1996 Drug Use Forecasting: Annual Report on Adult and Juvenile Arrestees . MENU TITLE: 1996 Drug Use Forecasting: Annual Report on Adult and Juvenile Arrestees Series: NIJ Published: 1996 44 pages 12,123 bytes National Institute of Justice Research Report Drug Use Forecasting 1996 Annual Report on Adult and Juvenile Arrestees ----------------------------- U.S. Department of Justice Office of Justice Programs 633 Indiana Avenue N.W. Washington, DC 20531 Janet Reno Attorney General U.S. Department of Justice John C. Dwyer Acting Associate Attorney General Laurie Robinson Assistant Attorney General Jeremy Travis Director, National Institute of Justice Justice Information Center World Wide Web Site http://www.ncjrs.org ------------------------------ From the Director of NIJ In last year's Annual Report I described 1995 as a momentous year for the Drug Use Forecasting (DUF) program. Calling 1996 momentous would not only be redundant, but would be an understatement. Since this Report was last published, the President has submitted a budget request to Congress to reengineer DUF into the Arrestee Drug Abuse Monitoring (ADAM) program. ADAM will build on DUF's success by more than tripling in size to conduct quarterly data collection in 75 urban areas. Equally important, ADAM will include an outreach component under which each urban site conducts annual data collection in a suburban, rural, Indian Territory, or other arrestee population. Outreach data collection will give us vital insights into the leading and trailing edges of drug epidemics and into the links between drugs and crime beyond our central cities. ADAM will vastly improve NIJ's geographic coverage, and will place a powerful research and evaluation tool in 75 cities and counties affected by drugs and crime. NIJ continued to provide leadership in the study of methamphetamine. DUF funded a San Diego Association of Governments (SANDAG) study of methamphetamine markets and use. When completed in 1997 this study will provide the first comprehensive, cross-site analysis of methamphetamine users, their use habits, and their buying practices. This study promises to be an invaluable aid to jurisdictions as they develop policy responses to this drug. The methamphetamine study will also be an important companion to a similar study on heroin, crack cocaine, and powder cocaine markets that NIJ conducted, in partnership with the Office of National Drug Control Policy, which will be published in 1997. DUF is also poised to make strong contributions to evaluating the impact of a program of universal needs assessment and drug testing combined with a tailored program of treatment, services, sanctions, and supervision and to our understanding of the community's influence on criminal history and criminal activity. The needs assessment, drug testing, and services program, called Breaking the Cycle, is being implemented as a demonstration project in Birmingham. In a step that NIJ hopes to see duplicated many times over with ADAM, the team evaluating Breaking the Cycle is using DUF to collect baseline data and measure program impact. The community study is being undertaken at the Denver DUF site and it promises to provide valuable insights as to links between criminal behavior and the community environment. NIJ's intramural research program continued apace in 1996. NIJ's "Homicide in U.S. Cities" project work shows strong links between drug activity and homicide in eight large U.S. cities. DUF data were a key element in illuminating and analyzing these complex relationships. Publications relating to the homicide study are expected in the first half of 1997. As DUF makes the transition to ADAM it becomes a more valuable research and evaluation tool. I urge you to keep up to date with these important changes by contacting our NIJ staff and by reading NIJ publications. Documents can be requested by contacting NIJ's National Criminal Justice Reference Service at 800-851-3420 or on the Internet at http://www.ncjrs.org. And I especially thank you for your support of these initiatives. Jeremy Travis, Director, NIJ ------------------------------ Table of Contents From the Director of NIJ New Features 1996 Marijuana Data 1996 DUF Annual Report on Drug Use Among Arrestees Methodology DUF Juveniles, 1991-1996 Recidivism Rates in the DUF Sample TELEDUF Pilot Project Guide to DUF Site Data Tables Adult and Juvenile Program Findings Atlanta Birmingham Chicago Cleveland Dallas Denver Detroit Ft. Lauderdale Houston Indianapolis Los Angeles Manhattan Miami New Orleans Omaha Philadelphia Phoenix Portland St. Louis San Antonio San Diego San Jose Washington, D.C. Site Reports Atlanta Cleveland Denver Detroit Los Angeles Manhattan Omaha Portland San Diego From DUF to ADAM ------------------------------ New Features Look for these features in the 10th anniversary edition of the Drug Use Forecasting (DUF) Annual Report: o A discussion of an important change in marijuana testing cutoff levels that affects interpretation of marijuana data in the Annual Report. o Updated coverage of methamphetamine issues that were reported in the 1995 Drug Use Forecasting Annual Report on Adult and Juvenile Arrestees and in the special 1996 National Institute of Justice (NIJ) report Methamphetamine Use Among Adult Arrestees: Findings from the Drug Use Forecasting (DUF) Program. o Extended analysis of juvenile drug issues using 1991-1996 data. o Analysis of recidivism using questions that were added to the DUF interview in 1995. o A report on the pilot testing of an automated telephone DUF interviewing system (TELEDUF). o The addition of site report pages that reflect the range of issues that DUF data are used to address. As DUF continues its evolution toward the Arrestee Drug Abuse Monitoring (ADAM) program, analyses such as these that focus on local issues, special populations, emerging trends in drugs and crime, and new areas of research and interviewing will become more prominent features of the Annual Report. ------------------------------ 1996 Marijuana Data Changes in drug detection methodologies and standards have resulted in changes to DUF reporting practices for marijuana. Changes in Technology and Reporting Levels Improvements in drug testing technology have increased the sensitivity (the ability of a test to detect the presence of a drug) and the specificity (the ability to distinguish a specific drug from other cross-reactants) of urinalysis tests used to screen for drug use, such as the widely used EMIT TM. In the case of marijuana, these improvements have resulted in a lowering of the standard cutoff level for a positive test report. This change was recommended by the Federal workplace testing program and new guidelines were issued by the Substance Abuse and Mental Health Services Administration (SAMHSA) of the Department of Health and Human Services, effective September 1, 1994. A lower cutoff is expected to identify a greater number of drug users, particularly those who may use marijuana occasionally or in moderate quantities. The concentration of drugs in urine is measured in nanograms (billionths of a gram) per milliliter (ng/ml) of liquid of the drug or the drug metabolite formed in the body as the result of the ingestion of a specific drug. The "cutoff level" is the concentration, stated in ng/ml, used to determine whether a specimen is positive or negative. Specimens with concentrations at or above the cutoff level are considered positive for the drug in question; all others are considered negative. In the case of marijuana, the cutoff level had previously been set at 100 ng/ml. SAMHSA now recommends a lower cutoff of 50 ng/ml. Impact of Changes When the DUF program began in 1987, DUF adopted Federal guidelines for cutoff levels which were then 100 ng/ml for marijuana. During 1995 as part of a feasibility study, urinalysis tests for marijuana were conducted using both the 100 ng/ml and 50 ng/ml cutoff levels. As a result of these analyses, beginning in January 1996, all DUF marijuana tests are based on the lower cutoff level. This analysis examines the impact of this change in cutoff levels on percentage positive for marijuana among subpopulations of DUF arrestees and in DUF sites. NIJ will publish a full version of this report in 1997. More than 34,000 samples from 1995 were tested at both cutoff levels from all sites. Adult males, adult females, and juvenile males were included. (See figure 1 on the next page.) Overall, lowering the cutoff increases the percentage positive for marijuana use about 5 to 7 percentage points. The greatest impact of the change in cutoff levels appears when positive tests are examined for different age groups. Greater numbers of younger (under age 15) and older (over age 30) arrestees test positive for marijuana when the cutoff level is lowered from 100 ng/ml to 50 ng/ml. The youngest and oldest age groups are likely to be less frequent users of marijuana or use it in lesser amounts than other arrestees. Hence, these individuals test positive at the lower cutoff level of 50 ng/ml, but not at the higher level of 100 ng/ml. Overall, the mean age of arrestees who test positive for marijuana in the 1995 data increases from 25 years to almost 28 years when the lower cutoff is used. The lower cutoff level had the same impact for blacks, whites, and Hispanics. A separate analysis examined trends in marijuana use using both the 50 and 100 ng/ml cutoff levels. The change in cutoff level has little effect on the overall trend in marijuana use demonstrated for 1995. For some quarters the increase in percentage positive is larger than for other quarters, but overall there is a 5 to 7 percentage point increase when percentage positive is determined at the 50 ng/ml level. Larger differences in percentage positive occur in some quarters for female and juvenile male arrestees. As discussed earlier, these are two of the groups for whom the change in cutoff level had the greatest effect. Implications for 1996 Marijuana Data Marijuana data for 1996 reveal that marijuana use has increased substantially in many DUF sites. Substantial increases have also been detected in many age groups and other subpopulations. In many, but not all, cases the increases in marijuana use substantially exceed the increases that can be expected from the change to the 50 ng/ml cutoff level. For example, marijuana positives for 31- to 35-year-old arrestees increased by 16 percentage points in Indianapolis and Atlanta, 15 percentage points in Cleveland, and 12 percentage points in Birmingham. Similarly, the median rate of marijuana prevalence for 15- to 20-year-olds increased 11 percentage points over 1995, but the rate of change for the youngest males varied across sites from a 6-point decrease in Houston to a 19-point increase in Indianapolis. While the increases recorded in marijuana use generally far exceed increases that would be expected from the change to the 50 ng/ml cutoff, and thus indicate true changes in marijuana use, readers should use caution when assessing the magnitude of any given change and its significance. Moreover, changes should be considered in the context of the specific site and subpopulation of interest. Readers should be particularly careful when comparing 1995 and 1996 marijuana test positive percentages for groups that are particularly affected by the change from the 100 ng/ml to 50 ng/ml cutoff levels, including female arrestees, younger arrestees, and older arrestees. Christy Visher, Ph.D., and Thomas E. Feucht, Ph.D., with K. Jack Riley, Ph.D. Office of Research and Evaluation, NIJ ------------------------------ 1996 DUF Annual Report on Drug Use Among Arrestees In 1996, DUF program sites located in 23 major metropolitan areas collected data from 19,835 adult male booked arrestees. Data were also collected from 7,532 adult female booked arrestees at 21 of these sites and 4,145 juvenile male and 645 juvenile female detainees at 12 sites and 7 sites, respectively. This report presents drug use detected through urinalysis for adult male and female arrestees and juvenile male arrestees/detainees. Because of small sample sizes, data on female juvenile arrestees/detainees are not included. Program findings are reported in three sections. The first section provides an overview of trends and issues in the 23 sites. The findings for adult males, adult females, and juvenile males are shown according to drug (marijuana, cocaine, and opiates), age group (particularly the youngest adults), and other categories (school status for juvenile males). The section concludes with a special analysis of methamphetamine. The second section of the report presents special topics and analyses, including the impact of changing cutoff levels for marijuana urinalysis (see "1996 Marijuana Data" on page 7 of this report) and an overview of TELEDUF. This section also includes analyses of juvenile DUF data and recidivism. In the third section, site-specific tables and graphical analyses for adults and juveniles are provided. To assist readers, the report includes a discussion of DUF data collection methodology on page 13 and a guide to the tables on page 20. The report concludes with selected DUF site reports on local and policy issues that have relied on DUF data. Drug Use Among Adult Male Arrestees o Marijuana use among adult male arrestees increased at almost every site, at rates exceeding those noted in recent years. o Compared to 1995 data, 12 sites showed decreased percentages of adult males testing positive for cocaine, 9 sites showed increased percentages, and 2 sites registered the same percentage. A general trend of increases in the fraction of arrestees testing positive for marijuana is apparent across sites. Only Phoenix reported a decline and San Jose reported no change in adult male marijuana test positive percentages. In contrast, regional patterns are more evident for cocaine, opiates, and methamphetamine. Cocaine, which has historically been the most commonly used drug among DUF arrestees in most sites, was surpassed by marijuana in popularity among male adult arrestees in many cities, but primarily in the Western United States. High rates of amphetamine use remain largely a Western U.S. phenomenon, while the highest rates of opiate use continue to be confined to a few large cities. Use of Marijuana: o In 1996, increasing rates of marijuana use registered across all age categories of adult males. This finding is in contrast to past years where increases were noted primarily in the juvenile and young arrestee populations. In nine DUF sites, the increase of marijuana positives from 1995 to 1996 among 31- to 35-year-old arrestees reached or exceeded 10 percentage points, and included increases of 16 percentage points (Indianapolis and Atlanta), 15 percentage points (Cleveland), and 12 percentage points (Birmingham). Use of Cocaine: o While cocaine use among male arrestees continued to decline or remain stable in many DUF cities, remarkable increases were noted in several sites. In Omaha, cocaine positives for adult male arrestees grew to 24 percent in 1996, up from 19 percent in 1995. In Miami, cocaine positives increased from 42 to 52 percent. Cocaine test positives rose 3 percentage points in Indianapolis. In other sites (Dallas and Houston) where there were overall decreases or a leveling off of cocaine positives in the adult male population, potentially significant increases nonetheless showed up among 15- to 20-year-olds (a finding that is discussed further below). Given the small number of cases, however, caution should be used when assessing the significance of the trend in this age category. Use of Opiates: o Opiate positives among adult male arrestees remained low relative to cocaine and marijuana, although a few sites reported rates of more than 10 percent. Opiate use among male arrestees continued to be highest in Chicago, Manhattan, Philadelphia, Portland, St. Louis, and San Antonio. In each of those cities, opiate test positives equaled or exceeded 10 percent in 1996. The highest recorded percentage among adult male arrestees was 20 percent, found in Chicago. In each of these sites, however, the rate dropped 1 to 3 percentage points from 1995, except in San Antonio where it remained the same. Use of At Least One Drug: o In the majority of sites (15 out of 23), the rate at which adult male arrestees were found positive for at least one drug increased over the last year. In 20 of 23 sites, more than 60 percent of adult male arrestees tested positive for at least one drug, and two more sites were within 3 percentage points of the 60-percent barrier. In only one site -- San Jose -- did less than 50 percent of the arrestees test positive for at least one drug. Several trends appeared to account for the overall higher rates of drug use in the adult male arrestee population nationwide. The greatest increases (5 to 10 percent) were seen in sites where both marijuana and cocaine positives are climbing. These sites include Denver, Fort Lauderdale, Indianapolis, Omaha, and San Antonio, all of which are cities not historically associated with the highest rates of drug positives in DUF data, but which are currently experiencing increases in prevalence in the arrestee population. On the other hand, cities traditionally showing high drug test positive percentages, such as San Diego and St. Louis, showed stability that is explained by a drop in cocaine positives and an increase in marijuana positives. Furthermore, in Manhattan and Philadelphia the rate of positives found for any drug among the adult male population decreased by 5 and 7 percentage points, respectively, despite the fact that these cities followed the nationwide trend of increased marijuana test positives. The declines in these two cities can be explained by significant decreases in cocaine and opiate test positives in adult male arrestees. Drug Use Among the Youngest Adult Male Arrestees o The percentage of the youngest males testing positive for marijuana increased sharply in most sites. The median rate of marijuana prevalence for this group was 64 percent, an increase of 11 percentage points over the past year. However, the rate of change varied across sites from a 6-point decrease in Houston to a 19-point increase in Indianapolis. o Recent cocaine use, measured through urinalysis, among the youngest male arrestees continued to drop in most sites, but increased noticeably in others. The decline in cocaine positives among young males noted in many DUF sites in recent years contrasts with increasing rates for this group in a number of sites in 1996, the most pronounced being found in Houston (14 percentage points). Other sites that showed increases are Omaha (11 points), Miami (10 points), and Indianapolis (8 points). Use of Opiates: o The median rate for opiate test positives was 2 percent among the youngest male arrestees. While the youngest adult male arrestee group exhibited the lowest prevalence rates for opiates among adult males in 1996, the percentage testing positive increased in nine sites. Of special note are Philadelphia and St. Louis in which, respectively, 12 and 14 percent of the youngest males tested positive for opiates. These are high levels for this age bracket and thus these figures bear watching to determine if they are indicative of an emerging or more widespread heroin problem in these communities. Drug Use Among Adult Female Arrestees o In 20 of 21 sites collecting female data, the fraction of adult female arrestees testing positive for marijuana increased. o Consistent with previous years, adult females exhibited higher prevalence rates for cocaine use than did adult males. Use of Marijuana: In 1996 adult females displayed notable increases in marijuana use. In five sites, increases reached 10 or more percentage points: Atlanta (13 points), Birmingham (10 points), Cleveland (11 points), Portland (10 points), and St. Louis (11 points). The highest rates of use were among those under age 21, with a median rate of 36 percent for that age group. Females 21 and older were detected as recent users of marijuana less frequently. Use of Cocaine: The median rate for cocaine test positives among adult DUF females continued to drop slowly -- from 50 percent in 1994 to 48 percent in 1995 and 46 percent in 1996. Despite the consistent decrease, there was significant variation among sites. At the majority of sites, rates began leveling off, with large decreases at five sites (New Orleans and Cleveland down 11 points, Birmingham down 9 points, and Dallas and Detroit down 8 points). On the other hand, some sites registered sharp increases, with Philadelphia up by 10 percentage points and Phoenix up by 9 percentage points. Increases of 5 and 6 percentage points for cocaine test positives were seen among females in San Jose and Portland, respectively. Use of Opiates: o Generally, opiate use among adult females remained stable or increased slightly. Two exceptions to overall stable rates of opiate use were seen in Manhattan and Portland. In each of those two cities, 8-point increases were reported, bringing the opiate test positives among adult female arrestees up to 27 and 26 percent, respectively. In Portland, the same percentage of adult female arrestees tested positive for opiates as tested positive for marijuana. Both Manhattan and Portland opiate figures were among the highest. San Diego, a third site with historically high rates of opiate positives among its adult female arrestees, however, demonstrated a decline among females and is currently at 10 percent prevalence. Use of At Least One Drug: o The percentage of female adults testing positive for at least one drug increased overall. In the majority of sites, overall drug use increased, and in all but two sites (New Orleans and San Antonio), the rate was more than 50 percent. Even with cocaine use slowing down in many sites, the decrease was offset by increased use of marijuana, amphetamines, and, to a lesser degree, opiates. Drug Use Among the Youngest Female Arrestees o Decreases in cocaine use for the youngest female arrestees were less dramatic than in the previous year. While several sites reported sharp declines in cocaine positives consistent with the previous year (percentages in Birmingham, Dallas, and Denver decreased by 12 to 16 points), percentages increased in 10 sites. The most notable increases in cocaine test positive percentages among the youngest females were in St. Louis (17 points), Ft. Lauderdale (8 points), and Portland and Washington, D.C. (6 points). Again, because of small numbers of respondents in these categories the percentage changes should be interpreted cautiously. o Female arrestees under age 21 had the highest percentage of marijuana positives among adult females. In every site, marijuana positives among the youngest female arrestees were more than 20 percent, with dramatic increases over 1995 in a few sites. In St. Louis, the rate for this population went from 26 to 50 percent, in Portland from 15 to 30 percent, in Cleveland from 18 to 47 percent, in Birmingham from 18 to 43 percent, and in Atlanta from 10 to 49 percent. Use of Opiates: o Prevalence of opiate use varied considerably across sites for the youngest female arrestees. Although nine sites showed rates of less than 1 percent for opiate use among this group, in some sites high rates were found even among the youngest female arrestees. In Manhattan, Philadelphia, and Portland, the numbers were at 15 percent or greater. In each of these sites, the percentage represented an increase over 1995. Drug Use Among Male Juvenile Arrestees/Detainees Interviews and urine tests were conducted with 4,145 boys in 12 sites in 1996. o Marijuana use rose sharply and cocaine use was up slightly among male juveniles. Drug use among boys was greater in every site but one (in San Diego rates held steady), with increases of 10 or more percentage points in 8 out of 12 sites. The increase was due mainly to marijuana use. The median marijuana test positive rate for boys was 52 percent in 1996, compared to 41 percent in 1995. Cocaine use, typically low among juvenile males, has fluctuated in recent years and in 1996 took a slight upturn, increasing in the majority of sites. Sites with the highest rates of cocaine use were Cleveland, Los Angeles, Phoenix, and San Antonio, ranging from 10 to 13 percent prevalence. Use of opiates among male juveniles remained very low overall. Drug Use and School Attendance: o Overall, cocaine use was much higher for boys out of school than for boys in school, a consistent finding for several years. Marijuana use rates for boys out of school continued to be higher than for boys in school; however, in three sites (Denver, Los Angeles, and Phoenix), the rates were within 3 percentage points of each other. The median prevalence rate for marijuana use by boys out of school (61 percent) was lower by 3 points than the median use rate for 15- to 20-year-old male arrestees -- the adult group with the highest rates of marijuana use. The median marijuana use rate for boys in school was 48 percent. Other Drug Use Use of Methamphetamine: o While methamphetamine use continued to be detected mainly in Western U.S. DUF sites, test positive rates fell significantly from 1995 levels. The eight cities (San Diego, Phoenix, San Jose, Portland, Omaha, Los Angeles, Denver, and Dallas) that were cited in the 1995 Drug Use Forecasting Annual Report on Adult and Juvenile Arrestees as having the highest methamphetamine test positive rates among adult arrestees all reported substantial declines in 1996. In San Diego, adult methamphetamine test positives declined from 37.1 to 29.9 percent; in Phoenix from 21.9 to 12.2 percent; in Portland from 18.7 to 12.4 percent; in San Jose from 18.5 to 14.8 percent; in Omaha from 8.1 to 4.3 percent; in Los Angeles from 7.5 to 7.0 percent; in Denver from 3.8 to 2.2 percent; and in Dallas from 2.7 to 1.3 percent. Only in San Antonio did the percentage of methamphetamine detections increase over 1995, from 1.5 percent to 2.1 percent. o In keeping with trends of recent years, adult females showed greater methamphetamine use than males. Females continued to lead males in terms of methamphetamine test positive percentages by 2 to 10 percentage points in most of the Western sites. Only in Denver did a larger fraction of males test positive for methamphetamine than females. The gap between females and males increased between 1995 and 1996 in five Western sites, and declined in three. Far greater numbers of females were arrested on prostitution charges than males, and drug test positive rates were often the highest in this charge category. However, even when female and male methamphetamine use was compared across similar charge categories, females were detected as methamphetamine users more frequently than males. o White arrestees continued to use methamphetamines in greater percentages than blacks and Hispanics. In San Diego in 1996, about 47 percent of the white arrestees were detected as recent methamphetamine users, compared to 25 percent of Hispanic arrestees and 10 percent of black arrestees. Note that in San Diego and Los Angeles, Asian arrestees had the second highest test positive percentage (behind whites), although these figures were based on a total of 41 and 28 Asian arrestees, respectively. The broad declines in methamphetamine use were approximately proportionate across racial and ethnic groups, although Los Angeles provided one exception. Overall, Los Angeles dropped from 7.5 percent methamphetamine positives in 1995 to 7.0 percent in 1996. This drop appears to have been driven by a change among blacks. The percentages of whites and Hispanics detected as recent methamphetamine users in Los Angeles increased, while the percentage of black users declined. Conclusions Although the data presented here cannot resolve the issue of how drugs and crime are linked, they clearly demonstrate that the relationship is strong and enduring. A median 68 percent of arrestees test positive for at least one drug at arrest. Moreover, DUF data likely understate recent drug use by arrestees, as urinalysis can only reliably detect drugs for approximately 48 to 72 hours after use. These data also indicate that regional and local trends can depart substantially from the national trend. In terms of general trends, the sharp declines in cocaine use that have occurred in some cities have been largely offset by substantial increases in marijuana use. Marijuana appears to have broadly replaced cocaine as the drug of choice among arrestees. At the disaggregated level, DUF data also suggest that there are significant regional, gender, and age cohort variations in drug use patterns that must be monitored carefully. Regionally, methamphetamine use among arrestees still appears to be confined to Western States and, in most cases, has abated substantially. Opiate use also appears geographically concentrated in the largest DUF cities and selected Western sites. Growth in drug use appears strongly in the youngest adult cohorts (ages 15 to 20), while older cohorts are generally experiencing a slow tapering of use, particularly with cocaine. Females continue to be more frequent consumers of cocaine and methamphetamine than males. The consistently large fraction of individuals testing positive for drugs at arrest and the substantial local variations in drug patterns combine to suggest that (1) point of arrest is an appropriate stage of intervention with respect to addressing substance abuse and (2) communities would benefit from having local knowledge about substance abuse patterns among their arrestees. NIJ is working to address both of these issues. NIJ is funding a demonstration program in Birmingham called Breaking the Cycle that offers universal testing and needs assessments for substance abuse at arrests, followed by the development of a tailored program of treatment, sanctions, and supervision. NIJ is also supporting the evaluation of this program to assess the program's impact on individual arrestees and on the community. If successful, comprehensive substance abuse intervention at arrest could result in reduced drug use, reduced recidivism, and improved functioning in areas such as employment and education. NIJ is also supporting the evaluation of a second program called Operation Drug TEST (Testing, Effective Sanctions, Treatment) which will operate similarly with Federal arrestees. Finally, the President has submitted a budget request to Congress that would allow NIJ to reengineer DUF into the 75-site ADAM Program. ADAM sites will be located in large urban areas and will collect data from arrestees four times per year. In addition, each ADAM site will conduct outreach data collection in a suburban, rural, Indian territory, or other arrestee population on an annual basis. Combined, Breaking the Cycle, Operation Drug TEST, and ADAM promise to greatly increase our understanding of appropriate community-level interventions against arrestee drug use and provide the research platform with which to measure progress. K. Jack Riley, Ph.D. with Nora Fitzgerald, Gabrielle M. Kyle, and Shu-Ahn Li Office of Research and Evaluation, NIJ ------------------------------ Methodology Each quarter, trained local DUF staff obtain voluntary and anonymous urine specimens and interviews from adult arrestees and juvenile arrestees/detainees who have been in a booking facility for not more than 48 hours. Data and urine samples are collected at selected booking facilities throughout the United States. At each of the 23 adult sites, approximately 225 adult males are interviewed each quarter. In all except 2 sites, approximately 100 adult females are also sampled. At 12 of these sites, data and samples are collected each quarter for approximately 100 juvenile male arrestees/detainees. Ten of these sites also collect data on female juvenile arrestees/detainees. However, given the small sample size of DUF female juvenile arrestees/detainees, these data are not presented here. Response rates for both adults and juveniles are consistently high. More than 90 percent of the total sample consent to be interviewed and more than 80 percent agree to provide a urine specimen. Adult arrestees are selected at the discretion of site personnel, who are guided by a target sample size and crime charge priority system. To obtain samples of adult male arrestees with a sufficient distribution of serious arrest charges, DUF interviewers, where possible, place a priority on felony arrestees and those arrested for offenses other than the sale or possession of drugs. Analyses have shown that those arrested for drug offenses are more likely than other arrestees to be using drugs; as a result, DUF statistics are likely minimum estimates of drug use among the population of those adults arrested for serious offenses. With the exception of Omaha, where all arrestees are included to obtain a sample of sufficient size, males charged with driving offenses generally are excluded from the sample due to DUF's emphasis on more serious crimes. Because they are fewer in number, all adult female arrestees and all juvenile male and female arrestees/detainees brought to the booking center during the data collection period are included in the DUF sample, regardless of the charge. At most sites, adult and juvenile catchment areas are identical. In 10 adult sites, the catchment area is the entire city. In another 10 adult sites and in 9 of 12 juvenile sites, it is the entire county or parish. (The city of Denver is Denver County in its entirety.) However, in Birmingham and San Diego, the catchment area includes the city and part of the county. St. Louis' catchment includes only the city and excludes the surrounding county; Los Angeles' comprises parts of the city and county. The Washington, D.C., catchment includes the entire District of Columbia. All urine specimens are sent to a central laboratory for analysis for 10 drugs: cocaine, opiates, marijuana, phencyclidine (PCP), methadone, benzodiazepines, methaqualone, propoxyphene, barbiturates, and amphetamines. All positive results for amphetamines are confirmed by gas chromatography to eliminate those caused only by over-the-counter medications. For most drugs, urinalysis can detect use within the previous 2 to 3 days; use of marijuana and PCP can sometimes be detected several weeks after use. ------------------------------ DUF Juveniles, 1991-1996 In recent years, DUF data have indicated substantial increases in juvenile drug use. The changes in juvenile drug use patterns have registered broadly across drug categories, with large percentage increases occurring in marijuana, hallucinogens, and amphetamines.(1) Prevalence trends, however, tell only a portion of the story. This section presents an overview of additional issues in the DUF juvenile population. Age and Initiation Alcohol and tobacco are typically the first drugs that DUF juveniles try. The mean age for first use of tobacco was 12.3 in 1996. For alcohol, the mean age was 12.5 in 1996. In contrast, the mean age of first use in 1996 was 13.0 for marijuana, 14.1 for LSD, and 14.4 for cocaine. Age of initiation into marijuana dropped significantly between 1991 and 1995, before increasing in 1996. Self-report of whether a juvenile has "ever tried" a substance suggests that differentiated progression to other drug use occurs. Juveniles who report having tried alcohol initiate, on average, marijuana use at a younger age (12.9 years) than juveniles who have not used alcohol (13.6 years). A similar relationship holds again between cocaine use and whether marijuana has been tried. Among those who have never tried marijuana, the mean age for cocaine initiation is 14.9 years, compared to a mean age for cocaine initiation of 14.3 years for those who have tried marijuana. Living Arrangements For the years 1991-1996, approximately 63 percent of DUF juveniles who answered the question reported living in a household where the father was not present. About 21 percent of DUF juveniles reported living in a household with both natural parents, and about 11 percent reported living with one natural parent and one stepparent. Almost 14 percent reported living in households headed by nonparental relatives such as grandparents and siblings and 5 percent reported living with unrelated guardians or in institutional settings. The percentages did not change substantially during the period, although the fraction living with their mothers only declined slightly, and the fraction living with nonparental relatives increased slightly. Urinalysis reveals that children living with unrelated guardians and in institutions test positive for drugs at much lower rates than children living in households headed by relatives, including parents. Lower rates may be due to increased levels of supervision experienced by this group. However, among juveniles living with relatives, the lowest test positive rates are among those living with both parents or a parent and stepparent. The percentage testing positive increases for those living with a mother or father only, and is highest among those living with nonparental relatives. Thus, while family structure has remained relatively constant in the DUF juvenile population, a relationship between family structure and drug test results appears to persist. Self-Report and Test Results Urinalysis results for marijuana match prevalence self-reports for about 79 percent of the juveniles. Some respondents' test results will not match their responses because of deception. Deceptive respondents may fear that their disclosures will be used against them, or that the government does not have the right to collect such information. In other cases, however, the respondents may not be aware of what they have consumed because they have been taken advantage of by a seller or because they have consumed adulterated drugs. Still others may simply overstate their drug use to project an image to peers and interviewers. From the standpoint of interpreting the DUF data, and from improving our understanding of how juveniles respond to the DUF interview instrument, it would be useful to determine what characteristics "true positives" share that are distinct from the characteristics of "falsely reported positives." An analytic technique called discriminant analysis can help with such categorization. Discriminant analysis correctly classified 74 percent of about 20,000 cases. About 75 percent of the juveniles who claimed not to have used marijuana in the last 72 hours, but tested positive, were categorized correctly in the falsely reported negative group. The model correctly classified about 70 percent of the true negatives, 72 percent of the true positives, and 94 percent of the falsely reported positives (those who said they used marijuana recently, but tested negative). Unlike regression coefficients, discriminant function coefficients cannot be directly interpreted. Larger standardized canonical discriminant function coefficients make larger contributions to a variable's score, and the aggregated scores determine categorization. As table 1 on the previous page indicates, it appears that cocaine urine test results, the number of drugs for which a juvenile tests positive, reports of no marijuana use during the past 30 days, and reports of never having used marijuana make particularly strong contributions to categorizing juveniles with respect to deception about marijuana use. By charge, only "drug possession" significantly aided classification. Juveniles held on drug possession charges may be concerned that self-report of drug use may be used against them. Interestingly, the variable drugsale, which indicates arrest on a drug sales charge, did not aid classification. Among the income variables, the drug dealing and legal income variables helped classification with regard to deception. Of particular interest in this model is the fact that interview year (intyr) made a statistically strong contribution to categorizing truthfulness. At one level, the interview year effect could be an indication of shifting norms about marijuana use. Deceptive behavior could vary from year to year as a function of how society, and peers, judge drug use. Periods of increasing societal disapproval of drug use may be accompanied by increased deception, and periods of decreasing disapproval may be accompanied by deceased deception. At another level, changes in deceptive reporting by year have important implications for analysts using DUF data. (1) Marijuana and amphetamine results are from urinalysis. Hallucinogen data are based on self-reports. K. Jack Riley, Ph.D., and Angela Moore Parmley, MPA Office of Research and Evaluation, NIJ ------------------------------ Recidivism Rates in the DUF Sample: Prior Arrests and Current Use of Drugs DUF data have always provided information regarding the current offense of the arrestee -- the charge under which the respondent is currently detained. What has been lacking is a link to prior arrest information; because the DUF interview is anonymous, it has not routinely been linked with official criminal histories. In a recent modification to the DUF interview, a series of questions was added to obtain self-report information regarding arrests during the prior 12 months. The questions ask about the number of arrests, the type of offense, and whether the respondent was sentenced to serve jail and/or prison time as a result. These data, coupled with urinalysis for drug use, provide important information about the role which drug use plays in the likelihood of recidivism. Overview Beginning in July of 1995, arrestees interviewed as part of the DUF program have been asked a series of questions regarding their prior arrest history. The questions are designed to solicit information about arrests during the 12 months prior to the interview. An additional question addresses the issues of whether the respondent served any jail or prison time during the same 12 months. o During the past 12 months, have you been arrested and booked for breaking the law, whether or not you were guilty? a. How many times during the past 12 months? b. What were the charges? o Have you served time in the past 12 months? Additional queries and probes are designed to clarify the questions and categorize responses. For this report, combined adult and juvenile arrestee data from all DUF sites for 1996 were analyzed. Figure 2 shows the overall percentages of male and female arrestees reporting at least one previous arrest in 12 months. Forty-one percent of male arrestees reported that they had been arrested previously at least once during the 12 months prior to the DUF interview, more than 40 percent of which had been arrested two or more times previously. Among female arrestees, the percentage was 33 percent, 42 percent of which reported two or more prior arrests in 12 months. White arrestees were more likely than black, Hispanic, or other arrestees to report having been arrested previously during the 12 months prior to the DUF interview. Forty-two percent of white arrestees had been arrested previously, compared to 38 percent of black arrestees, 36 percent of Hispanic arrestees, and 38 percent among other arrestees. The proportion of arrestees reporting at least one arrest during the 12 months prior to the interview was highest among younger arrestees, as figure 3 shows. Forty-seven percent of arrestees age 20 and under reported that they had been arrested previously at least once in the last 12 months. Among arrestees age 31 and older, only 34 percent reported having been arrested previously in the last 12 months. The number of arrests during the period 12 months prior to the interview did not vary significantly by offense type, however, as shown in figure 4. Figure 5 shows urinalysis results for those reporting a previous arrest in the last 12 months. The most striking difference is in terms of cocaine use. Among those who stated that they had not been arrested previously, only 33 percent tested positive for cocaine, while 44 percent of those reporting two or more arrests tested positive for cocaine. A similar but slightly smaller contrast is shown for marijuana test results. Results for methamphetamine, opiates, and PCP are inconclusive. The bivariate relationships linking previous arrests with gender, age, race, and cocaine use were sustained in a multivariate logistic regression analysis. The strongest effect detected in the analysis linked previous arrests with cocaine urinalysis results. Weaker (but statistically significant) effects were found for each of the other variables. The multivariate logistic regression analysis demonstrates that arrestees who test positive for cocaine are about 1 1/2 times as likely as others to have been arrested previously. This effect is independent of differences by gender, age, and race. Discussion These analyses support the general hypothesis linking drug use and other criminal activity; in large numbers, persons who use cocaine come to the repeated attention of the criminal justice system. These data suggest that arrests that occurred during the previous 12 months were important "missed opportunities" for assessing and intervening in the arrestee's use of drugs. Failure to have successfully identified those arrestees at risk for reoffending and associated drug use is manifested in the eventual rearrest of these individuals and their concurrent involvement in drugs as evidenced by urinalysis at the time of arrest. Thomas E. Feucht, Ph.D., and Gabrielle M. Kyle Office of Research and Evaluation, NIJ ------------------------------ TELEDUF Pilot Project Cleveland State University has been developing new methodologies for community research using telecommunications technologies. The TELEDUF Pilot Project sought to (1) develop an automated telephone interviewing system that could administer the DUF instrument, (2) pilot test the methodology with adult male, adult female, and juvenile male DUF respondents in Cleveland, Ohio, and (3) compare a cohort of traditional DUF respondents to a cohort of TELEDUF respondents. After obtaining the interview subject's consent, the researcher called the system and entered in an identification. The telephone was then handed to the arrestee, who listened to the series of prerecorded questions and answered by touch tone response. Respondents could skip questions by pressing the pound key. Using this process, up to four simultaneous interviews were conducted; data were immediately retrievable on completion of the interview. The project compared telephone- and human-administered response rates, rates of self-reported drug use, and the validity of self-report using urinalysis. As table 2 indicates, response rates did not vary significantly by method. Self-reported drug use was comparable within each group comparing the two modes of administration. There was a trend within the juvenile male group for higher self-reported drug use using the TELEDUF system. See tables 3, 4, and 5. Concordance rates were calculated comparing the self-reported cocaine use to urine results. For adult males, the concordance rate for TELEDUF interviews was 67% compared to 70% for the traditional DUF interviews. See table 6. Concordance rates for adult females were also very comparable. Concordance rate for juveniles indicated slightly less concordance for the TELEDUF group, although there was no self-reported cocaine use in either group. See table 7. This system is not without limits or disadvantages. Administration is impersonal and there is no opportunity to capture nonverbal or emotional data. Further, open-ended responses cannot be accepted. The number of response categories for each item must be limited, since respondents cannot listen to a long list of possible responses. Finally, in order for the system to work, it is clear that respondents need to be able to listen to the telephone and press a telephone key. However, TELEDUF does offer advantages over traditional administration of the DUF interview. Advantages: o Reliable delivery of questions/skip patterns. o Complete confidentiality. o Rapid-response data collection via direct data entry. o Laboratory data could be called and matched. o Simultaneous interviews. o Possible simultaneous languages. o Can add/change questions on command with few training implications. We conclude from this pilot project that there are possible advantages to the use of automated telephone-administered interviews. Among these advantages, the reliable delivery of questions and administration of skip patterns is appealing. Performing simultaneous interviews and direct data entry could have potential cost implications for major multisite research implementations. The TELEDUF pilot project supports the further consideration of automated telephone-administered designs in sites where phones are readily accessible and for very structured research protocols. Implications: o May be a useful tool for collecting self-report data in sites where phones are easily accessible. o May represent a rapid-response methodology for multisite collaborations. Sonia Alemagno, Ph.D., Stephanie Wolfe, John Butts, and Robert Pace ------------------------------ Guide to DUF Site Data Pages Pages 24 to 59 of this report contain one-page summaries of essential data from each of the juvenile and adult DUF sites. Results are for the sample of booked adult male and female arrestees and juvenile male arrestees/detainees. Data on adult female arrestees were collected at all but two sites. Sample selection procedures differ for male and female arrestees and for juvenile male arrestees/detainees (see Methodology). Catchment Area, Sample Size, Age, and Race Catchment areas for DUF collection of adult data vary across sites. In most cases, the entire county or city constitutes the area from which arrestees are drawn. At a few sites, the catchment area is a specific precinct or set of precincts within the city or county. Sample size indicates the number of adult arrestees and juvenile arrestees/detainees who completed the DUF interview and provided a urine specimen. Among juveniles, no distinction is made between arrestees and detainees. The sample is the total across four quarterly collection periods in 1996. For some results, sample size is reduced slightly due to missing data. Percentage distributions by age and race are shown for each site. Typically, these data are obtained from official booking records at the facility. While the minimum age of arrestees at adult facilities is usually 18, a small number of persons under the age of 18 are encountered among the adult arrestees. Because of this, the youngest age category for adults is reported as 15 to 20. For similar reasons, the oldest age category among juveniles is 17 to 18. For adult male arrestees, the percentage distribution of the offense at arrest reflects the facility population at the time of DUF data collection, subject to selection guidelines stipulated by NIJ. Specifically, adult males arrested for drug offenses are limited to not more than 20 percent of the sample, and a priority is placed on felony arrestees over misdemeanor and ordinance arrestees.(1) For adult female arrestees and for juveniles, the percentage distribution of the offense at arrest is a function of the composition of the population of arrestees/detainees at the facility at the time of collection. No categories of female or juvenile arrestees are excluded or under-sampled. Urinalysis Results by Offense at Arrest As with the age and race of the adult arrestee and juvenile arrestee/detainee, information on the offense at arrest is obtained from official booking records. For reporting purposes, violent offenses include robbery, assault, weapons offenses, extortion, homicide, kidnapping, manslaughter, sexual assault, and rape. Property offenses include larceny/theft, burglary, motor vehicle theft, arson, possession of stolen property, bribery, and pickpocketing. Drug offenses comprise drug possession and drug sales. Other offenses are public peace offenses, flight/escape, traffic offenses (primarily driving while intoxicated/driving under the influence), being under the influence of a controlled substance, probation or parole violation, family offenses, resisting arrest, sex offenses (other than sexual assault or rape), liquor law violations, obscenity, gambling, embezzlement, and other miscellaneous offenses including, for juveniles, violating curfew or home supervision. For females, prostitution is reported in a separate category. Urinalysis results for cocaine, marijuana, and any drug are shown for the total sample of arrestees, for each broad category of offenses, and for specific offenses. The number of arrestees in each category is shown in parentheses. Specific offenses were chosen because they constitute significant numbers of arrestees across sites. Though the distribution of arrestees across offense categories varies from one site to another, the same offenses are shown for each site. Unavoidably, for some sites the number of cases in a specific offense category is very small. Drug Use by Adult Male and Female Booked Arrestees and Male Juvenile Arrestees/Detainees This bar graph shows the percentage who were urine positive for drugs at the time of the DUF interview according to EMITTM. The graph shows the percentage positive for any of 10 drugs (listed at the bottom of each page), for cocaine, marijuana, and opiates, and for multiple drugs. The percentage positive is also tabulated by age and race. Drug Use Trends Among Booked Arrestees DUF data collection for adult arrestees began in 1987 in 12 sites while that for juvenile male arrestees/detainees began in 1988 at 2 sites. By 1990, this effort had expanded to include adults and juveniles in nearly all the sites reported here. Some, like Atlanta, started data collection in 1991. The last site to begin collecting juvenile data was Denver in 1991. To provide comparability throughout the tables, data series are reported beginning with 1992. Gaps in the line graphs represent periods when data were not collected or when sample size was insufficient for monitoring purposes. Graphs for both male and female adult arrestees are presented for 21 sites; data for male arrestees only are shown for 2 sites. Drugs Positive Among Male Juveniles by School Attendance Urinalysis results for any drug, for cocaine, marijuana, and for multiple drugs are shown by school attendance for juvenile arrestees/detainees. Data on school attendance are based on self-report and do not provide a measure of frequency of attendance or academic achievement. Results in the graph are presented for those still in school and for those who no longer attend but have not graduated. Results are also presented for arrestees/detainees who have successfully completed their schooling, although this number is typically too small for meaningful analysis. Juveniles interviewed during the summer months are asked if they plan to attend when school resumes in the fall. (1) NIJ is reviewing the sampling procedures currently used in the DUF program. Any change in sampling procedures will be noted here in future reports, together with implications of the change for comparison of series data. ------------------------------ Atlanta Annual Report This report focuses on Atlanta drug trends indicated by DUF data and some implications of those trends. Cocaine has continued to dominate the local drug market since 1992. While cocaine use leveled off in 1995, a trend also seen in other U.S. urban areas, the percentage of positive urine screens for cocaine reached a peak of 65.5 percent during the third quarter of 1996 -- coinciding with the 1996 Summer Olympic Games in Atlanta. A downward trend was apparent during the fourth quarter of 1996 (59.5 percent) to a level similar to what was reported in 1995. Similar to the cocaine trends, marijuana reports stabilized in 1995. However, this trend changed dramatically during the second half of 1996. Positive urine screens for marijuana reached levels higher than those reported at any time since 1992. While 14.5 percent of the individuals tested in 1992 had positive urine screens, this percentage had increased to 34.5 percent in the fourth quarter of 1996. Cocaine When exploring gender differences, it becomes clear that rates of positive urine screens for cocaine among women have been exceeding those among men since the second half of 1994. This may be due to practices that result in disproportionate arrests of prostitutes. During 1996, among female cocaine users approximately one out of five arrests involved prostitution, most frequently street prostitution -- a type easily recognized by local law enforcement officials. During the second half of 1996, among women, rates of positive urine screens for cocaine among whites (75.5 percent) exceeded those among African Americans (61.5 percent) by 14 percent (figure 1). The racial distribution during 1996 among male cocaine users reveals that rates of positive urine screens for cocaine among African Americans (60.3 percent) exceeded those among whites (42.0 percent) by 18.3 percent. Positive cocaine screens were most common among individuals 26 and older. This suggests an aging cohort of cocaine users, and may explain the leveling off in cocaine reports. Marijuana Rates of marijuana positives among males exceeded those among females for all reported quarters in 1996. In addition, rates tended to be highest among African Americans and individuals between the ages of 15 and 20. It appears that an increasing number of youth are becoming involved with marijuana use, an important finding from a policy point of view. Drug-Crime Connection When comparing the offense categories among male arrestees with positive cocaine screens to those who had positive marijuana screens, marijuana users are more likely to have been arrested for violent and drug offenses. This contradicts the public stereotype of the prevalence of violent offenses among cocaine users. Both males and females with positive cocaine screens exceeded those with positive marijuana screens only in the category of property offenses (figure 2). Kirk W. Elifson, Ph.D. In the third quarter of 1996, the Cleveland DUF site added a supplement to the DUF interview to examine self-reported risk factors related to tobacco, alcohol, drug use, social risk behaviors, mental health, and family support. A total of 80 juvenile male supplemental interviews were conducted. The majority of respondents reported that they smoke cigarettes and drink alcohol. One-third of respondents reported that they feel that they should cut down on their drinking. Almost half of those interviewed felt that they should cut down on their drug use. The outcomes of alcohol and drug use reported by the respondents were of great concern. About one-fourth of the respondents reported that they have been in an alcohol-related fight. Specific findings include: o 66 percent smoke cigarettes. o 88 percent report drinking alcohol. o 32 percent feel they should cut down on alcohol. o 22 percent have been in alcohol-related fights. o 13 percent have been in an alcohol-related car accident. o 44 percent use drugs more than once a week. o 47 percent feel that they should cut down on their drug use. o 18 percent have gotten into trouble due to drug use. A preliminary look at the mental health status of the respondents indicated that almost one-half of the respondents report symptoms of depression such as feeling sad most of the day or feeling like things are hopeless. One-third of the respondents reported that they use drugs to forget about their problems. Findings related to mental health include: o 47 percent feel sad most of the day. o 43 percent feel like things are hopeless. o 34 percent drink alcohol or use drugs to forget about problems. Clearly, this population is sexually active and engages in a self-reported high rate of unprotected sex. Most of the respondents admit that they have had sex while drunk or high. Findings related to sexual behavior include: o 95 percent have had sexual intercourse. o 65 percent have had unprotected sex. o 75 percent have had sex while drunk or high. In this preliminary pilot project, it appears that juvenile respondents were self-aware of the behavioral implications of substance use. Further research is needed that examines the interface of substance use with school, home, and neighborhood experiences. Multidimensional public health interventions in this target population should be considered. Sonia Alemagno, Ph.D., and Stephanie Wolfe ------------------------------ Denver Annual Report During the 7 years that the Colorado Division of Criminal Justice (DCJ) has been involved with the DUF program, a number of State agencies have used DUF data. In early 1996 the Piton Foundation contacted DCJ to request access to all of the Denver DUF data. The Piton Foundation was formed in the 1970s with the mission of strengthening Denver's low-income neighborhoods and families. In subsequent conversations, the Piton Foundation and DCJ discussed the opportunity to use DUF as a vehicle for research. DCJ joined with the Piton Foundation to submit a juvenile violence prevention proposal to NIJ. The Juvenile Violence Prevention Study will add several questionnaires to DUF. These questionnaires will provide information that will help law enforcement officials better understand the family and socioeconomic histories that precede criminal activity and the community context in which crimes occur in the City and County of Denver. The first step in this project is indepth analysis of historical DUF data. This analysis is currently underway. However, the following bullet points offer a preliminary snapshot of what the study has found in the 1995 juvenile DUF data. o In 1995 one in five juveniles still in school was at least 1 year behind their age-appropriate grade; more than 16 percent of Denver DUF juveniles were 1 year behind, 2 percent were 2 years behind, and 2 percent were 3 or more years behind. o 45 percent of juvenile detainees reported at some point having possessed a gun (48 percent male detainees, 37 percent female). o 56 percent of juvenile detainees reported that it is easy to get a gun illegally; 59 percent reported that there were lots of guns on the street in the neighborhood in which they lived. o More than one in four detainees (28 percent) reported that most of their friends have guns. o 41 percent of juvenile detainees believed a gun was needed for protection in their neighborhood; 23 percent believed their crowd respected them more if they had a gun. o Few juvenile detainees (5 percent) thought their friends looked down on them if they did not carry a gun and even fewer (4 percent) thought someone wounded by a gun was tough. o 60 percent reported having been threatened with a gun at some point in the past; 57 percent had been threatened by some other weapon. o Almost half of juvenile detainees (48 percent) reported they had been shot at, though relatively few (8 percent) had been injured by gunshot. o A larger proportion (27 percent) reported having been injured by a weapon other than a gun. o 59 percent of juvenile detainees, including 53 percent of female detainees, reported having been beaten up. o Combined, two-thirds (66.4 percent) of juvenile detainees reported having been injured by a gun or other weapon, or beaten up. o 18 percent reported they had been robbed at some point. Chris Webster ------------------------------ Detroit Annual Report The Detroit DUF site began using the Gun Addendum at the beginning of 1995. Through the fourth quarter of 1996, 1,375 male offenders were interviewed on questions dealing with gun ownership (or possession), gun acquisition, attitudes about having guns, and various experiences of the offender. Across all offenders, 57 percent tested positive for at least one drug, with one-third of the total offender group testing positive for marijuana only and 24 percent testing positive for other drugs. With regard to certain life experiences, this group of offenders reported a considerable level of victimization, with 52 percent indicating that they had been robbed, 54 percent had been shot at, and 25 percent had been injured by a gunshot. While these descriptions are informative, it is also instructive to consider their relationship with gun ownership (figure 1). Most notable is the lack of a relationship of gun ownership with drug use. Fifty-six percent of those who have owned a gun tested positive for any drug use, while 58 percent of those who have not owned a gun tested positive. Similarly, drug users were not any more likely to have owned a gun, with 33 percent of those testing positive reporting owning a gun and 35 percent of those not testing positive owning a gun. Figure 1 also notes some interesting and significant differences in gun-related victimization between gun owners and non-gun-owners. Sixty percent of those owning a gun reported being robbed compared to 48 percent of those not owning a gun. Almost 7 in 10 (69 percent) of the gun owners had been previously shot at compared to 46 percent of the non-gun-owners. About one-third (32 percent) of the gun owners had been injured by gunshots compared to 22 percent of the non-gun-owners. Those who had reported selling drugs were not significantly more likely to test positive for recent drug use (63 percent) than those who did not report having sold illegal drugs (57 percent) in Detroit (figure 2). Although the numbers are small, and the vast majority of those who had sold drugs did not indicate that they had also committed a gun crime, those offenders reporting previous selling of illegal drugs were more likely to have also indicated that they committed an offense with a gun than those who did not report selling drugs. Those offenders who reported selling drugs were significantly more likely to have owned a gun (58 percent compared to 31 percent) and to have been shot at (76 percent compared to 52 percent) when compared with those who did not report that they had previously sold illegal drugs. Drug sellers were also more likely than other offenders to have been robbed (62 percent compared to 51 percent) and injured by gunshot (34 percent compared to 24 percent), although these findings were not statistically significant. One of the more interesting aspects of this analysis involves a consideration of the differences between those who report having committed a gun crime and other offenders in Detroit. While this analysis may be instructive, it is based upon a relatively small number of offenders and thus caution should be exercised in generalizations. This group of more serious offenders was significantly more likely to report having themselves been victimized in gun offenses through being robbed (71 percent compared to 51 percent), shot at (83 percent compared to 53 percent) and injured by gunshot (45 percent compared to 24 percent). Such findings point to the importance of considering the lifestyles, activities, and circumstances that place individuals in situations to be both offenders and victims of criminal violence. Timothy S. Bynum, Ph.D., Tracy A. O'Connell, Nancy G. Becker, and Ed Banks ------------------------------ Los Angeles Annual Report Since Los Angeles became a DUF site in 1989, cocaine has been the drug most commonly detected by urinalysis of arrestees. The overall rate of cocaine positives has gone down slightly among both adult men and women, while positive rates for opiates, marijuana, and PCP have held fairly steady. Among adult men, the positive rates for amphetamines or methamphetamines have declined in recent years.1 (See figure 1 for methamphetamine trends by gender.) Test results show that among the women, rates of methamphetamine use have been rising over the past 3 years. Most of this increase has occurred in the subpopulation of Hispanic women, although the use of speed among Hispanics in general has been growing in the Los Angeles DUF sample. Local law enforcement officials were interested to see if the "ice" type of methamphetamine would make its way to Los Angeles from Hawaii. However, the methamphetamine addendum (see San Diego site report, page 71) administered to the DUF sample has not revealed a large population of ice users. Current data still show that few users of speed in Los Angeles use the ice form, perhaps due to reports of negative consequences that have spread by word of mouth and the media. The growing number of Hispanic methamphetamine users is linked by law enforcement personnel to the increased trafficking by Mexican nationals and by Mexican-Americans. The trafficking trend was noted several years ago by the Drug Enforcement Agency and Bureau of Narcotics Enforcement. The recent rise in methamphetamine use among Hispanics supports the suspected link. Among juvenile males in the Los Angeles DUF sample, there has been a sharp increase in marijuana positive rates since 1993. From 1990 to 1995, there has been a slow but steady increase in the positive rates of methamphetamine among juveniles. In 1996, approximately 5 percent of juveniles tested positive for methamphetamine. (1) It is important to note that the arrestees themselves often do not distinguish between the different types of speed. Kiku Annon, Ph.D., and Jeff Annon ------------------------------ Manhattan Annual Report DUF Manhattan staff have obtained support from the National Institute on Drug Abuse and the National Institute of Justice to analyze and document various trends among DUF arrestees in Manhattan and the Nation. Overall, the proportion of DUF Manhattan arrestees (age 18 and older) detected as users of cocaine, heroin, and marijuana has not changed dramatically from 1987 to 1996. The proportion of Manhattan arrestees is always among the top 3 cities (of 23 DUF sites) in the percent detected as cocaine/crack users and as opiate users. Major changes in drug use patterns have occurred among youthful birth cohorts (figure 1); age and period effects are generally not significant when birth cohort is held constant. Thus, when compared with their counterparts born in the 1950s and 1960s, birth cohorts of Manhattan arrestees born in the 1970s (and who reached age 18 after 1987 -- mainly in the 1990s): o Showed dramatic declines in the proportion detected as positive for cocaine/crack; this rate dropped from two-thirds of those born in the 1960s to about one-fourth or less of those born from 1973 to 1978. o Showed very important decreases in the proportion detected as opiate users; this rate dropped from about one-fourth of those born from 1950 to 1969 to less than one-tenth of those born from 1972 to 1979. Figure 1 illustrates these findings. While about one-fifth of DUF Manhattan arrestees born before 1970 are positive for marijuana at arrest, younger cohorts usually exhibit higher rates of marijuana use. Specifically, about one-third of DUF Manhattan arrestees born from 1970 to 1975 and half or more of those born from 1976 to 1979 were detected as marijuana users at arrest. These data show that among arrestees born before 1970, cocaine and crack remains the primary drug of abuse. This group represents older generations of drug users, who began their drug use careers before or during the height of the crack-cocaine epidemic (1984-89) in New York City. These cohorts have not significantly reduced their use of cocaine/crack as detected by DUF; they constitute the major cohorts of cocaine-dependent persons in New York City. While lower proportions are detected as heroin users, persons born in the 1950s -- who were youths during the heroin epidemic (1967-73) -- remain the core of the heroin injector pool in Manhattan. Among younger cohorts, the story is quite different. Youths coming of age in the 1990s appear to have avoided heroin and cocaine/crack in young adulthood -- but increasing proportions (and a majority born from 1975 to 1979) are being detected as marijuana users. Thus, the trend data in the DUF program provides a marvelous opportunity to monitor trends in specific drugs used among varied birth cohorts in Manhattan and other cities. Bruce D. Johnson, Ph.D. ------------------------------ Omaha Annual Report The relationship between drug use and gun ownership and different types of offenders has important implications at the national and local levels.(1) At the national level, these data speak to the commonly held assumption that drug use is highest among more serious offenders (i.e., the drugs/violent crime connection) and that guns are owned, carried, and used primarily by more serious offenders. In turn, these assumptions may affect law enforcement activities at a local level. If the assumptions are true, then law enforcement officers are aware of the risks they face when encountering different types of offenders, but if they are inaccurate, law enforcement officers may be placing themselves at risk for harm without knowing it. Results from the DUF interview and Gun Addendum help explore these assumptions. Using Omaha DUF data from 1994 to 1996, figure 1 shows that similar proportions of offenders test positive for cocaine use across offense type. To examine the relationship between gun ownership and drug use and offense type in Omaha, data from one cycle of the 1995 Gun Addendum were analyzed. 43 percent of respondents (N=274) reported ever owning a gun, and the most prevalent types were handguns and semiautomatic pistols/rifles.(2) To more clearly distinguish the relationship between drug use and gun ownership, gun ownership was compared to positive urine tests for cocaine. As shown by figure 2, a majority of cocaine-positive arrestees did not report owning a gun (81 percent). The proportions of gun ownership across positive and negative cocaine tests were nearly identical (19 percent and 22 percent respectively). Next, gun ownership was compared across misdemeanors and felonies, also shown in figure 2. More than 50 percent of felons own or have owned a gun, whereas 36 percent of misdemeanor offenders own or have owned a gun. These results illustrate that many arrestees, regardless of drug use and arrest charge, are involved in firearms. These findings should be assessed over time to study the implications for law enforcement officers, the courts, and other criminal justice organizations. (1) The sample of offenders in Omaha is unique compared to other DUF sites because nearly all adult male arrestees are eligible to participate in the DUF program. Consequently, the range of offenses within the Omaha sample includes traffic offenses as well as nontraffic misdemeanors and felony offenses. (2) Gun ownership was determined by ever owning a gun because only two respondents reported having a gun on their persons when they were arrested. This finding indicates that there may be some reliability concerns with portions of the Gun Addendum data. As a preliminary check, weapons charges were compared to whether respondents indicated that they were arrested with a gun in their possession. Only 25 percent of the respondents charged with a weapons charge reported that they had a gun when arrested. This discrepancy may be due to a different type of weapon (i.e., a knife), dishonesty due to the length of the interview, or dishonesty related to the sensitive nature of the material. Denise Herz ------------------------------ Portland Annual Report Use of drugs is astonishingly high in Portland among arrestees, with about 69 percent of combined male and female arrestees testing positive for any drug in 1996. Beginning in 1992, the DUF survey in Portland, Oregon, began showing a rise in cocaine use in the arrestee population. In 1992, the percentage testing positive for cocaine exceeded the percentage testing positive for marijuana. In 1993, 1994, and 1995, the numbers for cocaine use among both male and female arrestees went up steadily. The numbers from samples taken in 1996 indicate that the trend still continues. Among male arrestees, cocaine and marijuana are the two drugs that are most likely to be present in urine samples obtained near the time of arrest. In the second half of 1996, about 40 percent of male arrestees tested positive for cocaine, and about 31 percent of male arrestees tested positive for marijuana. For female arrestees, the popularity of cocaine is even more pronounced than it is for male arrestees. The percentage of female arrestees who test positive for cocaine vastly exceeds the percent testing positive for any other drug. In the second half of 1996, about one-half tested positive for cocaine, while only 25 percent tested positive for marijuana. In late 1994 and early 1995, Portland arrestees participated in the Substance Abuse Needs and Treatment Assessment (SANTA) project. After completing the DUF interview, arrestees were asked additional questions designed to look at substance abuse dependency among arrestees. Almost half (45 percent of the male arrestees and 42 percent of female arrestees) were diagnosed as dependent on alcohol or drugs. Based on the number of people arrested each month in Portland, these percentages add up to approximately 1,271 male arrestees and 292 female arrestees each month who need treatment for drug and/or alcohol dependence. About 64 percent of male arrestees and 63 percent of female arrestees who were diagnosed as drug dependent said they had not received any substance abuse treatment in the past year. As a direct result of the DUF survey, supplemented by SANTA data, the direction of substance abuse treatment may change in Oregon. In January 1997, the Office of Alcohol and Drug Abuse Programs (OADAP), in the Oregon Department of Human Resources, made recommendations based on the large population of arrestees that are categorized as dependent, but have not received treatment. OADAP is recommending that additional treatment resources and/or drug courts be considered as a rational alternative to reduce crime and additional criminal justice system costs. Diane Wiscarson ------------------------------ San Diego Annual Report The San Diego Association of Governments (SANDAG) Criminal Justice Research Division has a unique opportunity to provide timely DUF results to local policymakers because it functions as the criminal justice clearinghouse for the region. Local DUF results have been used to support legislation, to affirm the need for drug treatment, and to justify funding for law enforcement drug prevention and control efforts. In the late 1980s and early 1990s, methamphetamine use among San Diego DUF arrestees was unprecedented among DUF sites. Due to intensive law enforcement efforts and legislation regulating precursor chemicals, methamphetamine use among arrestees subsided temporarily. However, in 1995 and 1996, indications of methamphetamine use again surged not only among DUF arrestees, but in drug treatment admissions, Drug Abuse Warning Network (DAWN) data, lab seizures, and arrests. In response to these ominous reports, a member of the County Board of Supervisors convened a Methamphetamine Strike Force comprised of more than 40 experts from prevention, intervention, treatment, interdiction, and research (including SANDAG) from all levels of government. The strike force was charged with assessing, understanding, and developing an integrated set of recommendations to reduce methamphetamine problems in San Diego County. During the assessment process, SANDAG provided the team with a profile of methamphetamine users and compared recent users with those from 1990. As the meetings progressed, information gaps were noted and SANDAG sought funds to conduct a methamphetamine addendum through the DUF program to further explore methamphetamine use patterns and the dynamics of the illegal drug market. NIJ funded this effort, and 2 weeks following completion of the first cycle of interviews, SANDAG staff presented preliminary findings on 65 admitted methamphetamine users. An Advisory Group, composed of the U.S. Attorney, the District Attorney, Drug and Alcohol Services, the regional narcotics task force, and local law enforcement, had assisted SANDAG in the development of the methamphetamine addendum and reviewed the recent findings. It became apparent that there are characteristics about the manufacturing, distributing, and use of methamphetamine that differ from other illegal drugs and may warrant different intervention strategies. At the final meeting of the strike force, it was recommended that SANDAG seek funds to conduct the DUF-methamphetamine addendum with the treatment population. Commitment to this effort was acknowledged by the county supervisor who initiated the strike force and by the director of Drug and Alcohol Services. Ad hoc committees were formed to carry out the action plans of the strike force. The methamphetamine advisory group convened by SANDAG was asked to perform the role of a research committee, continuing to review and interpret the findings and implications of the methamphetamine addendum results. A guiding principle of the strike force is that "policy must be based on real information that can be targeted, measured, and evaluated." The DUF methamphetamine addendum and analysis demonstrates the utility of DUF as a research platform that informs and shapes policy decisions regarding methamphetamine prevention, intervention, and treatment issues. Susan Pennell ------------------------------ 1987-1988 In 1987 NIJ, with co-funding from the Bureau of Justice Assistance, unveiled DUF. DUF began collecting data and conducting urinalyses among arrestees in eight large cities. Adult male arrestees were interviewed in all eight locations and adult female arrestees in five locations. In 1988 DUF expanded to include juvenile arrestees in five of the DUF sites. A separate interview instrument was created for the juveniles. The first DUF Annual Report was published; special reports included "Injection in Arrestees and CDC Estimates of AIDS," "Regional Differences in Drug Use: Male Arrestees," and "School Dropout among Arrestees." 1989-1994 DUF established 14 new sites for a total of 22 adult male sites in 1989. The following year, DUF expanded to 21 female sites and 11 juvenile sites. Also in 1990, a special addendum, designed to track the spread of ice from Hawaii, was fielded in DUF sites. During 1992 and 1993, DUF operations and funding were consolidated in one organization, NIJ. In 1994, DUF moved to NIJ's Office of Research and Evaluation (ORE). 1995-1997 DUF revised the main instrument and developed a Spanish version of the interview. TELEDUF, a phone version of the DUF instrument, was pilot tested in Cleveland. DUF developed as a research platform, and several special topic addenda were introduced. The Gun Addendum was used at the majority of sites to collect data on the procurement of, use of, and attitudes toward firearms. In collaboration with the Office of National Drug Control Policy (ONDCP), NIJ conducted a special study of cocaine/crack and heroin drug markets and drug procurement patterns in six DUF sites. A third addendum provided data specific to heroin use. In the spring of 1995, NIJ proposed the development of the Arrestee Drug Abuse Monitoring (ADAM) program. The ADAM proposal preserved many of DUF's features, but included a call to expand operations to 75 sites; establish an annual outreach program to nearby and outlying arrestee populations; develop local coordinating councils; and redesign the data collection and sampling methodology. In 1996, the San Diego and Denver DUF sites were awarded the first two DUF challenge grants. San Diego began an examination of methamphetamine markets in five Western DUF sites. Denver initiated a study of community and environmental factors affecting juvenile criminal histories and deviant behavior. In 1997, the President included ADAM in the Administration's 1998 budget request to the Congress. The 10th annual DUF Site Director's Meeting/1st annual ADAM conference was held in Denver, Colorado. NIJ planning work on the transition from DUF to ADAM commenced, including development of a sampling protocol, review of interview instruments, and commencement of process for selecting new ADAM sites. 1998-2001 In fiscal year 1998, ADAM is expected to operate in 35 sites, including 12 new sites and 23 former DUF sites. In the subsequent year, ADAM is expected to enroll 15 additional sites, bringing the system up to 50 sites total. By fiscal year 2000, ADAM expects to be operational in 75 sites. Each site is expected to be collecting outreach data by fiscal year 2001. Building on DUF's successes, NIJ will expand data collection and reengineer DUF's methodology to transform DUF into ADAM. New ADAM program elements will include: o Establishment of an outreach program that will annually collect one additional quarter of data from a targeted population, such as in a suburban, rural, or Indian Nation site, allowing ADAM to monitor the leading and trailing edges of drug abuse. o Development of local coordinating councils that will guide the selection of outreach sites, generate local research projects for execution at the ADAM site, and assume a prominent role in disseminating the site's findings to policy, practitioner, and public constituencies. o Redesign of data collection and sampling methodology to provide policymakers and practitioners with a rigorous basis from which to assess local arrestee drug use and crime patterns and evaluate local policies. ADAM will provide local and state drug policymakers, courts, law enforcement agencies, treatment providers, and prevention specialists with information that can be used to conduct local research and evaluation. Federal agencies responsible for formulating national drug policies will also be interested in ADAM data to assess national trends.