Impact of a Real-Time Ventilator Management Dashboard With Alerts

Abstract

Background

Exposure to high tidal volumes in patients undergoing invasive mechanical ventilation is associated with increased mortality and worse outcomes. However, adherence to lung-protective ventilation (LPV) is suboptimal. No generalizable tool or approach exists that has been demonstrated to optimize implementation of this parameter of mechanical ventilation.

Research Question

We tested whether development and deployment of a real-time ventilator management dashboard with twice-daily alerts communicating synthesized ventilator information to clinicians would improve adherence to LPV (≤ 8 mL/kg ideal body weight).

Study Design and Methods

The Respiratory Insights Dashboard was developed to display key real-time patient and ventilator data for all patients in the ICU of a large academic medical center, as well as to send twice daily email alerts to ICU staff identifying patients with out-of-range parameters. The 5 ICUs with lowest tidal volume adherence were preselected for analysis in a quasiexperimental study comparing the periods before and after the dashboard.

Results

In the period before the intervention, 1,376 patients received ventilation, and in the period after the intervention, 3,064 patients received ventilation. Hourly adherence to LPV rose from 78% to 87% (P < .001). In patients initiated on ventilation with tidal volume (Vt) of > 8 mL/kg, the time to reach adherence fell from a median of 9 hours (interquartile range, 2-26 hours) to 5 hours (interquartile range, 1-15 hours; P = .048). Interrupted time series regression showed that temporal trends both before and after the intervention were not significantly different from zero; however, adherence level change after the intervention was 8.3 (P = .004). Increased adherence to LPV also was observed in multivariable regression (OR, 1.438; P < .001), along with an increase in adherence trend over time after implementation (OR, 1.001; P = .004).

Interpretation

In this study, a real-time ventilator management dashboard with Vt alerts was shown to significantly improve adherence to LPV hospital wide as well as more rapid correction of suboptimal ventilator settings. Implementation was associated with an increase in LPV adherence that has persisted for 15 months of follow-up.