Role of oxygen in a bubble model for predicting decompression illness.

Abstract

A mathematical model is developed for predicting the incidence of decompression illness (DCI) in dives involving nitrogen and/or helium. The model is based on the growth of a bubble that is predicted to form upon a condition of gas superstauration. The model compartment is assumed to be perfusion-limited, hence, the uptake and elimination of each gas is distinguished by its partition coefficient (solubility ratio of blood to tissue). Bubble evolution is predicted by assuming either a diffusion limitation in gas transfer or complete equilibrium between the gas and tissue phases. An oxygen effect is implemented by 1) considering the possibility that oxygen may contribute to the bubble in a gaseous form when its ambient level exceeds an estimated threshold and 2) taking into account the vasoconstrictive response induced by high levels of oxygen breathing. Risk of DCI is hypothesized to increase as the time-integrated bubble size increases. The model parameters described above were calibrated using the method of maximum likelihood against a total of 2,738 man-dives (715 involving helium) with a 4.56% incidence of DCI. The best fit of the data was obtain ed with the equilibrium approximation and the combined oxygen effect. <TRUNCATED>