role of occupational stressors. J Behav Med. 1984;7:259–275. 58. Krieger N. Racial and gender discrimination: Risk factors for high blood pressure? Soc Sci Med. 1990;30:1273–1281.
38. Walton MA, Blow FC, Bingham CR, Chermack ST. Individual and social/ environmental predictors of alcohol and drug use 2 years following substance abuse treatment. Addict Behav. 2003;28:627–642.
59. Krieger N, Sidney S. Racial discrimination and blood pressure: The CARDIA Study of young black and white adults. Am J Public Health. 1996;86:1370–1378.
39. Moos RH, Moor BS, Finney JW. Predictors of deterioration among patients with substance-use disorders. J Clin Psychol. 2001;57:1403–1419.
60. LaViest TA. Linking residential segregation to the infant-mortality race disparity in U.S. cities. Sociology and Social Research. 1989;73:90–94.
40. Substance Abuse and Mental Health Services Administration. Results from the 2002 National Survey on Drug Use and Health: Detailed Tables.
61. Polednak A. Black-white differences in infant mortality in 38 standard metropolitan statistical areas. Am J Public Health. 1991;81:1480–1482.
334
Cooper et al. RACIAL/ETHNIC DISPARITIES IN INJECTING
AEP Vol. 15, No. 5 May 2005: 326–334
62. Polednak A. Poverty, residential segregation, and black/white mortality ratios in urban areas. J Health Care Poor Underserved. 1993;4:363–373.
75. Johnson TP, Bowman PJ. Cross-cultural sources of measurement error in substance use surveys. Substance Use Misuse. 2003;38:1447–1490.
63. Polednak A. Trends in US urban black infant mortality, by degree of racial segregation. Am J Public Health. 1996;86:723–726.
76. Hoffman J, Klein H, Clark D, Boyd F. The effect of entering drug treatment on involvement in HIV-related risk behaviors. Am J Drug Alcohol Abuse. 1998;24:259–284.
64. Aneshensel CS, Sucoff CA. The neighborhood context of adolescent mental health. J Health Soc Behav. 1996;37:293–311. 65. Schultz A, Williams DR, Israel B, Becker A, Parker E, James SA, et al. Unfair treatment, neighborhood effects, and mental health in the Detroit metropolitan area. J Health Soc Behav. 2000;41:314–332.
77. Hartel D, Schoenbaum E. Methadone treatment protects against HIV infection: Two decades of experience in the Bronx, New York City. Public Health Rep. 1998;113:105–115.
66. Massey DS, Denton NA. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press; 1993.
78. Woods W, Guydish J, Sorenson J, Coutts A, Bostrom A, Acampora A. Changes in HIV-related risk behaviors following drug abuse treatment. AIDS. 1999;13:2151–2155.
67. Wacquant LJ, Wilson WJ. The cost of racial and class exclusion in the inner city. Annals of the American Academy of Political and Social Sciences. 1989;501:8–25.
79. Thiede H, Hagan H, Murrill C. Methadone maintenance and HIV and hepatitis B and C risk reduction among injectors in the Seattle area. J Urban Health. 2000;77:331–345.
68. Singer B, Ryff CD. New Horizons in Health: An Integrative Approach. Washington, DC: National Academy Press; 2001.
80. Kwiatkowski CF, Booth RE. Methadone maintenance as HIV risk reduction with street-recruited injecting drug users. J Acquir Immune Defic Syndr. 2001;26:483–489.
69. Williams DR, Williams-Morris R. Racism and mental health: The AfricanAmerican experience. Ethnicity and Health. 2000;5:243–268. 70. Friedman SR, Tempalski B, Cooper H, Keem M, Friedman R, Flom PL. Structural Factors to Guide Structural Intervention: Predictors of IDU Population Size, HIV Prevalence Among IDUs, and Prevention Program Coverage for IDUs. XV International AIDS Conference, Bangkok, Thailand, 2004. 71. Delva J, Mathiesen S, Kamata A. Use of illegal drugs among mothers across racial/ethnic groups in the United States: A multi-level analysis of individual and community-level influences. Ethnicity and Disease. 2001;11:614–625.
81. Metzger DS, Navaline H. Human immunodeficiency virus prevention and the potential of drug abuse treatment. Clin Infect Dis. 2003;37:S451– S456. 82. Janssen R, Holtgrave D, Valdisseri R, Shepherd M, Gayle H, De Cok K. The serostatus approach to fighting the HIV epidemic: Preventive strategies for infected individuals. Am J Public Health. 2001;91:1019– 1024. 83. Smedley BD, Stith AY, Nelson AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Healthcare. Washington, DC: National Academy Press; 2003.
72. Crum R, Lillie-Blanton M, Anthony JC. Neighborhood Environment and opportunity to use cocaine and other drugs in late childhood and early adolescence. Drug Alcohol Depend. 1996;43:155–161.
84. Anglin MD, Burke C, Perrochet B, Stamper E, Dawud-Noursi S. History of the methamphetamine problem. J Psychoactive Drugs. 2000; 32:137–141.
73. Cooper H, Moore L, Gruskin S, Krieger N. Characterizing perceived police violence: Implications for public health. Am J Public Health. 2004; 94:1109–1118.
85. Friedman SR, Tempalski B, Cooper H, Perlis T, Keem M, Friedman R, et al. Estimating number of injecting drug users in metropolitan areas for structural analyses of community vulnerability and for assessing relative degrees of service provision for injecting drug users. J Urban Health. 2004;81:377–400.
74. Guzman B. The Hispanic Population: Census 2000 Brief. Washington, DC: US Census Bureau; 2001.
