Ventilator Alarm Selection in the ICU

Abstract

Background:

Ventilator alarm selection in the ICU that has been poorly studied; which raises a concern for patient safety. We have previously reported that alarms are set very loosely probably due too the high number of false alarms and the fatigue they generate on the clinician. The results have been shared with clinicians and RT students who work/rotate at the institution where the study has been conducted. The goal of this study was to evaluate if selection of ventilator alarm parameters has changed over the last couple of years.

Methods:

A retrospective review of medical records of patients admitted to a medical-surgical ICU between May of 2013 and 2018 was conducted. The alarm settings selected for analysis were high respiratory rate (HI RR), high peak inspiratory pressure (HI PIP), and high and low minute volume (HI MV, LO MV), as they represent the most frequently documented alarms. SPSS 25.0 was used for data analysis. Means and standard deviations were calculated. A t-test was used to compare groups and statistically significant difference was set at a P < .05.

Results:

Recorded ventilator parameters and alarm settings of mechanically ventilated patients admitted to six ICUs at a university-affiliated, 496-bed hospital, in San Antonio, Texas. A total of 201ventilator checks were recorded. Patients monitored parameters were not significantly different between 2013 and 2014-2018 [19.6 vs. 19.8 ± 6.1 breaths/min); MV 10.3 vs. 8.1 ± 3.9 L/min); PIP 22.7 vs. 20.6 ± 7.3 cm H₂O); P = .12]. The selected alarm settings deviated from patient parameters in a range similar to the previous year (67% to 155% in 2013 vs. 61% to 159% ± 46.6% 2014-2018). In the 2014-2018 period, there was a significantly higher difference between the alarm and the patient setting for the HI PIP (P < .001), the HI MV (P = .009), and RR (P = .03) while no significant changes in selection occurred for LO MV (P = .29)]. Only 9 out of the 201 alarms (6.1%) were consistent with recommended limits.

Conclusions:

Our results suggest that there is an “alarming” need to review the practice of ventilator alarm selection among RTs who routinely manage patients on the ventilator. This practice appears to be poorly individualized to patient clinical parameters.