A NEW METHOD TO ANALYZE LUNG COMPLIANCE WHEN PRESSURE-VOLUME RELATIONSHIP IS NONLINEAR

Changes in dynamic lung compliance during inspiration and expiration c annot be modeled accurately with conventional algorithms. We developed a simple method to analyze pressure-volume (P/V) relationships under condition of nonlinearity (APVNL) and tested it in a lung model with k nown resistance and nonlinear P/V relationship. In addition, pulmonary mechanics in 22 infants, 11 of them with nonlinear P/V relationships, were analyzed with the new method. The findings were compared with th ose obtained by a recently introduced algorithm, multiple linear regre ssion analysis (MLR) of the equation of motion. The APVNL method descr ibed the changing compliance (C) of the lung model accurately, whereas the MLR method underestimated C especially in the first half of the b reath. In infants the MLR method gave highly variable, often nonphysio logical C values in the beginning of a breath. In contrast, the coeffi cient of variability of measurements obtained by the APVNL method was significantly smaller (p < 0.02), and the indices of model-fit showed better agreement between calculated and observed pressure than for the MLR method (p < 0.02). We conclude that the APVNL method accurately d escribes nonlinear P/V relationships present during spontaneous breath ing or mechanical ventilation. The method may be helpful in identifyin g and preventing pulmonary overdistention.