• ventilators
• patient monitors

With an increasing number of complex devices being used in neonatal intensive care units (NICUs) around the world, the alarms generated by these devices become a major burden on both patients and staff, leading to increased cardiorespiratory instability in patients and higher stress levels as well as a desensitisation to alarms (‘alarm fatigue’) among staff members.1 2 The latter may result in dangerously long response times to true alarms. Among these alarms, those generated by mechanical ventilators, although highly prevalent and often with particularly disturbing sound characteristics, have yet received surprisingly little attention. This is probably because, in contrast to patient monitors, there were no tracking systems available to objectively document occurrence rates for these alarms.

Belteki and Morley report results of a prospective study in which they electronically recorded data on the number and causes of alarms generated by a standard neonatal ventilator. They found an average of 10 such alarms per hour, similar to the rates reported for neonatal ICU monitors,1 3 and considerable variability between infants, with some spending more than 10% of their time with ventilator alarms. A large proportion of these alarms was potentially avoidable, as they resulted, for example, from inappropriate alarm settings, such as low expired Vt alarms caused by too low a maximum pressure setting during volume-targeted ventilation or inappropriate settings of the high …