Reference: Ther Adv Respir Dis. 2022 Jan-Dec;16:17534666221099729
Practice Point: Two days of levofloxacin in addition to steroids seems reasonable for most COPD exacerbations.
EBM Pearl: Unlike with “regular” superiority trials, using a per-protocol analysis in a noninferiority trial makes it less likely that the outcome is due to chance.
Let’s face it: non-inferiority trials are confusing. Most trials seek to show that a new therapy is superior to a placebo or control. This refutes a null hypothesis, which states the new therapy and control are equivalent. Noninferiority trials flip this paradigm and instead use an established therapy as comparator hoping to show that a new therapy is equivalent to the established one.
Along comes a study looking at a curtailed duration of levofloxacin for mild/moderate acute exacerbations of chronic obstructive pulmonary disease (AECOPD). It’s a nice-sized study: 155 patients in each arm, one arm receiving seven days of levofloxacin and the other receiving two days of antibiotics and five days of placebo, each in addition to oral steroids. For those of us who prescribe quinolone antibiotics, we are well familiar with the dose-dependent side effects of this powerful but troublesome class of drugs. It would be wonderful to be able to stop antibiotics at two days instead of seven.
The primary outcome evaluated was the rate of patients with relapse-free symptom resolution 30 days after treatment initiation. The noninferiority margin was set at a 10% difference, which is both arbitrary (as most noninferiority margins are) and clinically reasonable. At 30 days, 79% of people in the 2-day group were cured compared to 74% in the 7-day group. As such, the noninferiority criteria was met and several secondary (superiority) analyses were carried out evaluating the need for more antibiotics, ICU admission, and one-year exacerbation rates. Outcomes of these secondary analyses were all nonsignificant but numerically worse in the 2 day group.
As a rule, noninferiority trials should be analyzed per-protocol as the conservative estimate to reduce alpha (type I) errors, or the probability that the outcome is due to chance. The authors here state that there were no patients who experienced treatment crossover, dropped out, or did not have full results reported. In this fairly unusual case, the per-protocol and intention-to-treat analyses would be identical!
The bottom line here seems to be that 2 days of levofloxacin really does seem noninferior to 7 days in terms of cure rates at 30 days. Since AECOPD is a common problem, we’d like to see this study repeated with more patients, with more information about the secondary outcomes evaluated here. It would also be interesting to see if other authors confirm the selected inferiority margin. Lastly, in a large enough study there will be patients who do not follow protocol, allowing us to see if the per-protocol and intention-to-treat analyses continue to match up once the data get messy!
For more information, see the topic Acute Exacerbation of COPD in DynaMed.
DynaMed EBM Focus Editorial Team
This EBM Focus was written by Dan Randall, MD, Deputy Editor at DynaMed. Edited by Alan Ehrlich, MD, Executive Editor at DynaMed and Associate Professor in Family Medicine at the University of Massachusetts Medical School; Katharine DeGeorge, MD, MS, Deputy Editor at DynaMed and Associate Professor of Family Medicine at the University of Virginia; Nicole Jensen, MD, Family Physician at WholeHealth Medical; Vincent Lemaitre, PhD, Senior Medical Writer at DynaMed; and Sarah Hill, MSc, Associate Editor at DynaMed.