Variability in Usual Care Mechanical Ventilation for Pediatric Acute Respiratory Distress Syndrome: Time for a Decision Support Protocol?

Abstract

Objectives: Although pediatric intensivists philosophically embrace lung protective ventilation for acute lung injury and acute respiratory distress syndrome, we hypothesized that ventilator management varies. We assessed ventilator management by evaluating changes to ventilator settings in response to blood gases, pulse oximetry, or end-tidal CO2. We also assessed the potential impact that a pediatric mechanical ventilation protocol adapted from National Heart Lung and Blood Institute acute respiratory distress syndrome network protocols could have on reducing variability by comparing actual changes in ventilator settings to those recommended by the protocol.

Design: Prospective observational study.

Setting: Eight tertiary care U.S. PICUs, October 2011 to April 2012.

Patients: One hundred twenty patients (age range 17 d to 18 yr) with acute lung injury/acute respiratory distress syndrome.

Measurements and main results: Two thousand hundred arterial and capillary blood gases, 3,964 oxygen saturation by pulse oximetry, and 2,757 end-tidal CO2 values were associated with 3,983 ventilator settings. Ventilation mode at study onset was pressure control 60%, volume control 19%, pressure-regulated volume control 18%, and high-frequency oscillatory ventilation 3%. Clinicians changed FIO2 by ±5 or ±10% increments every 8 hours. Positive end-expiratory pressure was limited at ~10 cm H2O as oxygenation worsened, lower than would have been recommended by the protocol. In the first 72 hours of mechanical ventilation, maximum tidal volume/kg using predicted versus actual body weight was 10.3 (8.5-12.9) (median [interquartile range]) versus 9.2 mL/kg (7.6-12.0) (p < 0.001). Intensivists made changes similar to protocol recommendations 29% of the time, opposite to the protocol's recommendation 12% of the time and no changes 56% of the time.

Conclusions: Ventilator management varies substantially in children with acute respiratory distress syndrome. Opportunities exist to minimize variability and potentially injurious ventilator settings by using a pediatric mechanical ventilation protocol offering adequately explicit instructions for given clinical situations. An accepted protocol could also reduce confounding by mechanical ventilation management in a clinical trial.

Figure 1

Box Plot of PEEP/FiO₂ combinations on all patients on conventional ventilation modes in the 8 CPCCRN PICUs. PEEP/FiO₂ titration tables from clinical care for all 120 patients on conventional modes of mechanical ventilation. The Y-axis represents actual PEEP values as a function of actual FiO₂ used (X-axis). The superimposed dark blue boxes represent the pediatric mechanical ventilation protocol target combinations of PEEP/FiO₂, derived from the ARDS Network protocols. For all patients, there is variability in PEEP/FiO₂ combinations intensivists choose. In general, clinicians use less PEEP (peaking at just under 10 cmH₂O) than the protocol would recommend, particularly when FiO₂ climbs above 0.6.

1. Actual CPCCRN data (Light blue boxes); Mean value (◇), Median (Bar), IQR (Box), Range (Whiskers), Outliers (○), Protocol Targets (Dark blue bars).

2. Missing values are for 15 of the 3504 PEEP/FiO₂ observations for which no FiO₂ was recorded. In addition, there were 26 observations excluded where no PEEP was recorded.

Figure 2

The distribution of Peak Inspiratory Pressure (PIP), pH and Tidal Volume/kg actual body weight averaged for all 8 CPCCRN sites over Days 1, 2, 3 and 7 when treating patients with severe ARDS – PF ratio < 100 or SF ratio < 150. The Y-axis on the left shows the PIP (upper) and the VT (ml/kg) (lower) values and on the right, the pH values over the first 3 days of MV and Day 7, for all CPCCRN sites and all patients with severe ARDS.

1. Tidal volumes were maintained in the 7–8 ml/kg range for this group of patients with severe pediatric ARDS over the first three days. The resolution of disease over the week is implied by decreasing peak pressures and rising pHs with tidal volumes having increased up to approximately 11 ml/kg ABW.

2. Mean value (◇), Median (Bar), IQR (Box), Range (Whiskers), Outliers (○).