Appendix 1: Description of the Survey

I. Sample Universe

The States, SAMHSA, and other Federal agencies identify the NDATUS universe. Survey documentation broadly defines a specialty treatment provider as having:

• a formal structured arrangement for drug abuse and/or alcoholism treatment or recovery, using drug abuse and/or alcoholism-specified personnel;
• a designated portion of the facility (or resources) set aside for treatment services; and,
• an allocated budget for such treatment services. (2)

In practice, specialty service providers are identified mainly via administrative records related to public funding, licensing, or other regulatory functions performed by State and Federal agencies. Fifty-six States and other jurisdictions collaborate with SAMHSA in defining the universe and in conducting the survey. Federal agencies participate by providing lists of providers. These include the Bureau of Prisons (BOP), Department of Veterans Administration (VA), Department of Defense (DOD), the Indian Health Service (IHS), and the Food and Drug Administration (FDA). Some providers self-identify and report voluntarily, often to be included in the National Directory of Drug Abuse and Alcoholism Treatment and Prevention Programs [SAMHSA 1993b].

The complete list or universe of specialty treatment providers is called the National Facility Register (NFR). (3) It is maintained and continuously updated by SAMHSA as an automated data base. Prior to the 1992 survey, the NFR identified 12,331 specialty treatment providers. Non-treatment providers (i.e. providers who offer prevention and other treatment related services) are also included in the NFR universe and they also report to NDATUS. Prior to the 1992 survey, the NFR identified 2,412 non-treatment providers.

II. Data Collection and Processing Methodology

The NDATUS survey collects one-day census data on specialty provider and client characteristics. Since the 1980 survey, the reference date has been at the end of September, with the exception of 1987 when it was at the end of October. One-day census (or point prevalence) measures of services utilization are used, instead of annual measures, in order to minimize the burden of accounting and reporting client data.

Prior to the survey, the NFR is updated by sending lists of providers (treatment and non-treatment) to the States, to other reporting jurisdictions, and to Federal agencies for confirmation, correction, deletions, and additions. These updated NFR records are updated a second time, by providers themselves, when they are pre-printed on page one of each form that is mailed to providers. The 1992 NDATUS form was a 5-page paper form. Non-treatment providers completed (or verified) only the first page.

Survey forms were mailed to the States on September 15, for subsequent distribution to providers. Three out of four States made all contacts with providers in order to encourage completion of forms. The SAMHSA contractor sent postcards to providers in several of the remaining States that had low response rates.

Completed forms were returned to State agencies for initial editing before forwarding to the SAMHSA contractor. The latter edited each form, keyed the data into an automated record, and then conducted a final automated edit. A limited number of follow-up telephone contacts were made to clarify ambiguities and inconsistencies or to complete missing items. Missing or unknown client level data on each record were replaced with valid responses using ratio imputation procedures, based upon data reported by the same provider or similar providers [see Appendix 3, Section III].

A label file, with identification data for treatment and non-treatment providers plus main program characteristics, was used to publish the annual National Directory [SAMHSA 1993b] and then made available to the public. In addition to general use as a referral guide, this directory is used by the Center for Substance Abuse Treatment's National Hotline.

To make unbiased national estimates of provider and client level data, key data were estimated from a representative sample of NDATUS non-responders. Each provider in the sample was mailed another copy of the NDATUS survey and data were collected via follow-up telephone calls.

Return to default