The validity of peer responses as a tool for screening at-risk students: a preliminary analysis.

Students are becoming the majority of new amphetamine users in Thailand. This study compared urinalysis results with peer responses to individual characteristics related to substance use with the aim of identifying "at-risk" students. A randomly selected group of students from a public high school in northern Thailand was asked to fill out the names of classmates they viewed as having any of forty-three risk behaviors set out in a questionnaire. A total of 564 students were included, from whom urine specimens were collected on the first two days following the school break. An immunoassay test was used to screen the specimens and positive results were confirmed using thin-layer chromatography. About 4% of urinalysis results were methamphetamine-positive. Using urine test results as the standard, the sensitivity of peer responses to alienated behavior was 81.8%, while frequent class/school absenteeism and low concentration levels were somewhat lower, at 77% and 68%, respectively. Delinquency showed the least sensitivity at 50%. The McNemar chi2 test showed significant differences between urine test results and each peer response subscale (p < 0.001). This preliminary analysis has shown that peer responses with regard to substance-related behavior compare well with urine test results.