Differential Nonenrollment of Taiwanese Children in School-Based Substance Use Prevention Studies: A Multilevel...

The present study investigated the extent to which individual and school characteristics may differentially affect parental consent and child assent in the enrollment of a school-based substance use prevention study in Taiwan.

This study linked field notes on response and consent status during enrollment of the school-based prevention study with administrative survey data reported by the targeted students when they were in fourth grade (age 10-11) (N = 2,560; 53% male, 97.8% matched). The outcome variables, defined by the combined status of parental consent/child assent, were nonresponse and negative, discordant, and positive consent. Individual characteristics included family (parental education, employment) and child (psychological/behavioral, substance use) factors. Aggregate school-level substance use and percentage of aboriginal students and nonnative parents served as school-level factors. Multilevel multinomial regression analyses were performed.

Successful consent was obtained from only 820 students (32%). Male gender and feeling neglected by families were associated with failing to respond (adjusted odds ratio = 1.78 and 1.71, respectively). Higher parental educational attainment reduced the odds of negative consent by 30%, whereas having unemployed parents increased the odds of discordant consent by 326%. Children attending schools with a higher percentage of indigenous students were two times more likely to have nonresponse, negative consent, and discordant consent.

Nonresponse to the consent request or negative consent appeared to be associated with disadvantaged background and unfavorable parent-child interaction. This suggests complex pathways underlying ascertainment and a need to modify the consent practices in school-based prevention studies involving minors, especially in schools with higher ethnic minority composition.