Sedative and ventilatory effects of midazolam infusion: effect of flumazenil reversal.

The purpose of this study was to evaluate the effects of flumazenil (1 mg i.v.) on the ventilatory response of premedicated patients receiving a continuous infusion of midazolam for sedation. After assessing baseline ventilatory function using a modified Read rebreathing method for determining hypercapnic ventilatory drive, 16 healthy outpatients were administered fentanyl, 50 micrograms i.v., and midazolam 2 mg i.v., followed by a variable-rate midazolam infusion, 0.3-0.5 mg.min-1. Upon termination of the midazolam infusion, serum midazolam concentrations were measured and ventilatory function was reassessed. Then, 10 ml either saline or flumazenil (1 mg) were administered according to a randomized, double-blind protocol. Ventilatory function was subsequently measured at 5 min, 30 min and 60 min intervals after study drug. Compared with the baseline value, midazolam infusion reduced tidal volume and increased respiratory rate and alveolar dead space. However, midazolam did not decrease the slope of the CO2-response curve. Flumazenil reduced the degree of midazolam-induced sedation and the decrease in tidal volume (P < 0.05), but not the change in resting respiratory rate. In some patients, the ventilatory response to hypercarbia actually decreased after flumazenil administration compared with the immediate prereversal (sedated) values. It is concluded that midazolam infusion, 0.43 mg.min-1, did not impair CO2-responsiveness. Flumazenil's effect on central ventilatory drive was more variable than its reversal of midazolam-induced sedation.