Objectives: The concentrations of nitric oxide (NO) in the ventilatory circuits and the patient's airways were compared between sequential (SQA) and continuous (CTA) administration during inspiratory limb delivery.
Design: Prospective controlled study.
Setting: 14-bed Surgical Intensive Care Unit of a teaching University hospital.
Patients and participants: Eleven patients with acute lung injury on mechanical ventilation and two healthy volunteers.
Interventions: A prototype NO delivery device (Opti-NO) and César ventilator were set up in order to deliver 1, 3 and 6 parts per million (ppm) of NO into the bellows of a lung model in SQA and CTA. Using identical ventilatory and Opti-NO settings, NO was administered to the patients with acute lung injury.
Measurements and results: NO concentrations measured from the inspiratory limb [INSP-NOMeas] and the trachea [TRACH-NOMeas] using fast response chemiluminescence were compared between the lung model and the patients using controlled mechanical ventilation with a constant inspiratory flow. INSP-NOMeas were stable during SQA and fluctuated widely during CTA (fluctuation at 6 ppm = 61% in the lung model and 58 +/- 3% in patients). In patients, [TRACH-NOMeas] fluctuated widely during both modes (fluctuation at 6 ppm = 55 +/- 3% during SQA and 54 +/- 5% during CTA). The NO flow requirement was significantly lower during SQA than during CTA (74 +/- 0.5 vs 158 +/- 2.2 ml.min-1 to attain 6 ppm, p = 0.0001). INSP-NOMeas were close to the values predicted using a classical formula only during SQA (bias = -0.1 ppm, precision = +/-1 ppm during SQA; bias = 2.93 ppm and precision = +/-3.54 ppm during CTA). During SQA, INSP-NOMeas varied widely in healthy volunteers on pressure support ventilation.
Conclusions: CTA did not provide homogenous mixing of NO with the tidal volume and resulted in fluctuating INSP-NOMeas. In contrast, SQA delivered stable and predictable NO concentrations during controlled mechanical ventilation with a constant inspiratory flow and was economical compared to CTA. However, SQA did not provide stable and predictable NO concentrations during pressure support ventilation.