of questions about the organization and financing of alcohol and drug treatment facilities, and the characteristics of the clients that these facilities serve. This paper relies on data from the 1997 Alcohol and Drug Services Study (ADSS), which includes as one of its components a nationally representative sample survey of all types of alcohol and drug abuse treatment facilities in the United States. The Substance Abuse and Mental Health Services Administration (SAMHSA) contracted with Brandeis University to direct and analyze ADSS. The subcontractor for conducting the field data collection was Westat, Inc.
ADSS encompasses a three-phase research design and is based on a complex national sample of alcohol and drug abuse treatment facilities in the United States. Phase I consisted of a mail questionnaire collected by telephone interview of a stratified random sample of 2,400 noncorrectional alcohol and drug treatment facilities. Phase II consists of two components: (1) an administrator interview which collects more detailed cost information and other facility level data from a subset of approximately 300 facilities, and (2) record abstraction of over 6,000 clients. Phase III consists of up to six in-person follow-up interviews with Phase II clients, accompanied by urine testing to be conducted at six-month intervals. Data are being collected on treatment history, characteristics at admission to the index treatment, characteristics at follow-up including alcohol and drug use, employment, mental and physical health status indicators, illegal activities, and readmission to treatment. Facility level data collected in Phases I and II, combined with client level data collected in Phases II and III will allow for cost-effectiveness analyses, as well as other measures of treatment outcome.
This paper relies entirely on preliminary data from Phase I. The sampling frame was SAMHSA's 1995 National Master Facility Inventory augmented to encompass the universe of substance abuse treatment facilities. Phase I was conducted during early 1997 with data collected for the point-prevalence date of October 1, 1996 and for the most recent twelve-month reporting period of the facility. Data were collected which described facility characteristics as well as aggregate information on clients in the sampled facilities. Facility directors or administrators completed the questionnaire. The Phase I response rate was 92 percent.
The Phase I sampling design incorporates a stratified random probability sample that allows for estimates of parameters at the national level. The strata were selected to reflect the different modalities of care within the substance abuse treatment system. Since ADSS is based on sample data, weights have been developed to produce national estimates of facilities and characteristics of clients in treatment. The sampling weights adjust for