Statistical analysis of drug interactions in anesthesia

Anesthesiologists often use more than one drug in a patient to achieve a target response, such as a desired blood pressure.Isobolographic analysisis a standard mathematical method to test whether a drug-drug interaction exists and, if so, whether the interaction is synergistic or antagonistic. Several experimental protocols are suitable to collect data for interpretation by isobolographic analysis. Traditionally, each subject would receive a single dose of one or more drugs and the presence or absence of the target response in the subject at a specific time would be recorded. An alternative procedure is to infuse one or more drugs into each subject until the target response occurs. This procedure is clinically relevant to studies of anesthesia drugs, which are often titrated to achieve a target response. This latter method of testing for drug interaction requires fewer subjects than the traditional approach because more information is obtained from each subject. We present a statistical test for drug-drug interaction that uses the total doses of drugs given to each subject. This statistical test can be used as part of a complete analysis of drug-drug interaction. We consider (i) selection of appropriate sample sizes and doses; (ii) randomizing subjects to groups; (iii) effects of unequal group variances on the accuracy of the statistical test; (iv) appropriate methods to present the drug-drug interaction data graphically; and (v) tests to ensure that statistical assumptions are satisfied. Using the statistical methods described here, drug-drug interaction can be quantified and evaluated for statistical significance.