A Noninferiority Margin for Acne Lesion Counts
- PDF / 8,694,737 Bytes
- 9 Pages / 612 x 792 pts (letter) Page_size
- 82 Downloads / 163 Views
A Noninferiority Margin for Acne Lesion Counts
Christoph Gerlinger, PhD Senior Director Statistics. Bayer Schering Pharma AG, Berlin. Germany Gerald Stiidtler, MSc Clinical Statistician, lntendis GmbH. Berlin, Germany
Rolf Gotzelnonn, MSc Senior Statistical Programme6 Bayer Schering Pharma AG. Berlin, Germany
Klaus Graupe, MD, PhD Senior Medical Advisor. lntendis GmbH. Berlin, Germany
Jan Endrikat, MD, PhD Head Clinical Development and Drug Safety, Bayer lnc., Toronto. Ontario. Canada
Key Words Noninferiority margin; Acne; Choice ofdelta; Active control Correspondence Address Dr. Christoph Gerlinger, Bayer Schering Pharma AG. Miillerstrafie 178, 13342 Berlin, Germany (email: Christoph.Gerlinger@ bayerhealthcare.com).
When comparing diferent active treatments, a noninfen’on’ty-also called one-sided equivalence-study design is used. This study design requires the definition of a noninfen’ority margin, the threshdd value of clinical relevance. At present, a noninfen’ority margin of 10 percentage points is conventionally used for the change in acne lesion counts, but it lacks empirical validation. We analyzed the data of 4,081 patients with moderate to severe facial acne. The treatment efect was recorded by the investigatoras the rel-
I N T R O D U CTI 0 N Acne is one of the most common skin diseases (1). The majority of the patients are young individuals in adolescence or early adulthood. Acne is characterized by two major types of lesions: noninflammatory and inflammatory. The count of acne lesions is considered an objective measure for the effect of acne treatment in clinical studies, for example, by the FDA (2). When comparing the efficacy of different active treatments in the reduction of acne lesions, a noninferiority-also called one-sided equivalence-study design is often used. The methodological principles for the noninferiority trial design are described in the 1CH guideline E l 0 (3). The design of a noninferiority study requires the definition of a noninferiority margin, the threshold value of clinical relevance, in the clinical study protocol. There are statistical and clinical aspects to be considered when choosing a noninferiority margin (4). The major clinical requirement for choosing the noninferiority margin is that any treatment difference smaller than the noninferiority margin should not be of clinical relevance. The major statistical requirement for choosing a noninferiority margin is that the noninferiority margin is small enough
ative change in lesion counts fiom baseline (objective assessment). At the end of the treatment perial, patients rated themshes as having their acne condition improvedl unchanged, or w m ened (subjectiveassessment). We compared the changes in lesion counts with the patients’ self assessment to derive an empirically validated noninfm’oritymargin. We found that an empirically validated noninfen’ority margin of 10-Ls percentage points for facial acne lesion counts is appropriate.
to exclude the effect of placebo. The focus of this article is to empirically define the
Data Loading...
