• J W Skimming, P B Blanch, and M J Banner.
• Department of Pediatrics, University of Florida College of Medicine, JHMHC, Gainesville 32610-0296, USA.
• Crit. Care Med. 1997 Aug 1; 25 (8): 1410-6.

ObjectivesThis study was designed to test the hypothesis that the practice of infusing nitric oxide at constant flow rates directly into breathing circuits with intermittent (pulsatile) flow can lead to streaming and tidal pooling of the nitric oxide. This study was also designed to show the extent to which streaming and tidal pooling of nitric oxide affect nitric oxide delivery.DesignA series of five in vitro experiments was performed. For each experiment, either one or two features of the nitric oxide delivery/sampling system were varied, and the effects of these variations were evaluated with regard to measured nitric oxide concentration changes. The results from each experiment were analyzed using either one- or two-factor analysis of variance.SettingUniversity research laboratory.SubjectsBreaths were provided by a mechanical ventilator that was connected to a lung model. A standard, corrugated, adult breathing circuit was used. Gas samples were obtained from either the lung model's bellows or selected sites within the breathing circuit. Nitric oxide concentrations were measured, using an electrochemical gas analyzer.InterventionsThe system features that were varied included the cross-sectional position of the sampling site within the breathing circuit, the distance between the infusion port and the sampling site, the breathing frequency, the distance between the Y-piece and the infusion port, and the airway (deadspace) volume.Measurements And Main ResultsStreaming of nitric oxide within the breathing circuit was detected as far as 25 cm downstream of the infusion site (p < .0001). Pooling of nitric oxide was detected both near and downstream of the infusion site (p < .0001). Increasing the breathing frequency from 5 to 30 breaths/min increased mixing thoroughness (p < .005). Increasing the distance between the Y-piece and the infusion port from 15 to 180 cm decreased nitric oxide delivery to our lung model (p < .0001). Interestingly, increasing airway (deadspace) volume from 150 to 450 mL decreased nitric oxide delivery to our lung model (p < .0001).ConclusionsEstimates of nitric oxide delivery using a constant flow rate of nitric oxide infused directly into a breathing circuit during controlled mechanical ventilation can be confounded by streaming and tidal propagation of nitric oxide pools. Improved reproducibility of reported dose-response relationships is likely to be achieved through further study of nitric oxide behavior within the breathing circuits. Reduced toxicity associated with nitric oxide inhalation may also be achieved through a better understanding of this nitric oxide behavior.

23 keywords...

hide…