Derivation of a Clinical Prediction Rule for Pediatric Abusive Head Trauma*

• DOI: 10.1097/pcc.0b013e3182712b09
• Publication Date: 2013-01-10T17:16:02Z
• Authors: Kent P. Hymel; Douglas F. Willson; Stephen C. Boos; Deborah A. Pullin; Karen Homa; Douglas J. Lorenz; Bruce E. Herman; Jeanine M. Graf; Reena Isaac; Veronica Armijo-G...; Sandeep K. Narang
• Keywords: Clinical prediction rule; Head trauma

Objectives: Abusive head trauma is a leading cause of traumatic death and disability during infancy early childhood. Evidence-based screening tools for abusive do not exist. Our research objectives were 1) to measure the predictive relationships between isolated, discriminating, reliable clinical variables 2) derive reliable, sensitive, prediction rule that—if validated—can inform pediatric intensivists' decisions launch (or forego) an evaluation abuse. Design: Prospective, multicenter, cross-sectional, observational. Setting: Fourteen PICUs. Patients: Acutely head-injured children less than 3 years old admitted intensive care. Interventions: None. Measurements Main Results: Applying priori definitional criteria trauma, we identified that discriminating calculated likelihood ratios post-test probabilities abuse, applied recursive partitioning with maximum sensitivity—to help out if negative. Pretest probability (prevalence) in our study population was 0.45 (95 209). Post-test ranged from 0.1 0.86. Some these variables, when positive, shifted abuse upward greatly but changed it little Other negative, largely excluded increased only slightly positive. demonstrated poor inter-rater reliability. A cluster five available at or near time hospital admission 97% patients meeting trauma. Negative value 91%. Conclusions: more completeunderstanding specific qualities could improve accuracy. If validated, be used by intensivists calculate evidence-based, patient-specific estimate can inform—not dictate—their