Assessing the predictive validity of developmental milestones

Abstract

Parental report of development milestones influences the physician's decision to refer for developmental evaluation and early education. The predictive validity of commonly reported milestones was evaluated by collecting data from records in a comprehensive care clinic. Multivariate analyses determined the relationships among age of milestone achievement and Wechsler intelligence Scale for Children-Revised (WISC-R) intelligence quotient (IQ) scores. Other routinely collected medical and demographic information was then added, and stepwise regressions determined the best combinations of predictors of WISC-R IQ scores. Milestones alone predicted 24 to 42% of the variance in WISC-R IQ scores, suggesting that there are meaningful relationships between the age of attainment of milestones and later intellectual functioning. However, when other medical and demographic information was added to the prediction equation, the predictive power of milestones was overshadowed by that of the new variables. Milestones are validated as one criterion for referral for specialized early development services, but adding other objective data, including standardized developmental screening, is recommended for optimal decision making.