It is important to differentiate the IFI from the DFI, which has been reported in several studies30–33 (Table 2
). These studies,30–33 have a number of methodologic flaws. They used an end point of the time until the next AECB episode regardless of its etiology (bacterial or nonbacterial). The literature2,34 suggests that approximately 50% of all AECBs are nonbacterial. Subjects with initial failed treatment were excluded from the DFI group, and the proportion of the AECBs that were documented to be ABECBs were very low, similar to many short-term antibiotic studies.2,17,34 In one example, a recently published clinical trial33 compared moxifloxacin and standard therapy for treatment of AECB with monthly assessments up to 9 months after treatment. A longer mean DFI was observed in the moxifloxacin-treated group compared to the clarithromycin-treated group (132.8 days vs 118.0 days, respectively; p = 0.03), although no difference was observed in the occurrence of a new exacerbation during the follow-up period. However, only approximately 20% of the intent-to-treat population included in the follow-up analysis had a documented bacterial infection at study entry, and bacterial causes of the subsequent exacerbation were not reported. Considering these concerns of DFI studies, the value of the DFI as an index of efficacy for an antimicrobial therapy is questionable. IFI studies designed to avoid the previously noted negative factors provide a more accurate assessment of the effectiveness of a particular antibiotic for the treatment of ABECB. However, the controlled trials31–32 that have suggested an improved DFI with a quinolone have confirmed a higher bacterial eradication (particularly of H influenzae) with the respiratory quinolones.