As part of its welfare reform efforts, the state of NJ implemented an innovative program to address substance abuse among welfare recipients. One important element of this program was a new strategy to screen and identify substance abuse problems in welfare settings. Welfare caseworkers were required to administer a brief screening measure, the CAGE-AID (Brown, 1992), to all individuals applying for or seeking redetermination of TANF benefits. The CAGE-AID is a nine item measure designed to screen for alcohol and other drug use problems. Individuals responding positively to two or more questions were referred for further evaluation to a trained addictions counselor who was co-located at the local welfare office. Welfare caseworkers received training on measure administration and referral. An earlier well designed study of AFDC recipients in NJ estimated the prevalence rate of substance abuse problems at 20-30%. Thus, state planners anticipated that the screening program would lead to a high rate of identification and treatment referrals. We label the approach used statewide in NJ as "generic screening". Generic screening has three key features: front-line caseworkers with only minimal training conduct the screening; screening occurs for all recipients at the point of benefit determination; and the approach relies on paper and pencil measures, rather than interviews.
An examination of state records indicates that the generic approach did not yield the expected rates of identification and referrals. During the first six months of the program about 1% of TANF recipients received a referral for further evaluation. Figures for the last year (6/99-6/00) indicate that 4.4% of TANF recipients were identified and received a referral for further substance abuse evaluation. Our group conducted a brief evaluation to determine factors that might explain the low referral rates (Morgenstern, 1999). We found substantial differences between the structure of screening in medical settings the setting for which the screening tool was designed and that occurring in welfare settings. We found that using front-line caseworkers to screen might limit effectiveness because workers were not trained and saw the task as complex and demanding. In addition, we found a high level of reluctance on the part of welfare recipients to self-disclose a substance use problem indicating that reliance on paper and pencil measures was inadequate.