Home event recordings of oxygenation, breathing movements, and heart rate and rhythm in infants with recurrent life-threatening events.

For an evaluation of the usefulness of event recording in identifying mechanisms for apparent life-threatening events, 94 infants (median age, 3.1 months; range, 0.5 to 12 months) with a history of two or more apparent life-threatening events of hitherto unknown cause underwent event recording of transcutaneous oxygen pressure, arterial oxygen saturation, (not recorded in all patients), breathing movements, pulse waveforms, electrocardiogram, and instantaneous heart rate. Recordings were triggered by a fall in transcutaneous oxygen pressure to less than 20 mm Hg. During a median duration of event recording of 1.3 months (0.1 to 10 months), 52 events were recorded in 34 patients; 7 of the events had to be excluded because of uninterpretable signals. The following mechanisms were identified in the remaining 30 patients (45 events): a sudden change in skin perfusion but without hypoxemia (6 in 5 patients), hypoxemia induced by an epileptic seizure (6 in 5 patients), hypoxemia induced by suffocation by a parent (4 in 4 patients), and parental fabrication of events and medical history (7 in 6 patients). In the remaining 22 events in 12 patients, the precise mechanism of events could not be identified. Analysis of these as-yet-unexplained events showed prolonged abnormal hypoxemia (a fall in transcutaneous oxygen pressure to between 4 and 18 mm Hg and a fall in arterial oxygen saturation to 5% to 75%), lasting for 40 to 500 seconds in all. Only five of these events involved prolonged (> 20 seconds) apneic pauses, and only four an episode of bradycardia (heart rate < 80 or 60 beats/min). Thus event recording identified various mechanisms of apparent life-threatening events. In the events that remained unexplained, prolonged apneic pauses or episodes of bradycardia were found in only a minority. The identification of hypoxemia therefore may be more relevant to the early detection of these events than the identification of apnea or bradycardia or both. The relevance of these findings with regard to sudden infant death syndrome remains to be determined.