Differences in the reported prevalence of risky behavior
among childhood cancer survivors may be attributed to variations
across studies in regard to control group selection,
definitions of health behavior/use, method of assessment,
source of report, setting of data collection, or participant
age/developmental level