Prediction of serum theophylline levels

Because theophylline (T) has a narrow therapeutic index, serum concentrations resulting from previous outpatient therapy must be considered when planning intravenous (IV) aminophylline dosage in the emergency department (ED). Facilities to measure serum T concentrations rapidly are not available in all institutions. We therefore studied our ability to predict serum T concentrations in asthma or COPD patients, and assessed the probable outcomes if dosing decisions had been based on these predictions. Forty adult patients (mean age, 41.5 years) who had taken a theophylline preparation within 12 hours of ED presentation were studied. Serum T concentration was predicted based on each patient's medication history and the use of standard pharmacokinetic equations, and was compared with the patient's actual T level. The mean actual T concentration was 12.7 +/- 6.62 micrograms/mL, while the mean predicted T concentration was 11.4 +/- 5.37 micrograms/mL. If the predicted T concentrations had been used to plan subsequent IV aminophylline dosing, 68% of patients would have received therapy likely to place them within the range of 10 to 20 micrograms/mL; 15% would have received too little aminophylline, and 17% too much. Comparable results are noted if aminophylline dosing had been based simply on the elapsed time since the patient's last oral T dose. While IV aminophylline dosing decisions can be made safely in many patients by considering their medication history or predicting serum T concentration, inappropriate and possibly dangerous therapy will result in a sizable, unpredictable minority. Aminophylline dosing in the ED ideally should be done with knowledge of the patient's base-line theophylline level.