Abstract

BackgroundThe MIECHV (Maternal, Infant, and Early Childhood Home Visiting) program invests substantial federal resources to prevent child maltreatment and emergency medical costs. Eligibility is based on screening of demographic or clinical risk factors, but because screening accuracy in predicting poor outcomes is unknown, assignment to home-visiting might miss high-risk families or waste resources on low-risk families. ObjectivesTo guide eligibility decisions, this study tested accuracy of demographic and clinical screening in predicting child maltreatment and emergency medical care. Participants and settingA population-representative sample of 201 birthing mothers (39.8% Black, 33.8% Latina) in Durham, NC, was enrolled between July 2009, and December 2010, and followed through December 2015. MethodsParticipants were screened demographically (i.e., Medicaid, first-born, teenage, no high school diploma) and clinically (i.e., health/health care, parenting readiness, home safety, and parent mental health) at birth and followed through age 60 months, when Child Protective Services and hospital records were reviewed. Cox hazard models tested accuracy of prediction from screening variables. ResultsDemographic factors did not significantly predict outcomes, except having Medicaid/uninsured predicted more emergency medical care and being first-born was a (surprising) protective factor against a child maltreatment investigation. In contrast, clinical factors strongly predicted both maltreatment investigations (Hazard Ratio = 4.01 [95% CI = 1.97, 8.15], sensitivity = 0.70, specificity = 0.64, accuracy = 0.65) and emergency medical care (Hazard Ratio = 2.14 [95% CI = 1.03, 2.14], sensitivity = 0.50, specificity = 0.69, accuracy = 0.58). ConclusionsEven with added costs for clinical screening, selecting families for home visiting based on assessed clinical risk will improve accuracy and may yield a higher return on investment. The authors recommend a universal system of screening and care to support birthing families.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call