Analysis of purported erythromycin–warfarin interaction

In their article about a potential interaction between warfarin and erythromycin ointment, Parker et al.1 applied the Drug Interaction Probability Scale (DIPS)2 to assess the probability of a causal relationship between the purported drug interaction and changes in the International Normalized Ratio. Before applying a tool such as the DIPS, one must first establish that an interaction between the drugs could occur. The case report did not define the prescribed dosage of erythromycin ointment or how the ointment was actually used by the patient. Knowing the dosage would enable the authors to estimate the systemic exposure to erythromycin. Although the authors did not report any data describing the plasma concentrations resulting from the application of the ointment, one could assume 100% absorption and estimate plasma concentrations on the basis of the patient’s dosage. This worst-case scenario would at least provide some idea if adequate erythromycin was administered to produce an interaction with warfarin. If the estimated erythromycin concentration was below those reported to inhibit cytochrome enzymes, the erythromycin could not have altered warfarin metabolism, and trying to establish causation would be pointless.