appendix

METHODOLOGY

NTIES sought to answer a number of questions fundamental to an understanding of treatment effectiveness and the quality of Federally-funded treatment programs. Some of the questions it sought to answer were:

  • How were grant funds used? What improvements were made as a result of these grants? How many and what types of clients were affected by the awards?
  • Was the improvement approach validated? To what extent? What is the evidence that improved treatment services yielded effective results?

    • NTIES used a two-level study design. The first, or administrative level, covered the treatment orientation, size, budget, staffing distribution, and specific use of CSAT funds in all service delivery units eligible to receive support through the 157 grants, as reported by clinic directors. A second, or clinical-outcomes level, collected data when treatment began, ended, and approximately one year later from clients enrolled in eligible units.

    In all, 6,593 clients at 78 selected service delivery units were successfully enrolled in the clinical outcomes study. Clients were interviewed at admission to treatment, when they left treatment, and then at follow-up 12 months after the end of treatment. More than 82 percent of the recruited sample completed the follow-up interview, and 4,411 of those are included in the outcome analysis.

    The NTIES study, like many other major research surveys, relies primarily on self-reported data. Earlier methodological studies have shown this method of data collection to be generally valid and reliable. However, the NTIES study went a step further and collected additional data from outside sources to validate the self-reports in two key areas: illicit drug use and arrests. They collected a sample of drug testing data and a sample of arrest records that supported the NTIES findings.

    STUDY DESIGN

    NTIES measured the outcomes of treatment primarily through a method known as a "before/after" or "pre/post" panel design. This method compares behaviors or other individual characteristics in the same research subjects, measured in similar ways before and after an intervention. The results are expressed as a percentage of the occurrence of a behavior or circumstance in the NTIES population during clearly defined equivalent periods of time before treatment and after treatment.

    NTIES Table of Contents | Introduction | Conclusions | Appendix | NTIES Findings on: Drug and Alcohol Use | Changes in Criminal Behavior | Employment, Income, and Homelessness | Changes in Physical and Mental Health | Changes in Sexual Behavior | Variation in Treatment Outcomes | Costs of Treatment