Clinical prediction rules for abusive head trauma: a systematic review

Area of Science:

• Pediatrics
• Forensic Medicine
• Medical Informatics

Background:

• Misdiagnosis of abusive head trauma (AHT) carries severe implications for child welfare and families.
• Accurate and timely identification of AHT is critical in clinical settings.

Purpose of the Study:

• To systematically review and compare clinical prediction rules (CPredRs) designed to assist clinicians in evaluating suspected abusive head trauma (AHT).
• To identify the utility and limitations of existing CPredRs in the diagnostic pathway for AHT.

Main Methods:

• A systematic literature search was conducted across major databases (MEDLINE, Embase, PubMed, Cochrane) from January 1996 to August 2016.
• Inclusion criteria focused on externally validated CPredRs for AHT detection in clinical practice.

Main Results:

• Three CPredRs were identified: Pediatric Brain Injury Research Network (PediBIRN) (sensitivity 0.96), Predicting Abusive Head Trauma (PredAHT) (sensitivity 0.72), and Pittsburgh Infant Brain Injury Score (PIBIS) (sensitivity 0.93).
• These tools are tailored for different patient populations and diagnostic junctures, with limited overlap in predictive variables.
• None of the CPredRs are intended for definitive diagnosis but serve as clinical aids.

Conclusions:

• CPredRs like PediBIRN, PredAHT, and PIBIS offer valuable support at distinct points in the AHT assessment process.
• Widespread validation across diverse clinical settings is lacking for current CPredRs.
• Further impact analyses are recommended to ascertain the safety and effectiveness of these tools in real-world clinical practice.