ABSTRACT:

The two-treatment, two-period, two-treatment-sequence (2,2,2) crossover design is utilized frequently in drug development programs. Historically when completing such analyses, statisticians produce their own statistical analyses, tables, listings, and graphics by modifying either their own software or software which is on loan from other persons. This process is iterative and can produce redundant efforts both within and among individuals; errors in software revisions; delays in validation; and inconsistencies in the presentation of study results.

Recognizing the opportunity to improve this process, a SAS macro was developed to statistically analyze and efficiently display the 2,2,2 crossover trial. The macro produces report ready summary tables and listings in MS Word, simple and diagnostic graphics, and a text file for additional graphing and analyses. These outputs require no manual intervention. The SAS/BASE, SAS/MACRO, SAS/STAT, and SAS/GRAPH modules, as well as the rich text formatting language, were utilized.

BIOGRAPHY:
Amy Morris is a Senior Statistical Programmer at Merck Research Labs. A SAS user for 6 years, Amy has worked in both the Pharmaceutical and Insurance industries and also teaches a SAS/Computational Statistics course at Ursinus College. She has an MS in Applied Statistics from Villanova University and a BS in Quantitative Business Analysis from Penn State University.

Tom Bradstreet is a Senior Research Statistician at Merck Research Labs where he has been a consulting statistician on human and animal studies of new drugs for 18 years. He earned Ph.D. and MS degrees in Statistics; an MS in Biostatistics; and a BS in Biology. Tom's current interests include permutation tests, distributions of order k, the Behrens-Fisher problem, dose proportionality, statistical education, graphics, and increasing efficiency through statistical computing.

ABSTRACT:
This presentation will discuss two free products available on UNIX and PC environments called SCCS (Source Code Control System) and RCS (Revision Control System). When used in a multiuser environment, these systems give programmers a tool to enforce change control, create versions of programs, document changes to programs, create backups during reporting efforts, and automatically update vital information directly within a SAS program. The systems helps create an audit trail for programs as required by the FDA for drug trial. The major topics of the discussion will be: (1) Why Use These Systems? (2) The History of SCCS and RCS and FDA Requirements, (3) The Flow of a SAS Program Through a Version Control System, (4) How to use SCCS and RCS, (5) Automatic Updating of SAS Program Headers, (6) The Advantages and Disadvantages of Each System, and (7) Implementing a Version Control System.

BIOGRAPHY:
The authors Terek Peterson and Max Cherny are both clinical SAS programming consultants with Alliance Consulting Group currently working at SmithKline Beecham. The authors have worked at both CROs and drug companies, on reporting efforts for all phases of drug development, in varying theraputic areas, and with multiple operating systems. The systems being purposed are in response to FDA requirements for programming documentation and version control within a multiple user environment.

ABSTRACT:
Most SAS Software users working with large data sets often wonder how they can: a) make the data set "smaller," without eliminating necessary variables/observations or b) reduce the amount of time the SAS System needs to extract subsets of observations from the data set. This tutorial explains how data set compression works in the SAS System and the circumstances under which you might want to consider using it, as well as important pitfalls to avoid when considering applying this technique. The second section of the presentation discusses indexing and how it is implemented in SAS System software. Several "before and after" examples of both data set compression and indexing are presented, along with performance metrics showing when these techniques "worked" and when they did not.