DESCRIPTION (provided by applicant): Abusive head trauma (AHT) represents the most serious form of child abuse with a 13-30% mortality and significant neurological impairment in over half of survivors (1-6). The incidence of AHT is estimated to be 23-42 per 100,000 infants (7-10) with an average of approximately 30 per 100,000 infants. With a United States population of 300 million and a birth rate of 14.1 per 1000 population, this represents approximately 1250 cases annually in the United States. An effective primary prevention strategy that would reduce the incidence of AHT would potentially save the lives of many children and improve the lives of many others. The objective of this proposal is to test two simple, low cost interventions that have a high probability of success based upon our previous work. To evaluate these interventions, we will pursue the following four specific aims: Hypothesis #1: A hospital based post-natal primary prevention program directed to parents of all newborn infants in Pennsylvania will significantly reduce the incidence of AHT. Specific Aim #1: Assess the effectiveness of an established state-wide program of hospital based post-natal parent education about violent infant shaking, provided at a single consistent time point between the infant's birth and hospital discharge, in reducing the incidence of abusive head trauma (AHT). The principal investigator has recently established in Pennsylvania (PA) a statewide hospital based primary prevention program for AHT. Parents are provided written and video materials about violent infant shaking during the newborn period, and are asked to voluntarily sign commitment statements (CS) affirming their participation. The effectiveness of this strategy will be assessed by comparing the incidence of AHT cases reported to a statewide child abuse registry with both pre-intervention rates and with contemporaneous rates derived from the multi-state Kid Inpatient Database (KID). Additional embedded time series design and case cohort and sham cohort analyses will be used. Hypothesis #2: The 'Portrait of Promise' 8-minute video will be the most important mediator of the intervention effect; the intervention effect will be more due to changes in perpetrator rather than caregiver behavior; the intervention will be less effective among high risk families. Specific Aim #2: Identify which component(s) are the most important mediators of the intervention's effectiveness; determine whether the intervention effect is more directly related to changes in perpetrator or caregiver behavior; and determine the effectiveness of the intervention among various socioeconomic groups. The relative effectiveness of four program components (brochure, video, maternity ward posters, and discussion with the nurse educator) will be assessed by tracking exposure to the components on the CS. The frequency with which identified AHT cases are exposed to each component of the program will be contrasted with the corresponding exposure among a randomly selected cohort drawn from all PA births using a case cohort analysis. Whether the program changes perpetrator behavior or teaches parents to better select caregivers for their infant will be assessed by comparing reductions among various perpetrator groups and through post-natal parent surveys of parents' perceptions of the intervention in modifying behaviors. Finally, the relative effectiveness of this intervention among high and low risk families (as identified by four key demographic characteristics - parent age, educational attainment, marital status, and race/ethnic background - shown to be statistically associated with AHT in PA) will be compared.