Luft and colleagues [1] concluded that a 20-day course of amoxicillin is significantly more effective than a 7-day course of azithromycin for the treatment of erythema migrans skin lesions, a characteristic feature of early Lyme disease. They further state that their findings "[contrast] strikingly with the activity reported for various other agents that are currently being used as standard therapies for erythema migrans" and cite a previous publication of ours that reported the effectiveness of both cefuroxime axetil and doxycycline in treating erythema migrans-positive patients with early Lyme disease [2].

It should be noted that we subsequently reported the results of a second, independent study comparing cefuroxime axetil and doxycycline in the treatment of erythema migrans-positive patients with early Lyme disease [3]. Although this study was not cited by Luft and colleagues, it was similar in size to their study (232 compared with 246 patients) and again showed the effectiveness of cefuroxime axetil and doxycycline.

When one looks carefully at the paper by Luft and colleagues [1], it is apparent that methodologic differences between studies were not considered in reaching the conclusion that "... amoxicillin did extraordinarily well compared with previous reports from studies of this and other antibiotics." Luft and colleagues used less stringent criteria to define a favorable clinical response. Whereas patients in whom as much as 25% of symptoms occurred after treatment were assessed by Luft and colleagues as having had "complete response," a patient had to have no unresolved signs or symptoms to be assessed as a clinical success in our studies of cefuroxime axetil and doxycycline. Furthermore, in the study by Luft and colleagues, patients could have persistent erythema migrans and be assessed as having had "partial response" (such patients were combined with "complete responders" to give the overall clinical response), whereas any patient with persistent erythema migrans was considered to be a treatment failure in our studies. Thus, the rates of satisfactory outcomes (success plus improvement) achieved in our two studies (cefuroxime axetil, 93% and 90%, respectively; doxycycline, 88% and 95%, respectively) compare favorably with the reported rates for azithromycin and amoxicillin (97% and 100%, respectively) in the study by Luft and colleagues when the more stringent criteria used to assess satisfactory clinical outcomes in our studies are considered.

The paper by Luft and colleagues [1] illustrates clearly the difficulty of comparing the results of clinical trials that use different efficacy end points, particularly for a condition such as Lyme disease for which objective diagnostic criteria are limited. Although we agree that the double-blinding used by Luft and colleagues is desirable, we believe that when the differences in the outcome assessment criteria are considered, the findings of Luft and colleagues do not contrast strikingly with those reported in our studies of cefuroxime axetil and doxycycline.