Purpose of the Cincinnati Chapter of the ASA

The purpose of our organization is to encourage the use and appreciation of statistics and promote its proper application in all fields of study. We also strive to facilitate interaction among all persons in the greater Cincinnati area who have an interest in or are concerned with the field of statistics.

Upcoming Events

Fall 2007 Meeting - Tuesday, November 27, 3:30pm at Procter & Gamble's Mason Business Center (formerly HCRC)

Our fall 2007 meeting will be held at the Procter & Gamble’s Mason Business Center (formerly HCRC) which is located in Mason, OH. The meeting will be held in the Harvard Room on Tuesday, Nov 27 at 3:30 pm. The seminar will be presented by Dr. Dan Schnell from P&G on "Evalution of Covariate Adaptive Randomization by Minimization in a Clinical Trial. Abstract of the talk follows.

Abstract

The RDP-58 trial was a clinical trial to assess the efficacy and safety of a new treatment for ulcerative colitis. In this trial, patients were randomized to either a high dose of an active control (AC) or RDP-58 with a low dose of AC. Five factors were identified as important patient-level baseline variables for stratification of the randomization. Two additional factors were also used: clinical site (CS) and overall study balance. Various minimization algorithms to achieve balance across the 7 factors were explored by simulation. The final algorithm used a weighted squared difference objective function to determine the preferred treatment assignment as each patient was enrolled. Simulations suggested that the using minimization procedure would achieve good balance on all factors. Additional simulation findings suggested that linear mixed effects models with CS as a random effect and treatment and other stratification factors as fixed effects would preserve type I error probability for the test of treatment difference and provide unbiased estimates of treatment differences. A total of 156 and 138 patients were randomized (separately) in Phases A and B, at 52 and 46 study sites respectively. The properties of other possible randomization procedures applied to the actual sequence of patients in this study will be evaluated and compared to the actual randomization outcome. Dan Schnell obtained his PhD in Statistics from Iowa State University in 1987. He spent seven years as a Mathematical Statistician at US Centers for Disease Control in the Sexually Transmitted Disease/HIV group. In 1994 he joined P&G Pharmaceuticals where he has worked in cardiac, gastro-intestinal and musculoskeletal areas, specializing in Phase I and II clinical development. He has publications in a variety of medical and statistical journals coming out of the work done in clinical trials. Currently, Dan is Section Head, Pharmaceutical Statistics at P&G.

Recent Past Events

Fall 2004 Meeting on Thursday November 18, 2004 Statistical Analysis for Microarray Data by Arnold J. Stromberg, Ph.D. Director of Data Analysis, Microarray Core Facility, Department of Statistics University of Kentucky Summer Meeting on Thursday, June 3, 2004: Adjusting Sample Size for Repeated Measures ANOVA With an Internal Pilot by Dr. Keith E. Muller, University of North Carolina, Chapel Hill

Winter Meeting on Thursday, February 19, 2004: Future Trends in Statistical Science Robert Mason, President of the American Statistical Association

Short Course on Wednesday, March 31, 2004: An Introduction to Logistic Regression by Stanley Lemeshow, Ph.D., Director, Center for Biostatistics, Ohio State University