Adolescent preventive service guidelines emphasize screening and counseling for health-compromising behaviors. Most reports suggest providers deliver fewer services than recommended, but the validity of available data is unknown. We used audiotape records and postvisit adolescent and provider surveys from preventive visits by 11- to 21-year-olds at two health centers to assess the agreement and accuracy of adolescent and provider reports of issues discussed during encounters. Adolescent-provider agreement was estimated using Kappas (κ). Accuracy was assessed by sensitivity and specificity; issues were considered truly discussed if identified on the audiotape or reported by both the provider and the adolescent to occur.

Sixty-three of 70 (90%) eligible adolescents were enrolled. Audiotapes were available from 51 (81%) visits. Mean age was 15.4 years; 46% were female and 29% non-white. Agreement was excellent between adolescents and providers for having discussed seatbelt use (κ=.90); moderate for diet (κ=.69), smoking (κ=.63), drugs (κ=.62), HIV (κ=.50), drinking/driving (κ=.48), and condoms (κ=.45); and poor for alcohol use (κ=.37) and exercise (PE, κ=.36). Adolescent self-report was highly accurate. Provider reports had acceptable sensitivities, but poorer specificities. Table

Table 1
Full size table

Adolescent-provider agreement about issues discussed during preventive visits varies, but adolescents' recall is highly accurate. Adolescent self-report is an appropriate method to assess health counseling provided in preventive visits.