GAIN Short Screener IDs Needs of Adolescents in Washington, No Matter Which Door They Come In

In January 2007, the state of Washington mandated use of a 15-item, past-year version of the GAIN Short Screener (GAIN-SS) for all adolescents and adults seen by the Department of Social and Health Services in the state’s chemical dependency, mental health, child-welfare and justice programs.
 
Lucenko and colleagues1 recently used the data gathered through these sources to compile a report to the Washington legislature on the prevalence of co-occurring disorders in people presenting to these services. 
 
Figure 1 summarizes her results for adolescents. Consistent with prior research, the report shows that implementation of a simple screener identified youth with mental health, substance use or both in each setting.

 
To validate the measure, Lucenko and colleagues also examined all available administrative data (such as medical, treatment, urine results and arrest records) to identify people with any evidence of having co-occurring mental and substance disorders. They then compared the rate of co-occurring problems identified by the GAIN-SS with the results from all available clinical indicators. 
 
As shown in Figure 2, the rates are very similar. This suggests that the two-page GAIN-SS is a very efficient way to identify rates of co-occurring problems and its results are consistent with longer-term findings.
 
Two points to note here:
 
(1) The GAIN-SS does well at identifying the existence of a problem in a broad domain, but to better understand and treat it still requires more information and clinical input. The point here is simply that it is an efficient tool for initial identification of problems. 
 
(2)  It is important to note that in preparing the report, the researchers varied from the usual use of the GAIN-SS in several ways, including:
-- using 15 instead of 20 items (dropping crime/violence);
-- collapsing the internalizing and externalizing disorder screeners;
-- using a definition of 2+ symptoms on each subscale to indicate the presence of problems (versus the usual scoring of zero for low, 1-2 for moderate, and 3+ for high);
-- focusing only on the subscales (without looking at the total screener scores);
-- comparing the past-year symptom counts on the GAIN-SS with up to 2 years of clinical indicators (vs. only one year of clinical indicators); and
-- using the GAIN-SS's cutoff points to determine which youth had abuse/dependence (or were using at least weekly), but allowing "any use" to qualify -- e.g., including people with less than weekly use and no abuse/dependence -- when pulling data from other records for the comparison.
Each of these factors should have worked against the comparison. Yet the fact that the results of the GAIN-SS are so congruent with the data culled from all other available administrative sources for clinical indicators on the same youth makes the results even more striking.
 
The figures and text were created by the author with help from Melissa Ives and Tim Feeney.

1Lucenko et al (2009). Report to the Legislature: Co-Occurring Disorders Among DSHS Clients. Olympia, WA: Department of Social and Health Services. Retrieved from http://publications.rda.dshs.wa.gov/1392/