Drug Abuse Warning Network and Arrestee Drug Abuse Monitoring

Denise Herz & Christine Crossland. 21st Century Criminology: A Reference Handbook. Editor: J Mitchell Miller. Sage Publications. 2009.

At the turn of the 20th century, the U.S. government began instituting laws to reduce the availability of illicit drugs and to criminalize their use. The passage of laws continued and eventually culminated in 1971, when the first war on drugs was declared by President Nixon. As a result, stricter anti-drug laws were passed at the state and federal levels, and the Drug Enforcement Agency was created to enforce federal laws throughout the nation. Legislative reaction to illicit drug use primarily originated from concerns about marijuana, cocaine, and opiate use; however, the use of methamphetamine, club drugs (e.g., Ecstasy, LSD), and the illegal use of prescription drugs has garnered substantial attention from policymakers and law enforcement over the last 10 to 20 years.

Despite increased concerns over the use of these drugs, accurately documenting the extent of the drug problem was impossible until the 1970s because there were no standardized surveillance systems to measure the type or extent of drug use across the nation. To address this issue, the U.S. government began funding national data collection systems. Two primary data systems established during this time were the Drug Abuse Warning System (DAWN) and the Arrestee Drug Abuse Monitoring Program (ADAM, originally known as the Drug Use Forecasting Program).

Drug Abuse Warning Network

DAWN was established in 1972 as a national surveillance system to measure drug-related morbidity and mortality using data from hospital emergency departments (EDs) and medical examiners/coroners. Originally, DAWN was administered by the Drug Enforcement Agency and the National Institute on Drug Abuse, but federal law, Section 505 of the Public Health Service Act (42 U.S.C. 290aa-4), now requires the Substance Abuse and Mental Health Services Administration (SAMHSA), an agency of the U.S. Department of Health and Human Services, to collect nationwide data on morbidity and mortality. Currently, oversight of DAWN is managed by SAMHSA’s Office of Applied Studies.

For over a decade, DAWN data were collected from a convenience sample of hospital EDs, medical examiners, and coroners from across the United States. Sampling design limitations, however, rendered the data useless for producing national estimates of drug morbidity or mortality. To address this problem, DAWN data in 1988 were drawn from a representative sample of 24-hour hospital EDs in nonfederal short-stay general medical/surgical hospitals throughout the coterminous United States. Data collection from medical examiners/coroners, on the other hand, continued to be based on a nonprobability sample.

The primary purpose of DAWN was to be a first indicator of the serious consequences of drug use. In serving this purpose, it was anticipated that DAWN would help quantify the nation’s drug problem; monitor drug use trends over time; and guide resource allocation decisions at the local, state, and federal levels. Unfortunately, the realization of these goals fell short. In reality, DAWN was not able to produce results in a timely manner, its focus was too limited and did not consider broader issues of health, the quality of its data was questionable because of methodological limitations, and its methodology did not account for changes in the health care system. Consequently, the future of DAWN was questionable until a 1997 expert review panel convened by SAMHSA validated the importance of DAWN and recommended that it continue with substantial revision.

A 2-year evaluation of DAWN was commissioned in 1999 to identify how the program should be revised; specifically, the evaluation addressed whether (a) DAWN was collecting the right type of data to meet its goals, (b) the data DAWN collected were valid and efficiently collected, and (c) whether DAWN information could be delivered to policymakers and communities more effectively. The results of the evaluation (published in 2001) resulted in the development of New DAWN in 2003. In particular, New DAWN was born with a new sampling design, new case criteria, expanded data collection, and improved quality control.

A New Sampling Design

To collect data on drug-related ED visits, New DAWN uses a longitudinal, probability sample of nonfederal, shortstay, general surgical and medical hospitals with at least one 24-hour ED in all 50 states. More metropolitan sites were included in sample selection, and the boundaries for 13 of 21 original DAWN metropolitan areas were redrawn. Hospitals eligible to participate in DAWN were initially drawn from the 2001 American Hospital Association Annual Survey Database, which is a national registry of U.S. hospitals. The DAWN sampling frame (i.e., hospitals eligible for DAWN participation) is annually updated with newly established hospitals using this database.

New DAWN hospital samples were (and continuously are) drawn using a stratified simple random sampling approach. Stratification was based on (a) geography (i.e., Metropolitan Statistical Areas and the remainder of the nation) and (b) hospital ownership and size. This process was used to identify 54 geographic units. Fifty-three of these units represent metropolitan areas (e.g., 48 metropolitan areas, 2 subdivisions each for 3 of these metropolitan areas, and 3 subdivisions for 1 of these metropolitan areas). The 54th unit of the New DAWN sample is also known as the supplemental sample, which is intended to augment the metropolitan samples in order to yield a more complete national sample.

Sampling changes were less dramatic for medical examiners/coroners. As indicated earlier, a nonprobability sample of participating medical examiners/coroners was used for DAWN data collection even after probability sampling was implemented in 1988 for hospital EDs. A critical problem for improving sampling methods for drug-related mortality was the variability in criteria for death investigations across jurisdictions, which prevented the use of a probability sampling design. However, because DAWN is currently the only large-scale surveillance system that collects data directly from medical examiners and coroners, it was retained in New DAWN with sampling improvements rather than a redesign. To improve the sample of medical examiners/coroners, all medical examiners/coroners were recruited from the ED target areas and six statewide systems (Maine, Maryland, New Hampshire, New Mexico, Utah, and Vermont) were added to the sample. Because probability sampling was not possible for drug-related mortality, it is important to remember that national estimates for drug mortality are not possible.

New Case Criteria

Under Old DAWN, inclusion of ED cases required designated DAWN recorders at each site to infer the patients’ intent of use from medical charts; specifically, they were required to identify cases that were “induced by or related to” drug abuse; consequently, a range of visits related to drug use was excluded from data collection. New DAWN criteria no longer limit inclusion of ED cases to drug abuse. The program now requires a systematic review of all ED charts to identify any type of drugrelated visit. Medical examiner/coroner data now also include any type of drug-related death. Thus, New DAWN data comprise episodes and deaths that result from drug abuse and misuse, suicide attempts/completed suicides, overmedication, adverse reactions, accidental ingestions, malicious poisoning/homicide by drugs, underage drinking, patients seeking detoxification or drug abuse treatment (ED only), and other deaths related to drugs (medical examiner/coroner only).

Expanded Data Collection

In revising DAWN, data items without purpose were eliminated, and new data items were added to improve the description of cases and help distinguish different types of drug-related visits. Currently, DAWN data for ED visits capture patient demographics, the type of drug involved, the route of administration, the type and disposition of the case, presenting complaint, diagnosis, and case narrative. DAWN data for deaths include the demographics of the deceased, the type of drug involved, the route of administration, the involvement of the drug in the death, the manner of death, and the cause of death. The following types of drugs are included in data collection:

  • Illegal drugs, such as heroin, cocaine, marijuana, and Ecstasy
  • Prescription drugs, such as Prozac, Vicodin, OxyContin, alprazolam, and methylphenidate
  • Over-the-counter medications, such as aspirin, acetaminophen, ibuprofen, and multi-ingredient cough and cold remedies
  • Dietary supplements, such as vitamins, herbal remedies, and nutritional products
  • Psychoactive, nonpharmaceutical inhalants
  • Alcohol in combination with other drugs
  • Alcohol alone, in patients aged less than 21 years

Ultimately, the data changes expanded New DAWN’s coverage to include any ED visit or death “at any age, for any drug, for any reason.” This substantially increased the utility of the data and expanded New DAWN’s focus from documenting drug abuse to documenting drug use, misuse, and abuse and relating these dimensions of use to health consequences.

Improved Quality Control

One of the most debilitating criticisms of Old DAWN was that of poor data quality. As mentioned, identification of eligible cases was previously subject to reporter assessments about the drug-related nature of the visit or death, and training for reporters did not assure consistency across sites. Missing data was often a problem, reducing the usefulness of the data. Data reporting was done using outdated methods, substantially delaying the production and distribution of results.

To address these issues, New DAWN eliminated reporter assessments of intent and instituted rigorous training for reporters. In addition, strict quality assurance mechanisms were instituted, including data validation at data entry, automatic prompts for reporters, the use of a drug look-up database, and performance feedback for individual sites. Data are now submitted electronically, and deidentified data related to ED visits are made available to participating sites in real time via the Internet.

These changes substantially change the nature and strength of New DAWN data. They address the criticisms lodged at Old DAWN and help the program accomplish its original goals. These changes, however, make comparisons between Old DAWN data and New DAWN data impossible. Scholars must remember the incompatibility of the two iterations when reviewing, assessing, and reporting DAWN results across different time periods.

Selected Results for Emergency Department Visits

In 2005, hospitals in the United States delivered a total of 108 million ED visits, of which an estimated 1.4 million were related to the misuse or abuse of drugs:

  • 31% involved illicit drugs only
  • 27% involved pharmaceuticals only
  • 7% involved alcohol only in patients under the age of 21
  • 14% involved illicit drugs with alcohol
  • 10% involved alcohol with pharmaceuticals
  • 8% involved illicit drugs with pharmaceuticals
  • 4% involved illicit drugs with pharmaceuticals and alcohol

As just shown, about one third of visits involved one type of drug (e.g., illicit drugs only), whereas the remaining two thirds of visits involved some combination of drug types. Although the majority of drug misuse/abuse visits were associated with a single drug type, the typical drugrelated visit included multiple drugs within that type. In other words, an illicit drug only visit (i.e., not combined with alcohol or pharmaceuticals) often involved the use of multiple illicit drugs (e.g., cocaine and heroin).

Alcohol and Illicit Drug Use

Among all the drugs on which information is collected by DAWN, alcohol is unique. An ED visit related to alcohol use qualifies as a DAWN case under only two conditions: (1) The alcohol is found in combination with other drugs, regardless of patient age, or (2) the alcohol is found alone (i.e., not in combination with other drugs) in a patient under age 21. DAWN estimates that, for 2005, about one third (34%) of ED visits involved either alcohol in combination with another drug (all ages), or alcohol alone for patients under age 21. Of all these ED visits involving alcohol, about 7% involved patients under age 21 who used alcohol alone.

As indicated earlier, DAWN records ED patient demographics, such as gender, age, and race/ethnicity. In 2005, the male rate for visits involving alcohol in combination with other drugs was higher than the female rate. The rates between age groups varied little, but younger and older patients had lower rates. The data for race/ethnicity groups are more precarious than other demographic characteristics because of problematic levels of missing data (11%); however, on the basis of available data, 54% of alcohol-related visits involved white patients, 39% involved African American patients, and 12% involved Hispanic patients.

Over half of all drug-related ED visits were related to an illicit drug alone or in combination with another drug type. Of all visits involving illicit drugs, the drugs most often implicated were cocaine (55%), marijuana (30%), heroin (20%), and stimulants (17%; 13% methamphetamine and 4% amphetamines). Other illicit drugs were involved in less than 5% of cases. Rates per 100,000 also indicate that cocaine was the most likely drug involved in ED visits, followed by marijuana, heroin, and stimulants.

According to 2005 DAWN data, male rates (361.2 per 100,000) for all illicit drug visits were almost twice as high as female rates (192.1 per 100,000). In particular, male rates were double those of females for cocaine, heroin, and marijuana; however, the difference was much smaller for stimulants (57.8 compared with 36.2). In fact, the female rate for visits related to stimulants was equal to the female rate for heroin visits (36.2 and 36.9, respectively). With regard to age, the rates for all illicit drug visits were highest among 21 to 24-year-olds (581.5), but similar rates were found for 18 to 20-year-olds (517.5), 25 to 29-yearolds (528.5), and 35 to 44-year-olds. Rates were slightly lower for 30 to 34-year-olds (490.9) and substantially lower for 12 to 17-year-olds (197.9). As expected, the rates for patients less than 12 were 3.0 or fewer per 100,000. In general, rates for marijuana were highest for younger patients, whereas rates for cocaine were highest for older patients. The point at which this change in rates occurs appeared to be in the 21-to-24 age group. In addition, stimulant rates were higher than heroin rates for patients between the ages of 12 and 29. After age 29, the rates for heroin exceeded those for stimulants.

Similar to trends related to alcohol, the race/ethnicity data for illicit drug use visits are somewhat uncertain because of missing data: 13% for all illicit drug-related visits. On the basis of available data, however, 46% of the visits related to any illicit drug use involved patients who were white, 27% involved patients who were African American, and 12% involved patients who were Hispanic.

Selected Results for Drug-Related Deaths

The most recent DAWN data published for drug-related deaths are from 2003. In 2003, 122 jurisdictions in 35 metropolitan and 6 states participated in DAWN, and the response rate for medical examiners/coroners in these areas ranged from 9% to 100%. The 2003 DAWN mortality report focused on deaths attributed to suicides, homicides by drug, overmedication, all other accidental causes, and undetermined causes.

Overall, drug-related deaths ranged from 56.1 to 205.6 (per 100,000 population). The rates were lowest for children and adolescents, higher for 21 to 34-year-olds, and, with few exceptions, highest for 35 to 54-year-olds. More than half of the deaths occurred at home. Almost half of drug-related deaths involved alcohol or the use of an illicit drug (e.g., cocaine, heroin, and/or stimulants), and these deaths were typically attributed to multiple use of drugs rather than the use of a single drug. Single-drug deaths, however, were most likely to involve opiates, cocaine, or stimulants.

Significance of DAWN

DAWN contributes significantly to our knowledge of drug misuse and abuse in the United States. Since the revision of DAWN in 2003, it is better suited to be a first indicator of the health consequences of drug misuse and abuse while simultaneously providing an improved baseline for monitoring trends related to alcohol, illicit drug use, and nonmedical use of pharmaceutical drugs. Importantly, New DAWN allows for more accurate cross-site comparisons of drug use, identification of emerging drug use patterns, and investigation of suspected misuse/abuse patterns in particular areas. DAWN data are available quickly through DAWN Live! to participating sites and eligible public health organizations, and SAMHSA regularly publishes special reports using DAWN data. In turn, the availability of DAWN data capitalizes on its potential to inform policy and program development, document problems to support local initiatives, identify the link between drug abuse and other public health problems (e.g., sexually transmitted diseases), evaluate local anti-drug efforts, and contribute to academic research on drug abuse.

Given the improved utility of DAWN data, its desirability to a variety of entities has grown. At the local level, consumers of DAWN information include community epidemiology work groups, treatment agencies, prevention coalitions, and member hospitals. At the state level, state health and human services and law enforcement agencies utilize DAWN data, and at the federal level, several agencies rely on DAWN data, including the Food and Drug Administration, the Centers for Disease Control, the White House Office of National Drug Control Policy, the Drug Enforcement Administration, and the National Institute on Drug Abuse. In addition, the pharmaceutical industry utilizes DAWN data to monitor the use (or misuse) of its drugs.

Arrestee Drug Abuse Monitoring

From 1987 to 2003, the National Institute of Justice (NIJ) sponsored two innovative and ambitious efforts to provide needed information to local communities grappling with the issue of drugs and crime: (1) the Arrestee Drug Abuse Monitoring (ADAM) program and (2) its predecessor, Drug Use Forecasting program (DUF). The primary goals of these programs were to uncover national trends in drug use, provide local communities with early evidence of drug epidemics, support local planning officials with data linking crime to drug use, and measure the impact of efforts to reduce drug use and crime.

DUF and ADAM represent unique drug use surveillance systems compared with other national drug use indicators for several reasons. DUF/ADAM were the only programs that used recently booked arrestees—a population thought to comprise mostly hardcore or heavy drug users—as the target population. In addition, they were the only programs to collect self-reported drug use behavior and then confirm those reported drug use behaviors with bioassay results conducted on urine samples provided by the same respondents. Also, although DUF and ADAM played an important role in assembling a national picture of drug use and abuse in the arrestee population, both programs were both designed primarily as local information tools that would provide local users (law enforcement, treatment, prevention, and public health policymakers) with reliable estimates of the prevalence of drug use and related problems in their communities. Thus, DUF/ADAM data did not provide, and were not intended to provide, national estimates of arrestee drug use; instead, their purpose was to capture trends related to arrestee drug use in various communities across the United States.

Overview and Background

Beginning in 1987, the DUF program began collecting information on drug use among urban arrestees in 10 geographically dispersed cities. Data were collected quarterly for a 1to 2-week period using a convenience sample drawn from newly booked arrestees in the largest booking facility (i.e., jail) at each site. The target sample size for each site was approximately 225 interviews and urine specimens from adult male arrestees and 100 interviews and specimens from adult female arrestees. To collect DUF data, interviews were conducted with arrestees who had been arrested no more than 48 hours prior to the time of data collection. In addition to the interview, participating offenders were asked to provide a urine specimen, which was later tested for the presence of amphetamines, barbiturates, benzodiazepines, cocaine, marijuana, methadone, methaqualone, opiates, PCP, and propoxyphene. In cases where a specimen screened positive for amphetamine, the specimen was subjected to confirmation testing to detect whether a specific form of amphetamine—methamphetamine—was used.

Data collection under the DUF program was voluntary and confidential. In most sites, more than 80% of the arrestees approached agreed to be interviewed, and more than 80% of participating arrestees agreed to provide a urine specimen. A study-generated identification number was assigned to each interview instrument and its corresponding urine specimen container so that the data from each could be linked at a later date. Collection of the urine specimens allowed for comparison of the self-reported indicators of drug use and indicators of drug use based on urinalyses.

DUF data collection was conducted quarterly, for several reasons. Quarterly data collection generated new information more frequently than other national data collection programs. Quarterly collection and the timely release of findings allowed policymakers and analysts to view trends as they developed, potentially permitting earlier intervention into problems. Finally, quarterly collection helped adjust data for seasonal changes in arrest and crime patterns that occurred in some sites.

The DUF questionnaire was designed to collect information on demographic information, criminal offenses, the types of drugs used by arrestees, and their perceived need for alcohol and drug treatment. For more than a decade, baseline statistics were collected that detected trends in drug use that were then used to guide public health and safety policies. As is often the case with large-scale programs, though, important changes were made to the program during this 10-year period. An additional 13 sites were added as the program evolved, and in 1991, the composition of the DUF sample expanded to include juveniles in selected sites. Juvenile data were collected in 13 sites until 2002, when this part of the program was indefinitely suspended because of legal obstacles related to collecting information from juveniles.

In addition to the expansion in the number of DUF data collection sites and the population targeted, the original DUF instrument underwent three revisions. These revisions included slight changes in wording of existing questions, the exclusion of some questions, and the inclusion of new ones. Although DUF remained a rich, consistent data source for examining trends in arrestee drug use, it continued to receive criticism for its sampling limitations. To address this issue, the NIJ embarked on an ambitious agenda to increase the generalizability of data collected in its DUF program in 1996. In 1997, this effort resulted in the redesign of the DUF program into the ADAM program. Major changes included the expansion of sites from 23 to 35, the implementation of a new sampling design, the development of a new survey instrument, new quality assurance protocols, and expanded use of addenda for research purposes.

New Sampling Design

Under the ADAM program, a standard catchment (i.e., geographic) area for site data collection was defined as the entire county and applied in all selected sites. For example, in Los Angeles, under DUF, data collection occurred at one facility. ADAM expanded that sample to six locations to include a representative sample of facilities from all adult detention facilities in Los Angeles County. This change made it possible for local officials to make assertions about the entire county’s arrestee population based on ADAM data.

Changes in the sampling of arrestees at each site were also made. Like DUF, ADAM protocol still required trained interviewers to approach recently booked arrestees and administering a short voluntary and confidential interview; however, the ADAM program implemented probabilitybased sampling plans for male arrestees. Unlike the DUF samples, which were based on convenience samples, ADAM arrestees were selected for an interview and tested for drugs using disproportionate, stratified sampling plans tailored to each facility that accounted for several characteristics related to arrests, including day of week of the arrest, time of day of the arrest, reason for the arrest, and where the arrestee was booked.

New Data Collection Protocol

To improve the quality of the data collected, a new survey instrument was developed, and crosswalks from ADAM to other major drug and crime indicators were included. The new questionnaire expanded ADAM’s focus from self-reported drug use to the need for drug treatment and information related to drug markets in the area. Specifically, the ADAM questionnaire (a) preserved the key drug use measures while placing greater focus on five primary drugs (i.e., cocaine, marijuana, methamphetamine, opiates, and PCP) and their patterns of use over the prior year; (b) incorporated a validated drug use dependency and abuse screener; (c) included self-report information related to offender participation in inpatient, outpatient, and psychiatric treatment over the prior year and prior arrest history; and (d) added a section on drug acquisition and recent use patterns that provided greater insight into the dynamics of drug markets, use, and sharing. Also, questions (e.g., crosswalks) were included to directly link ADAM data to data produced by other national indicators, such as the National Survey on Drug Use and Health (formally known as the National Household Survey on Drug Abuse), the Treatment Episode Data Set; the Uniform Facilities Data Set; the System to Retrieve Information from Drug Evidence; Uniform Crime Report and National Incident-Based Reporting System; and the State Needs Assessment Household Telephone Surveys. Other surveys used during the questionnaire redesign were those conducted by the Center for Substance Abuse Treatment and the U.S. Census Bureau.

Improved Quality Assurance

To ensure that the program performed according to its mission and within a cost-efficient manner, the NIJ funded and directed all ADAM program operations with the assistance of a national data collection contractor and a national laboratory contractor. This structure provided the program with a centralized system of oversight that included fiscal management, rigorously standardized data collection procedures, minimum requirements for staff, and an ongoing accountability (operational and fiscal) from all data collection sites.

Whereas the NIJ was responsible for developing, shaping, and overseeing the infrastructure and methodology of ADAM to ensure accuracy and consistency across the different localities, the national data collection contractor maintained and implemented the technical and operational infrastructure for the program, including project management; communicating, defining, and enforcing the methodology and procedures; ADAM data entry, verification, and dissemination; and overall system support (technical support for the Web site, software, and standard programming systems). A Department of Health and Human Services certified laboratory contractor conducted analyses on all ADAM urine specimens. An immunoassay system was used to screen for the presence of drugs in urine. All ADAM specimens were analyzed using the same procedures with corresponding cutoff levels and detection periods.

The national data collection contractor was ultimately responsible for receiving and processing all data from participating sites and the national laboratory, which were then subjected to the same data management procedures by means of automated editing and entry programs. These data were then analyzed and disseminated in quarterly and annual reports by the national data collection contractor. These controls made it possible to compare findings from site to site.

Each research site in the ADAM network entered into subcontracts with the national data collection contractor and was held to minimum performance standards that included staffing, data collection, and fiscal performance requirements. Data collection at each site was managed by a local research team that included a site director and site coordinator. Site management teams were responsible for key operational issues, including directing and supervising data collection, establishing and maintaining contact with booking facility representatives, hiring and maintaining professionally trained interviewers and other site staff, overseeing data collection performance and monitoring site adherence to national data collection standards (meeting established data collection targets and minimum interview error rates), coordinating communication with the NIJ and the national contractors, and overseeing proper invoicing and fiscal monitoring of the site budget. Each site was evaluated every quarter and informed of any programmatic and/or administrative problems in the form of a feedback report and compliance letter.

The NIJ’s national standards for site staff involved the completion of a formal training schedule that used standardized materials. These materials covered training on interviewing techniques and on administration of the ADAM interview instrument. All site staff were required to attend and successfully complete 20 hours of training before they were permitted to interview arrestees. Interviewers were then required to demonstrate minimum levels of ability through their initial formal training and startup training, which was considered a probationary period. Contingent upon satisfactory completion of a minimum period of formal and on-site training, interviewers were then accepted for work on ADAM.

By and large, ADAM training was conducted just before data collection occurred at a site so that interviewer skills would be immediately applied to field conditions and so interviewers could be regularly observed and evaluated by national trainers. However, ongoing monitoring of the quality of their work determined their continued acceptability as ADAM interviewers. In addition, all interviewers were required to participate in a minimum of 6 hours of in-service training every year to maintain interviewer skills.

Research site staff was also responsible for disseminating local ADAM findings to a Local Coordinating Council composed of local, state, and federal law enforcement; representatives from the courts and corrections; social services professionals; treatment providers; policymakers; community-based organizations; the local research community; and other stakeholders. In turn, the Local Coordinating Council was responsible for focusing its efforts on integrating the data into local planning processes, meeting unique local information needs, and creating a local research agenda.

Use of Research Addenda

In addition to strengthening the program’s foundation, a greater utilization and recognition of the program as a research platform was supported. By appending specialized questionnaires to the ADAM instrument, several communities began exploring a wide range of topics that local policymakers wanted to investigate or address. The first of these studies focused on methamphetamine in five sites. Another study, conducted in partnership with the Centers for Disease Control, collected information on HIV testing, risk behaviors, and health care access among the arrestee populations at three sites. Another group of sites began collecting information on interpersonal violence and service needs among female arrestees, and two other sites began exploring the nature, patterns, and consequences of pathological gambling among the arrestee population. Ultimately, these studies and many others (e.g., acquisition and use of firearms by arrestees) helped improve the nation’s knowledge base about the social and behavioral correlates and consequences of drugs and alcohol among an arrestee population.

In sum, all of the enhancements made to the ADAM program resulted in a more comprehensive, nationwide source of information on drug use. Using ADAM data, the NIJ as well as local communities were able to identify levels of drug use among arrestees, track changes in patterns of drug use, identify specific drugs that were used and abused in each jurisdiction, alert officials to trends in drug use and the availability of new drugs, provide data to help understand the drug-crime connection, help evaluate law enforcement and jail-based programs and their effects, and serve as a research platform for a wide variety of drugrelated initiatives. In short, ADAM provided reliable and valid information that helped develop evidence-based policies to assist both local and national policymakers.

Unfortunately, the ADAM program was discontinued in early 2004. This action was taken in response to the significant reduction in appropriations received by the NIJ in that fiscal year for social science research. However, the Office of National Drug Control Policy, Executive Office of the President, introduced a modified version of the ADAM program in 2007. This program is called ADAM II, which is appropriate given that it uses the same research protocols introduced by the ADAM program. The only changes in ADAM II are the number of sites funded to collect data; a sole focus on adult male arrestees; and the questionnaire used, which was modified slightly to address issues surrounding the rising use and abuse of methamphetamine. Results from the new program are still awaiting release.

Selected Findings from ADAM

Over the years, the ADAM program collected drug use data from more than 89,000 adult male arrestees and from more than 17,000 adult female arrestees in 42 fully or partially funded sites. The level of drug use among the respondents was substantial in all years and at all sites. Every site reported that a majority of its arrestees tested positive for at least one drug. In more than half of the ADAM sites, almost 2 of every 3 arrestees tested positive for at least one of five drugs, including cocaine, marijuana, methamphetamine, opiates, or PCP. With only a couple of exceptions, marijuana percentage-positive rates, along with self-reported use, were higher among adult male arrestees, followed by cocaine, opiates, and methamphetamines. Female arrestees tested positive more often for cocaine, followed by marijuana, methamphetamine, and opiates.

In half of the sites, nearly 40% or more of all adult male arrestees reported that during the past 30 days, they had consumed drugs on 13 or more days. In 50% of the sites, almost 40% or more of adult male arrestees needed drug treatment, and nearly 30% needed alcohol treatment. Adult female arrestees also showed high rates of dependence, with 40% indicating a need of drug treatment and 23% in need of alcohol treatment. However, in no site did more than 20% of adult arrestees experience inpatient or outpatient drug or alcohol treatment. Potentially affecting this participation rate could be that nearly 60% of the adult male arrestees and 50% of the adult female arrestees reported having had no health insurance.

These consistently high percentages of overall drug use, however, masked the important differences in trends for specific drugs and specific segments of the arrestee population. In the late 1980s and early 1990s, the DUF data indicated serious problems with methamphetamine use in western sites. The problem of methamphetamine has shifted over time as evidenced by ADAM data, which showed that the southwestern and midwestern sites were reporting and testing positive for methamphetamine for the first time in the midto late 1990s, including areas considered more rural. In areas where urine screens had indicated virtually no use in the mid-1990s, methamphetamine rose dramatically and continued to grow into 2000 and beyond.

From 2000 to 2003, ADAM data showed that the percentages of adult male arrestees testing methamphetamine positive across sites increased and remained relatively high up until the program was decommissioned. Of the adult male arrestees who were tested by urinalysis for the presence of drugs, only 1.6% tested methamphetamine positive in 2000. The proportions increased to 2.6% in 2001 and to 5.3% in 2002, and in 2003, 4.7% tested methamphetamine positive. For the years 2000 through 2003, methamphetamine rates remained higher in western and most southwestern areas than in midwestern and eastern areas.

When asked by ADAM interviewers if they had acquired specific drugs in the past 30 days, large percentages of adult male arrestees in 2000 through 2003 admitted that they had acquired marijuana shortly before being arrested. Although much smaller proportions reported acquiring methamphetamine in the past 30 days, the percentage more than doubled from 2000 to 2002 (from 3.0% to 6.5%). In 2003, 4.9% of the adult male arrestees said they had acquired methamphetamine in the 30 days prior to arrest. In each of the 4 years (2000-2003), there were substantial increases in the percentages of adult male arrestees in states near the U.S.-Mexico border who reported having acquired methamphetamine in the 30 days prior to arrest. Once again, these differences suggest that appropriate policy responses to the drug problem vary widely from one community to the next and from one cohort to another, illustrating the need to tailor drug prevention, treatment, and enforcement strategies to local environments.

Significance of the Drug Use Forecasting Program and the Arrestee Drug Abuse Monitoring Program

Throughout the evolution from DUF to ADAM, a few characteristics distinguished these programs from other research programs. First, the DUF/ADAM programs focused on arrestees, a group that is more likely than other populations to be drug involved. Consequently, program data presented a different picture of drug use from that of studies that focused on household (National Survey on Drug Use and Health) or other populations (Monitoring the Future) and provided timely information to criminal justice policymakers and practitioners. They also provided a significant opportunity to conduct research on the association of drug use and criminal activity. Second, these programs were the only national research program studying drug use that used interviews and drug testing, providing a way to assess the validity of self-report data. By relying on the combination of self-report data and urinalyses of arrestees at the time of booking, these data were less susceptible to either exaggeration or denial of drug use than many other surveys. Finally, and perhaps most important, the DUF/ADAM programs were the only national drug use research programs built to specifically document and report the drug problem at the local level. Over the years, program data revealed that there was no single national drug problem; instead, there were different local drug problems that varied from city to city. This information was important, because communities often lack tools to monitor problems in a consistent, comprehensive manner and therefore have difficulty formulating appropriate policy responses.

Taken together, DUF/ADAM provided communities struggling with emerging and long-standing substance abuse problems with critical research tools to measure and understand the local drug problem and to evaluate programs and interventions that targeted the criminally active population. In addition, the DUF/ADAM programs provided a network of local drug use data that formed a foundation for understanding drug use across the country.


Attention to national drug use trends data has been important in the United States since the turn of the 20th century. In an age where communities face imminent, substantial budget cuts, understanding the drug problem in one’s community and sharing that information with policymakers and practitioners is essential, especially when an increase in service demand is coupled with a decreasing resource supply. Both the DAWN and DUF/ADAM programs have significantly contributed to the nation’s ability to document such information for over 20 years.

The history of the DAWN and DUF/ADAM programs underscores the difficulties in trying to measure drug use and its consequences across the United States. In particular, both programs received criticisms over their limitations and underwent significant revision in an effort to make the data they collect more reliable and valid. The data from these programs, in turn, improved in quality and usefulness to federal, state, and local policymakers and governments. As a result, the DAWN and DUF/ADAM data have been used in a variety of capacities to influence policymaking, program implementation, resource allocation, and an overall understanding of drug use and its consequences. The future of DAWN is promising, with substantial hope and resources devoted to the continued development of a strong national indicator of drug-related morbidity and mortality. The future of DUF/ADAM, on the other hand, is less certain. Without ADAM or something similar, national indicators of drug use will remain silent on the trends and patterns of drug use among criminal offenders.