Comparison between pulse oximetry and transthoracic impedance alarm traces during home monitoring.

OBJECTIVE

To compare transthoracic impedance (TTI/ECG) and pulse oximetry alarm traces detected during home monitoring in infants at risk of apnoea, bradycardia and hypoxaemia.

STUDY DESIGN

A retrospective evaluation of the monitor downloads of 67 infants who had undergone either TTI/ECG or pulse oximetry home monitoring using a device which can detect both parameters.

METHODS

The patients were categorised as: apparent life-threatening events (n = 39), preterm infants (n = 21) and miscellaneous (n = 7). TTI/ECG and pulse oximetry alarm traces were scored as either true or false alarms. Classification criteria were based on visual analysis of the impedance and plethysmographic waveforms captured by the memory monitor every time alarm thresholds were violated.

RESULTS

5242 alarms occurred over 3452 days of monitoring: 4562 (87%) were false and 680 (13%) true. The mean duration of monitoring was 51 days (range 5-220 days). There were 2982 TTI/ECG false alarms (65% of the total) and 1580 pulse oximetry false alarms (35%) (p = 0.0042). Of the 680 true alarms, 507 (74%) were desaturations not attributable to central apnoea and 173 (26%) were true TTI/ECG alarms (p = 0.0013).

CONCLUSIONS

Comparison of pulse oximetry and TTI/ECG alarm traces shows that true events were mostly attributable to desaturations, while false alarms were mainly provoked by TTI/ECG. The total number of false alarms is lower than reported in other studies using TTI/ECG only, thus indicating that monitoring using both pulse oximetry and TTI/ECG is suitable for home use. When the combination of both techniques is not feasible or not required, we recommend the use of motion resistant pulse oximetry alone.