We considered findings on generic screening in developing an alternative approach. We label this approach "specialized screening." There are three key elements of specialized screening. First, an interview format is used establish rapport with the recipient. Discrepancies and inconsistencies in response to interviewer questions are carefully probed. In addition, concerns about privacy and confidentiality are addressed. Second, specially trained staff, although not necessarily addiction professionals, conduct the interviews. Third, because of resource allocation and cost associated with conducting interviews, only high-risk groups, i.e., those recipients who are thought to have a substance abuse problem, are screened. Specialized screening is designed to augment, but not replace, generic screening. We implemented specialized screening programs with these features in two counties in NJ.
One specialized screening program was implemented in a county with a relatively small TANF caseload of 469 recipients. Welfare regulations in NJ allow caseworkers to mandate a substance abuse assessment for any recipient who has failed a work activity. The county welfare office decided that all sanctioned clients would be required to undergo a substance abuse assessment prior to lifting the sanction. Letters were sent to sanctioned clients indicating that the sanction process could be stopped and benefits restored if the recipient came in for an interview. The type of interview was not specified and the tone of the letter was inviting and not punitive.
Clients responding to the letter were interviewed by an addiction counselor who was co-located at the welfare office. As part of the interview, counselors administered several standardized measures, including one that assessed substance use disorder diagnoses. As part of the interview, counselors also assessed for barriers to employability related to medical, mental health, employment, family, or legal problems. Reports were prepared and provided to caseworkers based on these assessments. Overall, 352 letters were sent to sanctioned clients and 86 clients (24%) responded to the letter and were interviewed between February and October 2000. Of these 42 (49% of those interviewed) met criteria for a substance use disorder. In addition to the identification of substance abuse among sanctioned clients, the county also benefited from reports that detailed findings of other barriers to employability among sanctioned clients.
A second specialized screening program was implemented in a county with the largest caseload in NJ, 16,401 recipients. The large caseload made it impractical to conduct in-depth interviews on all sanctioned clients. Instead, two welfare caseworkers with an interest in helping substance abusers were identified. These caseworkers had extensive experience in implementing special programs in welfare settings, but had minimal training in substance abuse assessment or treatment techniques. These caseworkers interviewed high-risk clients. These groups included: clients reporting a lost electronic benefit card, those applying for emergency assistance, clients who another caseworker suspected of having a substance abuse problem, and clients who responded positively to one or more items on the CAGE-AID.
In order to evaluate the effectiveness of this program we examined two sets of data. Available data for the period of March through September 2000 indicated that 853 special screening interviews were conducted. Overall, 36.5% (n=312) resulted in a referral for further evaluation. Almost all those referred for an assessment are determined to have a substance abuse problem. Thus, specialized screening appeared to be an efficient method of identifying substance abuse problems. In addition, we compared the referral rates in the county where the special screening program was implemented to those of other NJ counties. Overall, the rate of referrals for substance abuse assessments in this county was 10.3% versus 4.4% for other NJ counties during the same 12-month period. Thus, specialized screening appeared to more than double the rate of referral for assessment. It should be noted that this method of evaluation has limitations since it is possible that other characteristics of the counties might explain differences in referral rates. Nevertheless, there are no obvious factors (e.g., differences in prevalence rate of substance use across counties) that could account for the size of the difference.