The federal government recognized the importance of early and accurate identification of children with developmental delay or disability through legislation, the Individuals with Disabilities Education Improvement Act (IDEIA), which requires states to set up methods to find eligible children. Epidemiologists suggest that 10% to 17% of the pediatric population could benefit from early intervention, yet less than half are identified by elementary school age. Physical and occupational therapists have the vision, knowledge, and skills to be actively involved in community child find efforts to identify these children. The purpose of this dissertation was to create a screening procedure, combining Dworkin’s model of surveillance with detection, and determine efficacy. The model used two steps, parent concerns and an observational screening tool, to identify children with developmental delay. An observational and correlational screening study was held in Houston County, Minnesota over a one year time period. Parents of eligible infants were invited to attend one of five clinics. Of the 213 eligible families invited to attend screening clinics, 55 infants were screened. Forty three were evaluated by the principal investigator, who was blind to all screening information. Parent concerns were measured in step one, by their choice to attend screening clinics and also their scores on the Parent Concerns Survey, a questionnaire including selected Parent Evaluation Development (PEDS) questions and two additional questions eliciting parent estimation of their child’s development and feeding concerns. Parents who chose to attend the screening clinics were later compared to a random group of parents who chose not to participate, to ensure that the two groups were comparable. Infants identified by parent concerns were tested with an observational screening test, the Meade Movement Checklist (MMCL) in step two. Parents attending the clinics had significantly more concerns on the modified PEDS than parents who did not attend. Sixty one percent of parents expressed concerns about feeding issues. Significant correlation was found between the MMCL and Movement Assessment of Infants and Bayley Scales of Infant Development II. No correlation occurred with scores between 4 and 6 months of age and the pass/fail on the Ages and Stages Questionnaire between 8 to 9 months of age. The specificity of the modified PEDS was 48% and sensitivity was 80%. The following psychometric qualities were found for the MMCL: specificity 88.2; sensitivity 77.8; positive predictive value (PPV) 63.6%; and negative predictive value (NPV) 93.8%. When Bayley Scale scores and medical diagnosis categories were combined to determine referrals, 15 infants were referred for further evaluation (27.7%). Referrals represented 7.04% of the 213 families’ invited and 34.8% of the independent evaluations completed. Screening 25% of a birth cohort identified by parent concerns, resulted in high specificity (88.2%) and high sensitivity (77.8%) in identifying infants at 4 to 6 months of age. Increasing specificity and combining two tests effectively increased PPV to 63.6% and NPV to 93.8% in this community. The model decreased professional time to successfully address current barriers to infant screening recommendations. Priorities for future research include further investigation of methods to estimate the number of clinic attendees, maintain screener reliability over time, create flexible testing times, and evaluate dollar costs of clinic and follow up personnel.
