Reproducibility of auditory brain-stem evoked responses as a function of the stimulus, scorer and subject

The reproducibility of auditory brain-stem evoked response (ABR) measurements was determined in a series of studies. ABRs were recorded from pre-term (32 and 36 weeks) newborns, term newborns, and adults to 3 different stimuli. Records were scored blindly with no knowledge of the subject, stimulus eliciting the response, or other records by that subject. Two scorers (inter-scorer reliability) and a single scorer over time (intra-scorer reliability) could score records using explicit criteria with a high degree of agreement. In contrast variability between replicate records to the same stimulus from the same subject was considerable. Intra-subject variability was not strongly affected by the age of the subject but was a function of the eliciting stimulus (the low intensity and fast repetition rate stimuli elicited more variable responses than loud, slow stimuli) and the wave measured (wave I latency was significantly more variable than the other measurements). Healthy, term newborns were studied to investigate the reproducibility of ABRs in the first few days after birth. There was a rapid reduction in ABR latency in the first 6-12 h after birth and a more gradual reduction thereafter. The prolonged latencies observed immediately after birth were associated with an elevation in ABR threshold. The reduction in latency was observed in vaginal and cesarean deliveries. These same effects were observed for amplitude of the ABR, but the trends were generally not significant. Despite the large changes in ABR latency within the first few days after birth, ABR latencies in response to some stimuli were significantly correlated between sessions separated by all intervals tested, ranging from 2 h to 2 days.