CHILDREN'S SELF-COMPLETION OF AN ASTHMA HEALTH STATUS MEASURE: DOES AGE OR READING LEVEL MATTER?
Linda Radecki, Lynn M. Olson, Mary Pat Frintner.. Department of Research, American Academy of Pediatrics, Elk Grove Village, IL.

BACKGROUND: Whether children should report on their own health status is a topic of ongoing debate and a critical issue in asthma management. Limited information exists about quality of child report when items are interviewer administered, less is known about children's ability to self-complete questions.
OBJECTIVE: To examine whether age or reading level matter in children's successful completion of a pencil and paper asthma health instrument.

DESIGN/METHODS: 270 children self-completed a pencil and paper version of the Children's Health Survey for Asthma - Child Version (CHSA-C). The CHSA-C includes 3 scales: Physical Health (PH), Child Activities (CA), and Emotional Health (EH). We examined descriptive statistics, reliability, validity, and feasibility (CHSA-C completion time). SORT-R standardized scores estimated reading level (RL).

RESULTS: Mean age=12.4 years (range=10-16); 58% male; 46% African American; 41% incomes <$30k/year; 53% mod/severe asthma (parent report). Descriptive statistics revealed no differences in scale scores by age or RL. Reliability (internal consistency): For most scales regardless of age, Cronbachs alpha met/exceeded accepted standards (.70); no consistent relationship emerged for RL. Validity: Children with fewer symptom days reported higher PH scores (indicating better outcomes) in every age group. The relationship was less consistent with RL included in the analysis, especially for older children. Feasibility: 10-12 year olds took longer for CHSA-C completion than older children (8.5 vs 4.8 minutes, respectively); those with RLs below the mean took more time (8.5 vs 5.6 minutes), regardless of age.

CONCLUSIONS: Children as young as 10 can successfully provide written information related to their own asthma health. Data quality and time required for completion may be mediated by child reading ability. Investigators and clinicians should consider these factors when collecting data from children and teenagers.