2008), trauma (Cohen, Kelleher, and Mannarino, 2008), maternal depression (Grupp-Phelan, Wade, et al., 2007), suicide (Aseltine and DeMartino, 2004), and drug abuse (McCabe, 2008), to name a few. The large number of tools available reflects the spectrum of problems and developmental stages to be screened, as well as perhaps the lack of standardization of approaches in this field.
The sensitivity (the ability to accurately identify individuals at risk) and specificity (the ability to accurately identify those not at risk) of available screening tools are important considerations (Meisels and Atkins-Burnett, 2005; Glascoe, 2000). On one hand, a high false-positive rate compounds the problem of stigmatization of potentially healthy children. On the other hand, an excessive false-negative rate will preclude many children in need from being identified and getting the early intervention services needed to keep them healthy. Most of the instruments reviewed have sensitivities and specificities in the 70-90 percent range, which is acceptable for screening. Positive and negative predictive values (the probability of disease among those with a positive test and the probability of no disease among those with a negative test, respectively) are usually not reported in these analyses. The committee did not systematically review the evidence related to all screening tools but was struck by the breadth of available tools.
Adaptation of screening tools for specific ethnic/cultural groups may be required. Psychometric properties are not always demonstrated for these groups (Pignone, Gaynes, et al., 2002). Children from culturally or linguistically distinct backgrounds may respond differently than majority youth not only to the screening instrument, but also to the screening process itself (Snowden and Yamada, 2005). In addition, behaviors and emotions that tools identify as dysfunctional may be adaptive in the sociocultural and physical environments of some ethnic minority children and families (Canino and Spurlock, 1994; Dubrow and Garbarino, 1989). Although race and ethnicity are often confounded with socioeconomic status, and socioeconomic status is the stronger predictor of MEB disorders, efforts to increase the cultural relevance, including the linguistic acceptability, of screening tools warrant attention.
Screening guidelines should be acceptable to the population and not cause labeling. Historically, the U.S. public has favored the opportunity to gain knowledge of potentially adverse medical situations or outcomes so that action can be taken to avoid the consequences. For example, all states have newborn screening programs in place, many of which test for 20, 30, or even more serious disorders. However, circumstances related to prevention of MEB disorders may frame this point of view differently. Some people do not want to acknowledge or think about mental illness. When screening results have the potential to adversely label or stigmatize young