Developing quality measures for adolescent care: Validity of adolescents' self-reported receipt of preventive services

Objective. To demonstrate the feasibility of directly surveying adolescents about the content of preventive health services they have received and to assess the validity of adolescent self-reported recall. Data Sources/Setting. Audiotaped encounters, telephone interviews, and char t reviews with 14-21 year olds being seen for preventive care visits at 15 pediatric and family medicine private practices, teaching hospital clinics, and health centers. Design. 537 adolescents presenting for well visits were approached, 400 (75 percent) consented, 374 (94 percent) were audiotaped, and 354 (89 percent) completed telephone interviews either two to four weeks or five to seven m onths after their visits. Audiotapes were coded for screening and counselin g across 34 preventive service content areas. Intraobserver reliability (Co hen's kappa) ranged from 0.45 for talking about peers to 0.94 for discussin g tobacco. The sensitivity and specificity of the adolescent self-reports w ere assessed using the audiotape coding as the gold standard. Results. Almost all adolescents surveyed (94 percent) remembered having had a preventive care visit, 93 percent identified the site of care, and most (84 percent) identified the clinician they had seen. There was wide variati on in the prevalence of screening, based on the tape coding. Adolescent sel f-report was moderately or highly sensitive and specific at two weeks and s ix months for 24 of 34 screening and counseling items, including having dis cussed: weight, diet, body image, exercise, seatbelts, bike helmet use, cig arettes/smoking, smokeless tobacco, alcohol, drugs, steroids, sex, sexual o rientation, birth control, condoms, HIV, STDs, school, family, future plans , emotions, suicidality, and abuse. Self-report was least accurate for bloo d pressure/cholesterol screening, immunizations, or for having discussed fi ghting, violence, weapon carrying, sleep, dental care, friends, or over-the -counter drug use. Conclusion. Adolescents' self-report of the care they have received is a va lid method of determining the content of preventive health service delivery . Although recall of screening and counseling is more accurate within two t o four weeks after preventive care visits, adolescents can report accuratel y on the care they had received five to seven months after the preventive h ealth care visits occurred.