Pre-hospital continuous positive airway pressure after blast lung injury and hypovolaemic shock: a modelling study

Abstract

Background: In non-traumatic respiratory failure, pre-hospital application of CPAP reduces the need for intubation. Primary blast lung injury (PBLI) accompanied by haemorrhagic shock is common after mass casualty incidents. We hypothesised that pre-hospital CPAP is also beneficial after PBLI accompanied by haemorrhagic shock. Methods: We performed a computer-based simulation of the cardiopulmonary response to PBLI followed by haemorrhage, calibrated from published controlled porcine experiments exploring blast injury and haemorrhagic shock. The effect of different CPAP levels was simulated in three in silico patients who had sustained mild, moderate, or severe PBLI (10%, 25%, 50% contusion of the total lung) plus haemorrhagic shock. The primary outcome was arterial partial pressure of oxygen (PaO 2 ) at the end of each simulation. Results: In mild blast lung injury, 5 cm H 2 O ambient-air CPAP increased PaO 2 from 10.6 to 12.6 kPa. Higher CPAP did not further improve PaO 2 . In moderate blast lung injury, 10 cm H 2 O CPAP produced a larger increase in PaO 2 (from 8.5 to 11.1 kPa), but 15 cm H 2 O CPAP produced no further benefit. In severe blast lung injury, 5 cm H 2 O CPAP inceased PaO 2 from 4.06 to 8.39 kPa. Further increasing CPAP to 10e15 cm H 2 O reduced PaO 2 (7.99 and 7.90 kPa, respectively) as a result of haemodynamic impairment resulting from increased intrathoracic pressures. Conclusions: Our modelling study suggests that ambient air 5 cm H 2 O CPAP may benefit casualties suffering from blast lung injury, even with severe haemorrhagic shock. However, higher CPAP levels beyond 10 cm H 2 O after severe lung injury reduced oxygen delivery as a result of haemodynamic impairment.

Modeling Mechanical Ventilation In Silico—Potential and Pitfalls

“…The potential for studies based on in silico models to inform treatment is also being increasingly recognized in areas of medicine where it is essentially impossible to carry out RCTs, such as emergency medicine and responses to mass casualty events. 112…”

“…The potential for studies based on in silico models to inform treatment is also being increasingly recognized in areas of medicine where it is essentially impossible to carry out RCTs, such as emergency medicine and responses to mass casualty events. 112 Beyond the realm of RCTs, the use of in silico models in preclinical research into MV will continue to expand, as such models begin to match or surpass the clinical relevance of animal or in vitro models. An important factor to consider here is the inevitable continuous improvement in the fidelity and predictive capability of in silico models, as more and better data become available for model construction and validation, and as the availability of low-cost computing power continues to increase.…”