The Role of Circuit Flow During Mechanical Ventilation of Neonates

• Background of the project

About 1.5% of newborns require mechanical ventilation. Although mechanical ventilation can be life-saving for neonates with respiratory failure, it may cause lung injury due to excess airway pressure (barotrauma), delivery of high tidal volumes (volutrauma) and repetitive closing/re-opening of lung units (atelectotrauma)1. Very preterm neonates are especially vulnerable to ventilator induced lung injury. Ventilator associated lung injury is one of several factors contributing to the burden of chronic lung disease of infancy, also called bronchopulmonary dysplasia (BPD).

Ventilators have several different modes. The most frequently used mode is, time-cycled, pressure limited, where the doctor sets inspired oxygen level, the peak inflation pressure (PIP), rate, and inflation time of the ventilator. In this mode the inflations are usually synchronised with the baby's breathing. This is either Synchronised Intermittent Positive Pressure Ventilation (SIPPV), where the ventilator synchronises inflations with all the baby's breaths, or Synchronised Intermittent Mandatory Ventilation (SIMV), where the ventilator only synchronises with a set number of breaths. These can be delivered with or without targeting the tidal volume delivered to a set value by adjusting the peak inflating pressure. This mode is called volume guarantee (VG). A similar mode of ventilation is flow-cycled or Pressure Support Ventilation (PSV) and can be combined with tidal volume targeting 2. In this mode the inflation is terminated as soon as the flow rate falls to 15% of the maximum flow during inflation.

These modes have continuous gas flow through the ventilation circuit. Inflation starts and pressure rises when the expiratory valve is closed. The rate of circuit flow alters the ventilation waveforms. With time-cycled ventilation the higher the circuit flow the faster the pressure rises, the lung is distended and potentially injured, and the earlier the set PIP is achieved. A high flow, with a relatively long inflation time will result in a "pressure plateau", that is the PIP is sustained after the flow, into the baby, has stopped because the lung volume is maximal at the PIP used (Figure 1). In PSV inflation stops when the flow has decreased to \~15% of the peak inflation flow and so the inflation time is shorter. Increasing the device flow shortens the inflation time and thereby reduces the mean airway pressure.

Despite the effects on ventilation patterns and the speed of lung distension and injury, consideration has rarely been given to the circuit flow. By protocol, it is usually set at 7-10 L/min when the ventilator is turned on and it is not changed.

This is relatively high, and usually produces a "square" pressure waveform with a rapid distention of the lungs and a sustained pressure plateau. (see A in figure 1) With PSV a high flow results in a relatively short inflation time (\~0.2 sec; it needs to be at least 0.3 sec for adequate lung aeration). There is no evidence these high flows are best for optimum ventilation and minimum lung damage. In a preterm lamb model there were no adverse effects on gas exchange or cardiovascular parameters until the flow was reduced to 3 l/min 3. In animal studies ventilation with high flow resulted in histologic and molecular changes of lung injury 4. The effect of lowering the ventilator circuit flow rate has never been investigated in clinical studies.

The Dräger Babylog VN500 ventilator has an alternative to setting a gas flow: the user can set the slope time instead, that is time required to reach the set pressure. In Cambridge it is invariably set to 0.08 sec which results in a flow \~7 L/min. As we use an inflation time between 0.33 - 0.45 sec, there is a pressure plateau lasting at least 0.25 sec and sustained inflation with little or no flow.

The investigators are the first to develop a unique system for downloading and analysing data from the Dräger VN500 neonatal ventilator. Using DataGrabber software obtained from Dräger Medical, the investigators can retrieve ventilator parameters at 100Hz frequency over long periods (hours and even days). To analyse and visualise the large datasets the investigators developed a data analysis workflow using the Python programing language and its add-on packages (Figure 2). With this tool the investigators can now study details of each inflation and spontaneous breath. The Investigators have recorded detailed data from 30 babies and shown it is feasible, accurate, and the investigators have the expertise (Belteki et al, submitted).

In this application the investigators propose to investigate the effect of different slope times (and therefore different levels of circuit flow) on ventilation parameters and gas exchange in preterm infants. The investigators hypothesise that a longer slope time (= lower circuit flow) will distend the lungs more gently yet maintain ventilation and gas exchange.

• Intervention:

The study is a within patient crossover design comparing short periods of ventilation with different slope times both in SIPPV-VG and PSV-VG modes with the following interventions:

A ventilator download tool and transcutaneous and expired CO2 monitors is attached to the ventilator and data download commences while the baby is ventilated with the parameters used by the clinical team. An arterial blood gas is performed. The ventilated parameters are changed as shown below. The order of these epochs is randomised. Another arterial blood gas is performed. Ventilator is changed back to the original parameters (or different as appropriate by the blood gas). Ventilator data and CO2 recording will be continued for another 30 minutes.

Interventions

Duration Ventilator More Slope time Inspiratory time[max] 15 min SIPPV-VG 0.08 0.40 15 min PSV-VG 0.08 [0.60] 15 min PSV-VG 0.16 [0.60] 15 min SIPPV-VG 0.16 0.40 15 min SIPPV-VG 0.24 0.40 15 min PSV-VG 0.24 [0.60] 15 min PSV-VG 0.32 [0.60] 15 min SIPPV-VG 0.32 0.40 15 min SIPPV-VG 0.40 0.40 115 min PSV-VG 0.40 [0.60]

Total study duration is 220 minutes. A researcher will be present continuously. FiO2 will be adjusted if needed to maintain saturations between 90-95%. If the FiO2 rises >15% or the end-tidal CO2 rises >1.5kPa over the pre-study level, that intervention will be abandoned.

Comparison:

At each slope time, the following parameters will be determined and compared with those recorded at 0.08 sec slope time: (1) peak inflating pressure, (2) inflation duration, (3) duration of inflation plateau, (4) duration of no gas flow, (5) expired tidal and minute volumes (mandatory/spontaneous, inspiratory/expiratory, (6) ventilator rate, (7) FiO2, (8) transcutaneous and/or end-tidal CO2 and, (9) interaction between ventilator inflations and baby's breaths. Values for SIPPV and PSV will also be compared.