Predictors of screening and injury in contacts of physically abused children.

Abstract

To determine rates of screening in contacts of children evaluated for physical abuse, and the relationship of clinical characteristics to screening recommendation and completion and injury identification. This is a planned secondary analysis of a prospective study of 1918 contacts of 1196 children referred for subspecialty abuse consultation in 20 US centers. We used multivariable logistic models to determine the relationship of index child characteristics, contact child characteristics, and shared characteristics to screening and injury identification. We identified injuries or disclosures of abuse in 180 (9.4%) contacts. Recommended screening was omitted in >20% of subjects for each screening modality. At least 1 screening test was more likely to be completed in contacts of index children of non-White race or Hispanic ethnicity (OR 1.45, 95% CI 1.13-1.87), with abuse-specific injuries (OR 2.15, 95% CI 1.63-2.83), with a confession (OR 2.18, 95% CI 1.17-4.07), when the history changed (OR 1.65, 95% CI 1.05-2.61), when an occult injury was found by imaging in the index child (OR 1.84, 95% CI 1.39-2.43), and when families lacked private insurance (OR 1.63, 95% CI 1.15-2.31). Completion of screening recommended for contacts of potentially abused children is relatively poor, despite high risk of injury. Several clinical and demographic factors were associated with increased contact screening.