Further Reading
Chang, M. N, A. L. Gould, and S. M. Snapinn. 1995. P-values for group sequential testing. Biometrika 82(3):650–654.
Chang, M. N, H. A. Guess, and J. H. Heyse. 1994. Reduction in burden of illness: A new efficacy measure for prevention trials. Statistics in Medicine 13:1807–1814.
Heiberger, R. M., and B. Holland. 2004. Statistical analysis and data display: An intermediate course with examples in S-Plus, R, and SAS. New York: Springer-Verlag.
Heiberger, R. M., and E. Neuwirth. 2009. R through Excel: A spreadsheet interface for statistics, data analysis, and graphics. New York: Springer-Verlag.
Heitjan, D. F., and D. Sharma. 1997. Modeling repeated-series longitudinal data. Statistics in Medicine 16(4):347–355.
Heitjan, D. F. 1999A. Causal inference in a clinical trial: A comparative example. Contemporary Clinical Trials 20(4):309–318.
Heitjan D. F. 1999B. Ignorability and bias in clinical trials. Statistics in Medicine 18:2421–2434.
Amit, O., R. M. Heiberger, and P. W. Lane. 2008. Graphical approaches to the analysis of safety data from clinical trials. Pharmaceutical Statistics 7(1):20–35.
Oxman, M. N., M. J. Levin, G. R. Johnson, et al. 2005. A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults. The New England Journal of Medicine 352(22):2271–2284.
Shih, W. J., H. Quan, and M. N. Chang. 1994. Estimation of the mean when data contain non-ignorable missing values from a random effects model. Statistics and Probability Letters 19(3):249–257.
Stine, R. A., and J. F. Heyse. 2001. Non-parametric estimates of overlap. Statistics in Medicine 20(2):215–236.
Heitjan, currently at the University of Pennsylvania, was another active scholar. He taught courses on longitudinal analysis with missing data and provided consultation on longitudinal trials and missing data. His primary collaboration was on a program involved with analyzing intra-ocular pressure data from glaucoma clinical trials using repeated-series longitudinal models, which led to a 1997 publication in Statistics in Medicine. He also worked on causal models with noncompliance and published papers about his work in Controlled Clinical Trials and Statistics in Medicine.
Stine’s (University of Pennsylvania) experiences as a Schor scholar were similar, but with different areas of emphasis. He worked on a prediction model of osteoporosis that led to statistical methods for measuring overlap in distributions, which he wrote about in a 2001 Statistics in Medicine article. Stine’s consulting and teaching was primarily in preclinical animal studies of blood pressure in a longitudinal setting, and he used his background in business statistics to help the marketing department with a promotion-response study using a mixture model with excess zeros.
Richard M. Heiberger, a statistics professor at Temple University, took a sabbatical at GSK’s Research Statistics Unit during the 2003–2004 academic year. He participated in a company-wide investigation into the use of graphics in clinical trial reports and submissions, culminating in a paper published in Pharmaceutical Statistics. He also gave seminars and taught two short courses based on his book Statistical Analysis and Data Display: An Intermediate Course with Examples in S-Plus, R, and SAS. Beginning in January 2009, Heiberger entered into a research contract with GSK through Temple University for a one-day-a-week collaboration. He began his current collaboration with a seminar on the RExcel interface, which is the topic of his new book, R Through Excel: A Spreadsheet Interface for Statistics, Data Analysis, and Graphics.
Heiberger has built upon the experience he gained in the design of clinical trials simulation experiments by designing Excel-based user interfaces, which allow clinical staff direct access to the underlying simulation software written in R. There are at least two journal articles in preparation based on this work.
These experiences illustrate that important problems in pharmaceutical statistics have been addressed by developing, applying, and publishing novel statistical methods in actual drug and vaccine research settings. These programs provide academic statisticians experience in solving real-world industrial problems, while allowing industry statisticians to maintain their theoretical foundation. They also can lead to long-term relationships between academia and industry.
Leave a Reply