Performance Improvement 1999. Services Research Outcomes Study (SROS)


The Services Research Outcomes Study (SROS) was initiated at the request of the Office of National Drug Control Policy (ONDCP) in 1990 and was to be the first national drug abuse treatment outcome study based on a nationally representative sample of treatment providers. Earlier outcome studies - all showing positive outcomes from treatment - had used samples based largely on urban treatment providers. It was thought this sample bias would exaggerate positive treatment outcomes. The SROS confirmed these earlier findings of positive treatment outcomes and dispelled the notion that sample bias accounted for positive results. The SROS principal finding was as follows:

Graph: Treatment Reduces Drug UseA nationally representative survey of 1,799 persons confirms that both drug use and criminal behavior are reduced following inpatient, outpatient and residential treatment for drug abuse.

Because use of every substance decreased nationally in the five years after treatment, switching from one substance to another after treatment cannot account for the positive results.

The one exception to the overall decrease in drug use following treatment was for those who were adolescent when discharged from treatment. Following treatment, adolescents increased their use of alcohol by 13 percent and the use of crack by 202 percent, albeit from a low pre-treatment base of five percent. Specific reasons for the difficulties in treating adolescents remain unclear. Consequently, this treatment subgroup poses an ongoing challenge for addiction researchers and those who work with adolescent substance abusers.

To conduct a drug abuse treatment outcome study using a nationally representative sample, and to update earlier findings from ten and twenty years ago.

The Services Research Outcomes Study (SROS) was initiated in the National Institute on Drug Abuse at the request of the Office of National Drug Control Policy (ONDCP) in 1990 and moved to the Office of Applied Studies with the creation of SAMHSA in 1992. SROS was to be the first national drug abuse treatment outcome study based on a nationally representative sample of treatment providers. Earlier outcome studies such as DARP, TOPS, and the more recent DATOS are based on samples in which urban treatment providers predominate. It was thought that this urban sample could exaggerate positive treatment outcomes by overemphasizing the older and better-funded programs. SROS was designed to test this assumption by using a random national sample of treatment programs. SROS was also intended as an update of drug abuse treatment outcome findings to provide a periodic check on national treatment effectiveness.

These positive findings were determined through a survey of 1,799 (71.4 percent male and 28.6 percent female) persons discharged from treatment in 1989-1990. Lost through death were 277 potential respondents. Out of an eligible 2,756 respondents, 1,799 were interviewed for a survey capture rate of 65 percent, which is the expected rate for drug abuse surveys. Substance abuse records for each of the clients were provided by a nationwide sample of 99 drug treatment facilities. All 1,799 clients were fully informed of the purpose of the survey and gave their informed consent to be interviewed five years following discharge from drug abuse treatment. By weighting up to national levels, the respondents translate to 976,012 individuals discharged from treatment in 1990. SROS asked respondents to participate in a urine test of recent drug use and 76 percent agreed. The urine test results agreed with self-reported 'drug use in the last week' from 98-90 percent, depending on the drug. Urine tests agreed with self-reported alcohol use by 64 percent. The urine test is not the best measure of alcohol use, however, making it difficult to interpret these results for alcohol.

The SROS clearly demonstrates an overall national drop in the use of illicit substances five years after treatment episode. Reported use of any illicit substance dropped by 21 percent five years after treatment, which is an increase in abstinence from all illicit drugs covered by the survey.

Besides the overall increase in abstinence, each substance also showed increased abstinence, meaning that the surveyed sample was not switching from one substance to another but decreasing overall drug use. For instance, any use of alcohol decreased by 14 percent, any marijuana by 28 percent, any cocaine by 45 percent, any crack by 17 percent, and any heroin by 14 percent.

Treatment had positive outcomes for women, men, and all age groups except adolescents. Adolescents treated in 1989-90 showed a 13 percent increase in alcohol abuse and a 202 percent increase in crack use five years following treatment. Adolescent employment, however, increased dramatically. These findings describe the known tendency for adolescents to experiment unwisely, but treatment for the group must be improved to stop substance use quickly before it destroys a young life. Programs of improvement for adolescent treatment are underway at SAMHSA.

Survey results confirm those of previous studies showing that treatment for substance abuse can significantly reduce crime. Most criminal activity, including breaking and entering, drug sales, prostitution, driving under the influence and weapons use declined by between 23 and 38 percent after drug treatment. As one would expect, older age groups were more likely to reduce their post-treatment criminal activities than were younger groups.

Not only crime, but violence against oneself and other declined, specifically involvement in physical abuse and suicide attempts declined in the five years following treatment.

The positive counterpart of reductions in crime and violence was a noticeable shift toward regaining and retaining child custody after drug abuse treatment. Family stability was further enhanced by reports of more reliable housing obtained following treatment.

In locating respondents for the survey, a higher than average death rate was observed in this group, compared to the U.S. population. Due to the small number of observations, reliable estimates are not possible, but it was striking to note that white males discharged from treatment had roughly eight times as many deaths as expected (108 deaths, rather than the expected 13). White females discharged from treatment had nearly 18 times as many deaths as expected (35 deaths, rather than the expected two). Black males discharged from treatment had about five times as many deaths as expected (67 deaths, rather than the expected 13). Black females discharged from treatment had seven times as many deaths as expected (14 deaths, rather than the expected two).

Use of Results
Because SROS is based on a nationally representative sample of treatment providers, the findings describe drug abuse treatment in the U.S. as a whole This makes the study suitable to support national drug control and treatment policy. For example, knowing that treatment reduces the number of drug users nationally provides a rationale for expanding and improving treatment to reduce demand for illicit drugs and thus to shrink the illegal markets, at least in the U.S. Every drug treatment outcome study (DARP, TOPS, DATOS) has shown treatment reduces drug use, but until these findings were confirmed in a nationally representative sample, the findings were criticized as biased toward the larger urban and perhaps most effective treatment programs. SROS confirms that the consistently positive finding that drug abuse treatment reduces national drug use is not the result of sample bias. Further support for policy decision-makers will be provided by a replication of the SROS, the Alcohol and Drug Services Study (ADSS), now underway in the Office of Applied Studies, SAMHSA. The methods are those used for SROS with some additional refinements and improvements.

Just as important to policy as the overall positive results shown in SROS are the weaknesses found in the national drug treatment system. In 1989-1990, treatment was not effective for the majority of participating adolescents. While this may be understandable given the inexperience and experimentation known to characterize adolescents, treatment improvement for this group is clearly indicated. Efforts are now underway at SAMHSA to improve treatment for adolescents.

Finally, substance abuse is known as a relapsing disorder, and new studies that look at repeated treatments over time (longitudinal studies) may begin to tell us how many treatment episodes by which groups should be expected on the road to full recovery.

AGENCY SPONSOR: Substance Abuse and Mental Health Services Administration

FEDERAL CONTACT: Barbara A. Ray, Ph.D.

PHONE NUMBER: 301-443-0747

PERFORMER ORGANIZATION: National Opinion Research Center (NORC), Chicago, IL