Validation of respiratory mechanics software in microprocessor-controlled ventilators

BACKGROUND AND METHODSSeveral microprocessor-controlled ventilators, available for clinical use, contain optional computer software programs capable of performing near-instantaneous determinations of airway resistance and lung compliance. This study was undertaken to determine the validity of the measurements for airway resistance and lung compliance obtained by the software packages on three microprocessor-controlled ventilators. Three ventilator models were studied. An artificial ventilator-patient circuit was constructed using a test lung and an endotracheal tube. Airway pressure and gas flow curves were recorded using conventional means. Static lung compliance and airway resistance were calculated using standard equations, while automated measurements were obtained from the ventilators. The following parameters were then varied to simulate a wide variety of clinical situationstidal volume, peak inspiratory flow rate, respiratory rate, endotracheal tube, and test lung compliance. RESULTSAutomated measurements were highly correlated with values obtained manually (resistancePuritan-Bennett 7200a r = .94, Bear 5 r = .98, Veolar r = .96; compliance7200a r = .93, Bear 5 r = .97, Veolar r = .97). Calculated limits of agreement between the two methods demonstrate that although not in absolute agreement, the software-determined values for airway resistance and lung compliance differed from the manually derived values in a ventilator-specific, predictable fashion. CONCLUSIONSThe correlation and agreement demonstrated between values of airway resistance and lung compliance measured by the respiratory mechanics software packages and those values derived manually suggest that these software packages may be useful for measuring trends, as well as responding to treatment in the clinical setting. These results apply only to the controlled, mechanical ventilation mode. Further studies are indicated to validate this software in patients capable of generating spontaneous breaths. (Crit Care Med 1992; 20:1152–1156)