teria for a psychiatric disorder are evaluated by a clinician, who determines whether the student needs to be referred for treatment or further evaluation. Ideally, the program also includes a case manager who contacts the parents of students who are referred and establishes links with a clinic to facilitate treatment adherence.

Evaluation results indicate that most of the adolescents identified as being at high risk for suicide through the program were not previously recognized as such, and very few had received prior treatment. About half of the students referred for treatment attended at least one treatment visit, however. In addition, the program's requirements of a clinician and a case manager may be a resource burden for many schools.

The screening strategy developed by Thompson and Eggert (1999) as part of their comprehensive Reconnecting Youth program (discussed in detail in Chapter 24) is based on a public health prevention model that emphasizes the identification of at-risk students on the basis of observable behaviors. The first level of screening involves a review of high school attendance registers to identify students having high absenteeism. Teachers and guidance counselors are asked to recommend students they deem to be at risk. Identified youth are then assessed by means of the Suicide Risk Screen (SRS). Those with elevated risk for suicidal behaviors are given an appropriate intervention within the school setting or are referred for further evaluation and treatment (Thompson & Eggert, 1999).

Recent screening initiatives for college students include the Comprehensive College Initiative (CCI), developed by Jacobs (2003) to identify students at risk for depression and facilitate them to get treatment. The program has been offered at a large number of colleges in conjunction with National Depression Screening Day. In addition to the in-person screenings offered at this annual event, the program includes an on-line year-round screening component.

In campuses where it has been implemented, the CCI has been described by its developers as effective in identifying at-risk students and motivating them to seek treatment (Jacobs, 2003). Almost 20% of students taking the screening measure scored “very likely” to be suffering from depression and 5% reported persistent suicidal ideation. Both student participants and college officials were reported to have positive reactions to the in-person and on-line program components. No data have yet been reported, however, on treatment follow-up or outcomes, or on changes in suicidal behavior on the participating campuses.

Another recent program is the College Screening Project developed by the American Foundation for Suicide Prevention (Haas et al., 2003). This project, which is currently being pilot tested at selected colleges, uses the campus e-mail network to target students and encourage them to complete a Depression Screening Questionnaire, which is found on a project-developed Web site. This instrument is an adaptation of the Patient Health Questionnaire, which has been established to be an effective tool for identifying depression among community samples (Spitzer, Kroenke, Williams, & The Patient Health Questionnaire Study Group, 1999; Spitzer et al., 2000). In addition to depression, the screening questionnaire includes items dealing with current suicidal ideation, past suicide attempts, anxiety and other affects, drugs, alcohol, and eating disorders. Students use a self-assigned user name and password to log into the Web site; the user name is the sole identification on the submitted questionnaire.

Assisted by a computer program, a clinically trained counselor evaluates the responses and assigns the student into one of three tiers on the basis of their suicide risk. The counselor then writes a personalized reply that the student accesses on the Web site, using their user name and password. Students with significant problems as determined by a well-defined set of criteria are urged to come in for a face-to-face evaluation. The Web site also contains a “Dialogue” feature that allows students to communicate with the counselor on-line to discuss concerns they may have prior to an evaluation.

During the face-to-face meeting, treatment options, including medication and psychotherapy, are discussed and referrals are made to appropriate services on and off campus. In an effort to evaluate treatment effects, the project collects data on an ongoing basis from treatment provid

ers on student adherence, treatment progress, and disposition.

Initial reports indicate that about 80% of the students who respond to the screening questionnaire indicate some mental health problems, with almost half of all respondents falling into the highest-risk tier. Fewer than 15% of identified students, however, comply with recommendations for evaluation, which suggests that recommendations need to be refined to make them more acceptable, or that innovative strategies need to be developed to encourage greater numbers of at-risk students to seek help. Almost all students who receive a clinical evaluation through the College Screening Project are referred for treatment. Over 90% of students coming for evaluation have reported that the screening questionnaire and the counselor's responses were critical factors in their decision to seek help (Haas, 2003).

One other Web-based screening program for college students, the ULifeLine program, has recently been developed by the Jed Foundation (2003). This program provides computer-generated results to students who complete the screening instrument. Although identified students are provided with recommendations regarding treatment possibilities, no follow-up is offered. It is not clear whether without a personal connection, such Web-based screenings will succeed in motivating students in need to seek treatment.

In their basic assumptions, screening programs as implemented within both high school and college settings closely conform to scientifically validated premises regarding the causes of suicide—i.e., that suicide risk is not randomly distributed, but rather is conferred by certain factors that are both identifiable and, to a considerable extent, alterable. At the same time, such programs face a number of challenges.

Screening measures with acceptable test characteristics (e.g., a sensitivity of 80% and a specificity of 70%, figures similar to screens for depression) will necessarily miss some in the population who will go on to make suicide attempts, while identifying many more as at risk when they are not. The often transient or episodic nature of suicidality among young people makes screening this population even more difficult. Given that costs are involved each time a segment of the target group is screened, most school-based screening programs assess students only once a year, and in some cases, only once during a several-year period. The timing of the screening may increase the likelihood of identifying students in need of referral (e.g., close to exams, at the beginning of high school or college, or during the senior year) or at other times may reduce this likelihood.

Both high school-and college-based screening programs report relatively low adherence with treatment recommendations among those identified through the screening instrument to be at risk. Although this is likely due to a range of problems that are beyond the scope of the screening effort (e.g., lack of parental support, perceived quality of available treatment, and attitudes of treatment providers), additional strategies appear to be needed to encourage students at risk to access and make effective use of needed treatment services. In this regard, better integration of suicide education, gatekeeper training programs, and screening programs may be helpful.

All school-based suicide screening programs need to be mindful of the availability and quality of mental health services for students who are identified as at risk. On college campuses, this is sometimes a formidable problem. It is estimated that only 38% of colleges provide mental health services (Gallagher, 2001), and most of those that do limit the number of sessions or offer only group therapy that may not be appropriate for students at risk for suicide. Although many colleges require students to have health insurance, most students (as well as most people in the general population) are not adequately covered for acute or long-term mental health services.

Even when implemented under ideal conditions, there is no clear evidence that screening for suicide in general populations improves rate reduction outcomes. In addition, as yet, no data have been reported on the effectiveness of high

school-or college-based screening programs in reducing suicide risk factors, including depression and suicidal ideation, or suicidal behavior at the schools where screening programs are being implemented.

Within high schools, there is evidence that administrators prefer suicide education and awareness programs over screening programs (Miller, Eckert, DuPaul, & White, 1999). Many colleges and universities have also expressed reluctance about implementing depression and suicide screening programs. This appears to reflect, in part, concerns about the liability schools may assume in the event that students identified as at risk for suicide do not follow through with treatment recommendations and actually engage in suicidal behavior. Identification of at-risk students may also put universities into a difficult legal and ethical position with respect to parents. Because students over the age of 18 are considered adults, parents of students cannot typically be contacted without written permission from the student. Although confidentiality can be waived in situations in which threat to life is concerned, universities are reluctant to become embroiled in such matters. Further, monitoring students identified as in need of mental health services is difficult because of their diverse living arrangements and lack of supervision by other adults.

Although Web-based programs show promise as a tool for suicide screening with youth, one complication is the recent Health Insurance Portability and Accountability Act (HIPAA), which limits the use of electronic technology to transmit identifiable health information, because of the potential threats to patient confidentiality. This has been interpreted as requiring that a student's actual identity not be revealed on-line, making it impossible for the counselor to intervene to help a student believed to be suicidal unless he or she presents in person for evaluation.

Finally, as was earlier noted in discussing suicide prevention education, most screening programs directed at young adults are designed specifically for college students. Although screening programs are expensive to administer and monitor, creative strategies are needed for integrating and supporting screening into existing health-care settings that reach all youth.

Based on research indicating markedly different rates of suicide among different Native American tribes, May and Van Winkle (1994) suggested that high suicide rates among certain tribes were linked to a loosening of social integration within the tribe as members become increasingly acculturated into the broader society. The underlying assumption for a small number of programs is that instilling certain personal traits and social skills in Native American youth will counter the negative effects of the acculturation process and protect these youth against suicidality.

The Zuni Life Skills Curriculum for preventing suicidal behavior (LaFramboise & Howard-Pitney, 1995) is illustrative of programmatic efforts in this category. This program, developed specifically for Zuni youth, featured a 30-week, three-times-a-week, course focused on building self-esteem, helping youth identify feelings and stresses, improving communication and problem-solving skills, decreasing self-destructive behavior, and setting goals. The curriculum also provided information about suicide and training for intervening with suicidal peers. Results of the program were mixed, with students showing a decrease in hopelessness but not depression after the intervention. Although the program was not specifically addressed to suicidal youth, some of those who participated reported decreased suicidal behaviors. Adult judges rated the impact of the skills training program as positive, but youth overall reported few effects on social functioning.

Although the Zuni curriculum demonstrated some success, more specific evaluation of program efficacy is needed that incorporates a control-group design and links outcomes to specific program components. In particular, studies that suggest differential acculturation to be pivotal in explaining suicide rates among Native American youth have not controlled for other variables such as psychopathology or family influences. It should also be emphasized that no empirical evidence has been put forth that supports a link between high suicide rates among Native American youth and deficits in personal or social skills.

The program has not been replicated in other at-risk tribes. Resources available for the development, implementation, and evaluation of suicide prevention programs for Native American youth appear to be limited (Middlebrook, LeMaster, Beals, Novins, & Manson, 2001). An additional observation is that although programs targeting Native American youth are based in part on the premise that external forces in the social and cultural environment contribute to the difficulties these young people face, the strategies focus on changing individuals rather than the external influences themselves.