Reliability of Motor Evoked Potential Identification in Pediatric Populations

Area of Science:

• Neuroscience
• Developmental Pediatrics
• Biomedical Engineering

Background:

• Transcranial magnetic stimulation (TMS) elicits motor evoked potentials (MEPs) for assessing neuromotor development.
• Reliability of rater-based MEP identification in pediatric populations, especially infants, is not well-established.
• This study evaluates interrater reliability (IRR) and consensus for MEP identification in adolescents and infants.

Purpose of the Study:

• To assess the interrater reliability (IRR) of motor evoked potential (MEP) identification in pediatric populations.
• To compare MEP identification reliability between adolescents and infants.
• To identify factors influencing disagreement in MEP scoring.

Main Methods:

• Three experienced researchers independently scored electromyography (EMG) trials from adolescents and infants using a 5-point confidence scale.
• Interrater reliability (IRR) was calculated using intraclass correlation coefficients (ICC).
• Consensus rates and signal characteristics were analyzed to predict scoring disagreements.

Main Results:

• Excellent IRR for adolescent MEP scores (ICC=0.973) and good IRR for infant MEPs (ICC=0.784).
• Consensus rates were significantly lower for infants (76.8%) versus adolescents (89.1%).
• Higher baseline EMG variability and lower peak-to-peak amplitudes were linked to reduced consensus.

Conclusions:

• Rater-based MEP identification is feasible in pediatric cohorts but more challenging in infants.
• Good reliability is achievable in infants with multiple trained raters and consensus scoring.
• Findings support consensus-driven best practices for MEP identification in early developmental research.