A: In universal prevention, the program is implemented to everyone in a given population. But in selective intervention or selective prevention studies, we use population-based epidemiological evidence to identify a group at risk. For example, we know that for children, exposure to violence or exposure to poverty increases the risk for mental health conditions; therefore, using this epidemiological knowledge, we can select a group.

We might choose schools that are located in neighborhoods where there's a high crime rate, which we ascertain on the basis of census data or police data. We might go to schools where a high proportion of the children who are enrolled as students are eligible for free or reduced lunch. So we know it's about the population, not about individual children.

A: We also may choose the participants on the basis of clinical information, or indications that they are at-risk, have been exposed to domestic violence, or are failing at school. We have to know something about the children themselves to identify our target group. Or we might actually collect information on children's mental health status to make a decision as to who's eligible and who's not.

Developmental transitions are a good way of identifying a target group. For example, children with emotion and behavior problems are particularly vulnerable to the stresses associated with the high school transition. Just like with the selective intervention, where we might go to a school that's in a high risk neighborhood, here we're targeting a high risk developmental period. We're choosing this moment as being a time in the life of the child where a vulnerable child could get into big trouble.

A: Screening sounds easy, but there can be tremendous ethical implications of obtaining the mental health status of the student. In our study the High School Transition Study, only a small proportion of the students, even the high risk students, are going to eventually participate in the preventive intervention program. Of those who are eligible, half of them are randomly assigned to the control condition. Some of them will get no treatment, and many of the others will wait a while before they get anything. Some of them are screening so high for depression, that we're really concerned about their immediate distress. We have a whole team of mental health professionals who come in and meet with each of those students who look to be acutely distressed and assess them and make recommendations for treatment or other sorts of services and supports.

Remember that the optimal approach to case identification depends on the goals of the intervention, developmental stage, resources available obviously, and then logistical considerations about your relationship to the particular school or the particular population.