Name of Project: A Method for Integrating Estimates
of Need for Treatment
Principal Investigator: Mary-Lynn Brecht, Ph.D
Funding Agency: Caliber Associates
Funding Period: November2002-November 2003
Scope of Project: This study, as part of the
National Evaluation Data Services (NEDS II) project, examines
the feasibility of using a hierarchical model approach and generalized
estimating equations (as might be used for the meta-analytic synthesis
of effects of a set of research studies) for synthesizing estimates
of treatment need from different subpopulations and disparate
data sources. The study will use estimates of treatment need from
the California State Treatment Needs Assessment Program, which
included data from household, arrestee, and youth populations.
Model assessment will include examination of the sensitivity of
results to specificity of sub-population estimates, sample size,
and violation of model assumptions. Assessing the critical treatment
gaps documented for substance abuse treatment requires reasonable
estimation of the need for substance abuse treatment and often
relies on separate estimates from specific subpopulations and
disparate data sources, since complete population enumeration
is rarely feasible. Combining such estimates into a single comprehensive
picture is a current challenge, usually approached conceptually
by explaining subpopulation coverage, overlap, and omissions,
without a statistical basis for integration. Models developed
in the current study may provide an empirical and analytic basis
to improve accuracy and credibility of estimates of need for treatment
For more information, contact Mary-Lynn
Brecht.
Last updated - 09/29/2003
Last reviewed - 08/4/03