Parental reports of apnea and bradycardia: temporal characteristics and accuracy

A prospective examination was made of the temporal course of parental observations in response to a monitor alarm of apnea (apnea setting of 20 seconds) or bradycardia (bradycardia setting of 80 beats per minute). Data were obtained from 155 subsequent sudden infant death syndrome siblings followed up at home, during the first 20 weeks of life, on an apnea/bradycardia monitor with an attached event recorder. In addition, parental reports were compared to an objective recording of the pattern of cardiorespiratory activity surrounding each monitor alarm. Only those parental observations were considered which reported the infant to be asleep with no apparent equipment malfunction following an apnea alarm (with or without pallor, cyanosis, or the provision of external stimulation) or a low heart rate alarm associated with pallor, cyanosis, or stimulation. Observations were analyzed within each of five age periods (less than 29, 29 through 56, 57 through 84, 85 through 112, 113 through 140 days). The percentage of infants reported to have prolonged apnea, prolonged apnea with stimulation, or bradycardia with stimulation was found to decrease with age. An examination of the linked event recordings failed to document an episode of apnea as long as 15 seconds for any of the reported episodes of apnea. Furthermore, bradycardia as long as 5 seconds in duration could be documented in only 3 of 422 reported episodes of bradycardia. These results indicate the potential for considerable error when total reliance is placed on parental observations and point to the necessity for objective event recordings when using home monitors in the clinical management of at-risk infants.