Pharmacokinetic optimisation of aminophylline infusions in critically ill patients

The method of Chiou et al. was used to predict theophylline kinetics in eleven critically ill patients with either acute severe asthma or bronchoconstriction. Following the commencement of an accurately metered infusion of aminophylline, venous blood samples were taken at approximately 1, 5 hours and 7-12 hours for measurement of plasma theophylline concentration. The 1- and 5-hour levels were used to estimate total body clearance and plasma concentration of theophylline at the 7-12 hour sampling time. Using these values, the infusion rate was adjusted if necessary and the protocol repeated. Initial predictions were unreliable in two patients because of continued absorption of theophylline from pre-infusion therapy with aminophylline suppositories or slow-release theophylline tablets. In the remaining studies there was a significant correlation (y = 0.9x + 0.55, r2 = 0.93, p less than 0.01, n = 19) between predicted and actual plasma concentrations at the 7-12-hour sampling time. In three patients, sequential estimates of theophylline clearance showed an approximate twofold variation and in another two patients, there was evidence of concentration-and/or time-dependent theophylline kinetics.