The following protocol was carried out in the course of 1 day. Following informed consent, baseline measurements of respiration and surface electromyography (sEMG) signals (Viking IV; Viasys Healthcare; Madison, WI) were obtained. Rectified and moving-time averaged sEMG were measured from the sixth, seventh, and eighth intercostal spaces for eight consecutive spontaneous breaths while the patient was receiving continuous positive airway pressure, 5 cm H2O with no pressure support, through the ventilator (Puritan Bennett 7200; Tyco Healthcare; Pleasanton, CA). Maximal sEMG activity during inspiration to minimal sEMG activity at end-exhalation was examined. Simple measures of ventilation (tidal volume [Vt], minute ventilation, and frequency [f]) were captured through the ventilator display, and a serum theophylline level was obtained. The rapid shallow breathing index (RSBI) [f/Vt] was calculated from the obtained f and Vt. IV aminophylline was then administered at gradually increasing doses with the intention of reaching the therapeutic level that yielded the best results in terms of sEMG and respiratory measurements. A bolus of 2 mg/kg was administered over 30 min, followed by a continuous infusion of 0.6 mg/kg for 30 min. Measurements of respiration, sEMG, and serum theophylline were obtained immediately following completion of the continuous infusion. Separate measurements were made following the same protocol after boluses of 4 mg/kg of aminophylline followed by 30 min infusion of 0.6 mg/kg, and finally 6 mg/kg aminophylline followed by 30 min of a 0.6 mg/kg infusion. Several hours elapsed between each increase in drug dose.