Statistical Comparisons of Two Methods for Particle Size Distribution of Orally Inhaled Products

• Activity Number: 79
• Type: Contributed
• Date/Time: Sunday, July 29, 2012 : 4:00 PM to 5:50 PM
• Sponsor: Quality and Productivity Section
• Author(s): Monisha Dey*+ and J. David Christopher and Ziqing Pan
• Companies: Merck Research Laboratories and Merck Research Laboratories and Consulting Statistician dDimension LLC
• Keywords: orally inhaled products (OIPs) ; cascade impactor ; aerodynamic particle size distribution (APSD)

Abstract

The FDA currently uses a grouped-stage approach for detecting changes in the aerodynamic particle size distribution (APSD) of orally inhaled products (OIPs) as measured by multi-stage impactor testing which represents different portions of the APSD profile. An alternative approach using two metrics, the ratio and the sum of the large particle mass (LPM) and the small particle mass (SPM) of the mass-weighted APSD, is being proposed by an industry working group. In this talk, the alternative approach will be compared to the current grouped-stage approach via multivariate modeling of the APSD profiles, realistically mimicking the non-normal mass distribution exhibited by many OIPs. Various aspects of the choice for the cut-point used to define LPM and SPM will be discussed. A comparison of the performance of these two approaches in terms of Bias, Type I errors and Type II errors will be done.