HOME QUICK SEARCH AUTHORS KEYWORDS HELP BUGS

PharmaSUG 2002 Papers - Index

85 papers. Source: PharmaSUG (Pharmaceutical Software Users Group) conference proceedings.

Back to Conf
go 2003
go 2001
# Title
×		 Authors
×		 Abstract
×		 Keywords
×		 Extras Pages
1 %Addval: A SAS Macro Which Completes the Cartesian Product of Dataset Observations for All Values of a Selected Set of Variables Rich Schiefelbein It is often useful in clinical programming to create additional observations to add to an existing dataset as "placeholders." These additional observations then complete the "Cartesian product" of the combinations of all values for a selected set of variables. For example, the output dataset from the PROC FREQ in a SAS /g210 analysis of count data does not include observations for combinations of variables that are not present in the input data, but may still be of interest (i.e. the result dataset does not include an observation with a count value of "0" for "Caucasian/Females" if that particular combination is not present in the input data). This macro is a very flexible tool which allows one to quickly and easily create additional "combination" observations in a dataset in=, out=, proc freq, macro variable, proc sql, macro call, proc step, sample 6
2 %Qexcel Tackles the Challenges of Converting Data to SAS from Excel Jay Zhou One common programming task in the pharmaceutical industry is to transfer external data such as lab, ECG, PK, and other non-case- report-form data to SAS from EXCEL. Spreadsheets can be read with the SAS import wizard or procedures such as excel, dataset, proc step, label statement, macro function, macro variable, infile statement, data step, macro definition, input statement, proc import, macro loop, audit trail, filename statement, macro parameter, variable, call symput, informat statement, sas log, cpu time, proc printto, macro quoting function, %if-%then-%else 6
3 (Web) Software Development: Best Practices for Developing Enterprise Applications Gregory S. Barnes Nelson data validation rules, proc sql 8
4 A Comparison of XML and SAS Xport Files for Electronic Submission Tianshu Li The eXtensible Markup Language (XML) is the next generation of ASCII text file. It was developed to describe, transfer, deliver and share data across different platforms and applications. As the use of XML gains momentum, it may be accepted by the FDA as a data format for electronic submissions. The SAS transport file has worked as an FDA archival standard for several years but it has its drawbacks. For example, although SAS transport files are portable data sets, only a limited number of other technologies can process data to or/from SAS data sets or SAS transport files. An experimental XML engine was therefore created in SAS 8.0 to import and export XML data files. This paper will introduce the concepts of XML and SAS Xport files; compare XML and SAS data structures and assess the pros and cons of these two data structures for electronic submissions txt, ascii 4
5 A data dictionary. A more complete alternative to Proc contents. Angelina Cecilia Casas A data dictionary is an alternative to the output from PROC CONTENTS for the description of one or more datasets in one or more data libraries. A data dictionary can be a more complete description of a dataset because it can include information from one or more format catalogs associated with a dataset. This paper gives an example of how to present information of one or more datasets in two libraries proc step, proc sql, data dictionary, proc contents, proc format, proc report, proc print, create table, macro variable, adverse events 4
6 A Data-Driven Macro Automating the Data Presentation Process by Generating Tailored, Customizable SAS Code: Relax, let %TABGEN do your work! Generating Tailored In many clinical trials it is useful or even required, perhaps for quality control purposes, to generate a selected data printout of desired variables in any/all datasets for a given directory. This paper presents a simple-to-use yet feature-laden macro designed to lighten this time-consuming task. The macro, %TABGEN, generates for each dataset an independent SAS program containing a tailored, customizable PROC REPORT. %TABGEN gleans dataset attributes using PROC SQL and SCL functions to generate a detailed DEFINE statement with appropriate width, format, and label information for each variable. %TABGEN allows sub-setting, the global dropping or keeping of variables, and the specification of ID and order status as well as line and page break variables. Several header, report type, and directory specification options are also available, and empty datasets are processed to show dataset structure proc report, proc step, macro variable, proc sql, data step, macro call, %if-%then-%else 6
7 A SAS Macro for Producing Data Summary Tables Yajie Wang,
Lan Zhao,
Surai Thaneemit-Chen		 The Department of Veterans Affairs (VA) Cooperative Studies Program (CSP) conducts large scale, definitive research in multi-site clinical trials on health issues important to the nation's veterans. To provide oversight committees and study investigators with the information they need to monitor study progress, on a regular basis CSP staff produce summaries of patient intake, patient characteristics and outcomes, and data quality. Since only the oversight committees may see results broken down by treatment group, we produce two sets of data summaries. Typically, the formats of the tables and graphs in the two sets are similar as both include summary statistics for categorical and continuous variables, and differ only in how the data are broken down (by site in one set and by treatment group in the other). This paper presents a SAS macro module to generate data summary reporting tables macro variable, macro definition, data step, macro loop 6
8 A Web Based Clinical Data Browser Eric Grelak,
Chester Springs		 A problem many pharmaceutical companies have is making its clinical trial (CT) information easily accessible to authorized non- programmers. This situation can be greatly compounded when information is scattered over multiple platforms and legacy systems. The authors describe an n-tier application, which allows the user to access and analyze CT information. The front-end tier is a web browser providing users with a single interface to clinical trail information. Access to information is controlled such that each user receives a different list of available information depending on security privileges. The back-end tiers decide how available information should be accessed, and then retrieves the requested information using a variety of applications (ASP, SAS/INTRNET, Perl, etc.). In addition, programmers can publish their own SAS programs/web applications to the user community adverse events, clinical trial 6
9 Access to Relational Databases Using SAS Frederick E. Pratter proc sql, proc step, dataset, libname engine, sql pass-through, sql server, excel 4
10 Adding a TIFF Preview to Encapsulated PostScript Graphics within the SAS System Gerhardt Pohl There is increasing use within the pharmaceutical industry of PDF (Portable Document Format) for clinical trial reports and other regulatory documents. The natural graphics format for figures to be included in such documents is PostScript. Unfortunately, in work environments where Microsoft Word is the primary authoring tool, PostScript graphics can be problematic. PostScript graphics are not visible in Microsoft Word documents prior to printing or rendering as a PDF. To overcome this difficulty, a visible preview of the graphic can be inserted into the Encapsulated PostScript (EPS) file prior to insertion into MS Word. A technique will be demonstrated using the DATA STEP to bind SAS/GRAPH EPS and TIFF files into an EPS file with TIFF preview. When inserted into MS Word the TIFF portion of the file is displayed while the EPS portion passes without translation or distortion into the final PDF proc step, pharmaceutical industry 4
11 An Automation System for Generating Case Report Form Tabulations David C. Izard,
Shi-Tao Yeh		 This paper describes an automation system for producing a Case Report Form Tabulation (CRT) listing. For the purposes of this paper, the authors define a CRT listing as a clinical report that interleaves a series of individual data listings (e.g., demographic characteristics, laboratory data, and vitals) by patient. The system approach used in this paper applies an efficient data integration and patient listing reporting process while providing a simple yet powerful programming interface to produce CRT listings. The SAS product used in this paper is SAS BASE on a UNIX platform crf, macro definition, proc step, proc report, dataset, sample, macro variable, concomitant medications 6
12 Bayesian Analysis of Population Bioequivalence Using the Independence Chain Algorithm in PROC MIXED Chain Algorithm The statistical test for population bioequivalence given in the current FDA guidance document on this matter ignores the dependence between summary statistics. It is very conservative in cases where there is a high degree of correlation of subject responses between test and reference formulations, which is usually the case in bioequivalence studies. Adapting the ideas of Kass and Wolfinger (2000), we use the PRIOR statement in PROC MIXED to approximate the distribution of the population bioequivalence parameter Θ PBE, from which we can construct tests based either on the probability of the upper tail region (i.e., Pr(ΘPBE>Θ0)), or on the one-sided upper 95% confidence bound of ΘPBE,. We show that this test can conclude population bioequivalence when the FDA method does not. We also examine the statistical properties of this approach. The Kass and Wolfinger approach is also compared to another alternative given by McNally et al which is also more powerful than the FDA sample, proc mixed, confidence interval, proc step, normal distribution, proc means, covariance, restricted maximum likelihood, likelihood function, null hypothesis, posterior probability, prior probability, correlation 6
13 Building and Using Macro Libraries Arthur L. Carpenter While many users take advantage of the SAS Macro Language, few take full advantage of its capabilities to build and maintain libraries of macros. Management of large numbers of macros can be problematic. Names and locations must be remembered, changes must be managed, duplication and variations of individual macros must be monitored, and efficiency issues must be taken into consideration. These problems come even more to the forefront when macros are shared among programmers - especially in a networked environment. Macro libraries allow you to control and manage the proliferation of your macros. Libraries can consist of a formal use of the %INCLUDE statement, a group of programs defining macros that can be automatically called, or even collections of macros that were previously compiled and then stored. Macro libraries are extremely useful, and not really all that complicated. If you are not currently using them as part of your macro programming efforts, read on! macro definition, stored macro, autocall macro, options statement 8
14 Bulletproofing and Knowledge Encapsulation in Statistical Macros John K. Troxell Most SAS programmers agree that SAS macros are a wonderful tool for automating repetitive tasks. More controversial is bulletproofing (making a macro robust to incorrect parameters or data). Less commonly considered is the potential that macros offer for delivery of statistical expertise. Statistical knowledge can be built into macros with advanced rules for validating macro parameter values, implementation of good statistical practice, implementation of otherwise unavailable statistical methods, or delivery of data-driven advice and interpretation in English. Pros and cons of bulletproofing and knowledge encapsulation are discussed, and some specific techniques are presented. Many of the techniques are useful in non-statistical macros as well. Examples are drawn from macros for stepwise regression analysis and analysis of covariance regression analysis, macro put, proc step, ancova, predictor variable, proc glm, covariance, macro loop, anova, %if-%then-%else, proc genmod, autocall macro 4
15 Choice of Development Tool for the User Interface of a Client-Server Application in a SAS Environment Barry R. Cohen Application developers in SAS environments regularly face the question about what tool to use to build the client-based user interface of their client-server applications. These are environments where: (1) the data is in server-based SAS data sets, (2) the primary processing is done with server-based SAS Software applications that cover file management, analysis, and reporting, and (3) the user interface, or front-end, is client-based and is used primarily to direct and control the processing and to present results. Although the server side is all SAS-based, the client side user interface can be, but does not have to be, written with SAS Software. The choice of a front-end tool in this environment has never been simple because of trade- offs involved sas base, dataset, logistic distribution 6
16 Clinical Data Model and FDA Submissions Shy Kumar Development of data model in clinical trial management environment is an ongoing process but of utmost importance for a good clinical information management. Companies are trying to come up with efficient global standards in data modeling to work with. Global standards are also important and imminent in view of the new electronic submission guidelines of Food and Drug Administration (FDA) for New Drug Application. Most important factors to consider when designing the analysis and reporting data structure are standard data structures, metadata, documentation, and data normalization. The main objective of this paper is to provide a data model that will make FDA submission more efficient in new drug approval process. The paper discusses specific areas such as developing data model, features of data model, defining source data structures, defining derived variables, and also discusses the strengths and weakness of this modeling approach normalization, snowflake schema 6
17 Clinical Trial Online - Running SAS on the Web without SAS/IntrNet Quan Ren During clinical trial, it is very important for the project management to have the most recent updated clinical trial information. The best solution is dynamic access to clinical trial data: dynamic data management and dynamic data analysis. SAS System almost became a standard programming language in clinical data analysis. With the development in Internet, people have been exploring every technical possibility to use SAS System to manage and analyze clinical trial data dynamically through Internet. SAS/INTRNET software has provided an option to use SAS System through Internet. However, there is an additional cost. This paper will present another way to use SAS System through Internet without SAS/INTRNET. An experimental example - Clinical Trial Online will be presented. Through Clinical Trial Online , the following technique will be demonstrated: 1. Remote data entry, data browsing and data editing through Internet. 2. Running SAS programs remotely and bringing the adverse events, clinical trial, proc means, data science, survival analysis, anova, vital signs 8
18 Convert ASCII Output Files to SAS Data Sets Lara E.H. Guttadauro This paper describes two macros used in converting ASCII output files to SAS data sets. The first macro (READASCI) reads the ASCII output into SAS and creates a data set. The second macro (UNWRAPTX) takes information in one column that is displayed on multiple rows and 'unwraps' that information so that it is displayed on one row proc print, macro loop, %if-%then-%else 4
19 Create Directory on Windows Without the Fleeting DOS Window Hsiwei Yu On newer Windows environments, the pipe feature on a FILENAME statement can be used to create directory without the fleeting DOS window. This is better than the X command approach. Any DOS commands can be used in the FILENAME pipe feature. Finally we will show example of managing directories on remote machine from local SAS session libname statement, macro put 2
20 Data Definition Tables - Definition and Automation Matthew J. Becker One of the elements of a successful FDA electronic submission is the data definition table (DDT). Three types of data definition tables are needed for a submission: 1) a list of all datasets and what type of information each contains (eg: Adverse Events, Demographics, etc.); 2) for each dataset in the database, a list of all variables, with variable type, variable format, and, if the variable is derived, algorithm; and 3) an alphabetical list of all the variables in the database, along with the dataset in which each resides. Using macros originally developed by Iza Peszek and Frank Liu, we have taken a step towards automating DDT generation for a submission. We will discuss the components and process used to arrive at the ultimate goal: automated DDTs! dataset, create table, proc sql, excel, label statement, proc step, order by 4
21 Data Dependent Footnotes using PROC REPORT Angelina Cecilia Casas Sometimes it can more convenient to create tables and listings using PROC REPORT, instead of DATA _NULL_, even though there is greater flexibility with DATA _NULL_ for titles and footers. This paper gives an idea of how to manipulate data so the easiness of PROC REPORT can be used while keeping some of the flexibility of DATA _NULL_ proc report, proc step, numeric variable, proc sql, adverse events, macro variable, proc sort 4
22 Data Exploration in the Pharmaceutical Industry Andrew T. Fagan The SAS P21 Data Explorer (P21DE) is a tool based on, but not restricted to, the emerging CDISC (SDS) guidelines. It enables users to browse patient population data in a clinical context. Unlike most generic data visualization tools, the P21DE does not require the user to understand the structure of the underlying data. Rather, the user can directly analyze demographic, safety, and efficacy data with regard to specific clinical measures. For instance, using the P21DE users can analyze safety data by selecting specific laboratory measures and viewing how these values change from baseline or between visits or compared to lab normal ranges. With traditional data browsers, the user would first have to create the appropriate data set with the analysis information. This usually means merging several data sets and generating the comparative data - tasks that can only be accomplished if the user has knowledge of the data structure adverse events, raw data, baseline, clinical trial, proc means, clinical data, dataset, cdisc 8
23 Date and Time Values in Clinical Trial Data George Li Date and time values are widely used in clinical trial data for computing duration, cumulative time, and for windowing issues. Partial or incomplete dates in the raw data and the wide variation of formats across countries are issues that arise when working with such data. This paper discusses ways to overcome these problems using SAS functions, formats, and other features in processing, analyzing and presenting date and time values missing value, clinical trial, adverse events, clinical data 4
24 Design Considerations Used to Develop Clinical Reporting Software Steven Light,
Paul M. Gilbert		 To streamline the production of SAS-generated reports used in regulatory submissions, many companies are purchasing commercial SAS-based clinical reporting software or developing custom SAS-based software. This SAS-based software usually consists of several layers of SAS Macros. This paper discusses the design of the CR Toolkit , commercially available SAS-based clinical reporting software. The discussion will focus on the overall architecture of the software and specific design aspects of the report modulesIntroduction regulatory submission, proc freq 4
25 Discriminant Analysis (PROC DISCRIM) to Separate a Study Population by Treatment Subgroups in a Clinical Trial with a New Antidepressant Lev Sverdlov,
Park Ridge		 Data were evaluated from a pilot phase 2 clinical trial (N=55) with a new antidepressant and multivariate discriminant analysis (PROC DISCRIM) was used to separate the drug-treated from placebo populations by treatment subgroups. The list of variables for discriminant analysis included the percent of change from baseline for the primary efficacy variable (21-item Hamilton Depression Rating Scale - HAMD-21) for 11 major time points from Day 3 (the first day of evaluation after treatment) to Day 40 (the last day of evaluation at the end of the trial). Separation with discriminant analysis was very effective for any combination of two of three treatment subgroups (from 82.4% to 86.7% of subjects with the correct classification by treatment subgroups) and for the three treatment subgroups together (80% of subjects with correct classification) discriminant analysis, change from baseline, clinical trial, proc discrim, sample, proc step 6
26 Documenting the Database: A SAS Macro to Automate the Process William C. Murphy Creating the database for your client was straightforward, but now you have to list all of the data sets in the database, along with variable lists and sample contents. You could just run a PROC CONTENTS on all members of the library, but the client wants the variable list interleaved with sample printouts from each data set. You can easily accomplish this by alternating the PROC CONTENTS with PROC PRINT, along with appropriate TITLE statements for each data set in the library. However, to do this every time you must report on a database could be an onerous programming task. But by using ODS, PROC DATASETS, and a %DO loop, you can create a macro to handle your variable and data set documentation needs for any database proc step, macro variable, proc contents, proc print, macro loop, do loop, sample 4
27 Don't copy this program: use SYSPARM instead Steven Wong Often a production program needs to be slightly modified to run in an optional way. That program is copied and the copy is modified per the new specifications. Now both programs reside in production and one or the other is run depending on which option the user needs. One obvious drawback to this approach is when the common portion of the programs needs to be fixed or modified, both files require the same modifications. However there is a technique to run the one program with the modifications/options as an input parameter. SYSPARM is a base SAS system option, which allows the user to pass information to the program from the operating system. SYSPARM is available for all operating systems, the syntax is slightly different for each operating system. For this paper I will be using the UNIX syntax. This paper will be directed to the beginner/moderate skill level, basic knowledge of the macro language will be required macro variable, if-then-else, options statement, %if-%then-%else, macro definition 4
28 Dynamic Letter Assignment in Efficacy Tables in Clinical Trial Research Zaizai Lu The paper introduces a macro we created to assign letters to different treatment groups. Treatments that are not significantly different from each other will be assigned the same letter, while treatments that are significantly different from each other will get different letters. The order of letters assigned will be based on the order of the parameters of interest. The macro program can automatically determine how many treatment groups there are in the analysis, and then automatically assign letters to different treatment groups, based on user-defined order of certain parameters, the overall p values for the treatment effect and pairwise comparison p- values among different treatment groups. This macro has great flexibility and automation, and is easy to use p-value, macro variable, macro definition 6
29 Efficient Interpolation Calculation Julia Z. Zhang Often times, programmers and statisticians encounter missing values in data points while constructing data sets for statistical analysis. Depending on the type of analysis being performed, it may be necessary to impute these missing values. Linear interpolation i.e., replacing the missing values with results derived from non-missing, adjacent values, is an example of imputation. Generally, linear interpolation implemented in a DATA STEP has involved use of LAG function, which returns adjacent values stored previously in the lag queue. There is, however, no corresponding SAS function to return subsequent, non-missing, adjacent, values from observations not already read/SET into the step. Retrieving the necessary non-missing, adjacent values can be accomplished using a combination of the OPEN, FETCHOBS, VARNUM, GETVARN, GETVARC, and CLOSE functions in Base SAS data step, dataset, missing value, imputation, sas base, cpu time, set statement, proc step, proc sort, proc sql 4
30 Expanding the Functionality of Min()/Max(): Finding Tied Minimum or Maximum Values and Their Contributing Variables Kent Nassen,
Subrahmanyam Pilli,
Luai R. Alzoubi		 SAS provides functions to find the minimum or maximum value within a variable or across a set of variables. But what happens when you need to determine, across several variables in a record, which variable is the minimum or, if there are ties at the minimum value, which variables had the tied value? Three SAS macros are presented and compared, which go beyond the functionality of the MIN() or MAX() functions by determining the minimum or maximum value or tied values, as well as indicating the source variable(s) that account for the minimum or maximum value(s). Methods demonstrated are macro loops, arrays, and PROC SQL. These macros were originally developed to produce a derived data set for a clinical Positron Emission Tomography (PET) cardiovascular study where comparisons across regions of the heart were needed and tied minimums or maximums in the data would be handled differently, depending on the region(s) with the minimum or maximum value(s) proc sql, proc step, do loop, macro loop, macro definition, memory usage, %if-%then-%else, data step, cpu time, proc print, macro variable 8
31 Fast Two-Sample Permutation Tests, Even When One Sample is Large, that Efficiently Maximize Power Under Crude Monte Carlo Sampling John D. Opdyke I present a method for quickly performing multiple nonparametric two-sample permutation tests on continuous data in SAS, even when one sample is large. I maximize statistical power (within the context of a crude Monte Carlo approach) by "oversampling" - drawing more permutation samples than desired, deleting duplicates, and then selecting the desired number of samples from the remainder. I determine the optimal number of samples to "oversample" based on sampling probability and the runtime of a sampling procedure (PROC PLAN). Implementing "oversampling" with nearly optimal numbers of samples increases start-to-finish runtime typically by only 5%, and always by less than 10%. Using telecommunications performance measurement data from multiple sources with a wide range of sample size pairs, I benchmark start-to-finish runtime against a) another SAS procedure (PROC MULTTEST), b) another SAS program written for the same purpose, and c) Cytel's PROC TWOSAMPL, with very favorable results sample, proc step, permutation test, sample size, power analysis, monte carlo, confidence interval, in=, out=, macro definition, %if-%then-%else 10
32 Have a Strange DATE? Create your own INFORMAT to Deal with Her Venky Chakravarthy Whatever the title may lead you to believe, this is a serious discussion of dates that are strange to the SAS Software. You will read about date values that cannot be readily produced with the supplied date FORMATS and some that are not recognized by the supplied date INFORMATS. Enhancements to the date FORMATS enable you to customize dates in several ways. While these enhancements are welcomed, there are some limitations when it comes to inputting date values. Date INFORMATS, which are the tools for converting to SAS date values, are lagging behind in comparison. This paper deals with date values that are out of the range of current SAS date FORMATS/INFORMATS and suggests ways to output/input such values. The bulk of the discussion, however, focuses on how to INPUT such date values and how to create your own INFORMAT to do so date variable, sysdate9, proc format, proc step, data step 6
33 How are we doing? The importance of metrics and standard practices in clinical data management and biostatistics. Kent Letourneau Assessing how well you are doing on a given process can be described as a 3-step process. First you must determine which attributes of your process are most important. Second you must come up with quantitative measurements that will numerically describe these attributes. Finally you must define some standard practices that can be used for comparison to determine if your process is achieving the attributes that you desire. This paper will go through this 3-step process and attempt to define which are the important items, metrics and standard practice for clinical data management and biostatistics database lock, elapsed time 6
34 How to fill up the TABLE with all the categories when the actual DATA has only a few: A beginner's TUTORIAL - via the DATA STEP. Venky Chakravarthy This paper addresses one of the common problems encountered by a beginner in producing a report. The requirements and specifications dictate the presence or absence of categories in a table. However, the data does not always cooperate. After summarizing we may find that the final data set does not have all the required categories. This paper emphasizes the following four easy steps to ensure that the table meets the requirements: (1) Create a template data set out of the requirements. (2) Extract the trial data and summarize if needed. (3) Reshape the final data, if needed, to get it to the format of the template created above. (4) Update the template with the final data set. This gives a definite structure and keeps the process simple for a junior programmer. Only the creation of a template data set may be a little challenging. With some experience this can be easily overcome proc step, proc print 4
35 Implementing a Standardized Data Model - A Real Life Experience Curtis Wolf,
Paul M. Gilbert		 As organizations strive for gains in efficiencies in processing and using clinical information, the need for a common set of terminology and data standards becomes vitally important. As partnering and use of outside organizations become more common, this need takes on an even broader challenge. One such company was the setting for a project we will be discussing in this presentation. The intent of the project was to facilitate the sharing of data across multiple platforms and systems thereby enhancing and creating efficiencies in the data management, review and reporting process and enabling sharing of resources and unique developments across the organization. This model is being used as the basis for the designing care report forms (CRFs), clinical databases and analysis files and consequently for Case Report Tabulations (CRTs) for submission to the FDA and other regulatory agencies, as well as the data source for a variety of other reporting tools clinical data, data standard, clinical research 4
36 Imputation of Longitudinal Incomplete Data in Clinical Trials Naum M. Khutoryansky A common problem in longitudinal clinical trials is missing data caused by the withdrawal of some patients from the study before completion (that is, drop-outs). Drop-outs, in general, yield unbalanced data with unequal numbers of measurements for each patients. Some drop-outs appear completely at random. In other cases the drop-out process depends on the response variable of interest. Fo r the latter situation, ignoring missing data may lead to biased comparison of the treatment effects. This paper compares four different techniques of estimating the mean and standard deviation of the response variable at the end of study for some longitudinal clinical trials with large percentage of dropouts. The techniques being considered include the last observation carried forward missing value, clinical trial, last observation carried forward, imputation, standard deviation, longitudinal data, sample, repeated measures, regression analysis, proc step 4
37 IND Annual Reporting at a Glance Kristi Wiser An Investigational New Drug (IND) Application is a request for authorization from the Food and Drug Administration (FDA) to administer an investigational drug or biological product to humans. During a new drug's early development, the sponsor's primary goal is to determine if the product is reasonably safe for initial use in humans. An IND Annual Report is required to be submitted in yearly intervals within 60 days of the anniversary date of the IND effective date. This report is required annually until the final Clinical Study Report for studies filed under the IND has been submitted to the FDA or the IND has been withdrawn. Some of the IND Annual Report required information includes a brief summary status of each study in progress or completed during the previous year and the safety of the drug as defined by adverse event tables and listings. A more complete description of the process for developing an IND Annual Report is found in the poster adverse events, csr, clinical trial, clinical research, proc freq 6
38 Introduction to SAS/GRAPH Philip A. Mason sample, bar chart, proc step 8
39 It's Easy If You Know How: Importing, Processing, and Exporting CDISC XML with SAS Michael C. Palmer The Clinical Data Interchange Standards Consortium (CDISC) has published an XML-based clinical data standard that will make it easier for a pharmaceutical company or other study sponsor to develop a single, low maintenance, vendor-neutral gateway for clinical trials data, including central lab results and FDA-compliant data archiving. The tutorial will cover the basics of data-centric XML, the structural and semantic requirements of CDISC's Operational Data Model (ODM) standard, and techniques for importing, processing, and exporting clinical data in SAS using the ODM. The tutorial is needed because SAS users accustomed to flat or relational files have found it challenging to import, process, and export CDISC XML. Trouble has come from two characteristics of XML. First, it is a text stream, and, second, it is hierarchical in structure. SAS, on the other hand, is oriented towards data in fields and flat or relational file structures clinical data, adverse events, cdisc 4
40 Java Application Development for Analyzing Clinical Pharmacogenomics Data Zongwei Shen,
Robert Tracy,
Richard Judson		 Genaissance Pharmaceuticals has developed its proprietary HAP/Ge4 technology for establishing the complicated relationships between genomic patterns present in the human population and clinical data. The DecoGen /Ge4 Platform, a sophisticated Java- based software available at Genaissance, has demonstrated its power to accomplish analyses in which huge genomic analysis data are involved. With the recent progress in the development of Java-based products such as AppDev Studio /Ge4, the computing power of SAS system can be remotely accessed via Java clients. This paper describes major issues associated with Java-based software integrated seamlessly with SAS so that detailed statistical analyses and related data manipulation can be pushed to the SAS server. Topics and examples presented will include how pharmacogenomics analysis data and PROCs can be dynamically issued from Java clients to the SAS server dataset, proc step, clinical data, response variable, predictor variable, data step, proc glm, sample, proc genmod 4
41 JMP Density Ellipses for Pharmaceutical Analysis Charles E. Shipp JMP is a data discovery tool used in both (1) the exploratory phase and (2) the analysis phase, of experimental and theoretical work. The density ellipse shows the extent of data, the center of mass, the least squares fit line, and the degree of correlation between a selected dependent variable and a selected independent variable. When superimposed, the ellipses of similar variables for different classes show and contrast their sameness and differences. The example used is pharmaceutical dummy data in the area of cancer research predictor variable, response variable, correlation, class variable, proc sgplot 6
42 Let ODS in SAS V8 Help to Speed Up Your Electronic Clinical Study Report (e-CSR) Jim J. Wang,
Sherry L. Pinnell,
Tikiri B. Karunasundera		 An electronic clinical study report ( e-CSR) contains in-text tables within the body of the report, and also post-text results (tables, figures, listings) as its supplements. It is a common practice to create in-text table shells in MS- Word, and then keypunch in the numbers based upon the post-text results, which is time-consuming and vulnerable to human errors. Prior to the birth of SAS Output Delivery System (ODS), all the post-text results had to be in the ASCII format, which led to another tedious task of converting ASCII files to Adobe Portable Document Format (PDF) files. This paper will present how ODS may help to speed up the generation of your e-CSR by delivering PDF and/or Rich Text Format (RTF) post-text ods rtf, csr, output delivery system, clinical trial, ods pdf, file statement, proc step 4
43 Macro Functions: How to Make Them - How to Use Them Arthur L. Carpenter The SAS macro language includes a number of macro functions, such as %UPCASE, %QSCAN, and %SUBSTR. In addition a number of Autocall macros, such as %LEFT, %TRIM, and %VERIFY, are also supplied that act like macro functions. Macro functions are especially useful as programming tools, because the function call is replaced directly by the result of the call. Many macro programmers are unaware that they too can write useful macro functions. Fortunately macro functions are not difficult to program. Not difficult, that is, if you know the simple techniques required to turn a macro into a macro function. This paper shows you how to create a macro function, the statements that you must avoid, and the technique used to "pass" the macro function, macro variable, macro definition, %if-%then-%else, macro call, macro facility 6
44 Macros for Model Building, Summarizing, and Displaying Results of Stepwise Logistic Regressions Lori S. Parsons When performing stepwise logistic regressions with PROC LOGISTIC, analysts often want to summarize the results of the model without including the full selection procedure. Beginning with SAS 7.0, the Output Delivery System (ODS) provides users with a method to print a subset of the default output. There is additional information about the stepwise model that should be included in a summary, but cannot be obtained directly from ODS code. Graphical representations of the model are also helpful in a logistic distribution, proc step, predictor variable, logistic regression, proc logistic, regression analysis, macro variable, odds ratio, confidence interval, proc contents, output delivery system, response variable, missing value, dataset, out=, stepwise selection, ods output 6
45 MacStat: A Statistical Analysis and Reporting System using SAS Macros, Visual Basic, and Microsoft Word Richard Thomas Two of the many challenges that exist for statisticians at AstraZeneca is one, generate statistical analysis without reinventing the wheel and two, how to convert text-based SAS output into Microsoft Word tables (and eliminate the manual typing process). This paper details the MacStat system was designed to solve these problems data science, confidence interval, sample, response variable, proc step 6
46 Making SAS Output Pretty - Creating Customized Documents for your Clients Elaine Dempsey SAS has long been valued for its strength in statistical computing. Indeed, statistical analysis is the basis for the creation of SAS over 25 years ago. Today, in the clinical research industry, SAS is almost always used in conjunction with a separate word processing package in order to present output in a richly formatted manner that will be pleasing to the eye of the client. However, SAS itself can produce the same aesthetically pleasing output by utilizing the DATA STEP Graphics Interface. The examples shown in this paper are based on SAS version 6.12 under Windows 98, but the principles presented apply to all operating systems and subsequent releases of SAS data step, proc print 4
47 Managing SAS Services: A System Administrator's Guide to Enterprise Applications Gregory S. Barnes Nelson Enterprise software development activities usually involve robust management and monitoring tools that enable systems to run seamlessly and often without intervention. Over the past 25 years, SAS has evolved into a true enterprise-class suite of solutions. As companies turn to SAS software for solving critical business needs, those that manage these applications require mature tools and techniques to help ensure the stability of these environments. In this paper, we will discuss SAS from a systems administration perspective. That is, how can we manage the environment in which SAS operates. Here, we will discuss various methods for managing database, compute and web services in both a Microsoft Windows NT environment as well as UNIX environments. We will focus on how to start and stop services, techniques for monitoring systems and methodologies for how to do this in multi-tier environments 2
48 ODS to RTF: Tips and Tricks Paul D. Hamilton Since the introduction of the ODS RTF destination, traditional character-based displays of tables and listings are passé. The new capabilities also bring new tribulations as the pharmaceutical community strives to produce high quality statistical reports, maximizing the amount of information on a page, while still conforming to FDA guidance. This paper examines some obscure ODS to RTF formatting tricks available by getting creative with SAS version 8.2. Topics covered include selective bolding and graying of data values, insertion of blank separator rows, selective graying of report rows, management of widow and orphan data lines, control of table cell widths and heights, font control, and selecting and displaying special characters (footnote notations, scientific and mathematical symbols, typography characters) proc report, proc step, character variable, numeric variable, adverse events, proc template, proc sort, macro loop, %if-%then-%else, macro definition 10
49 Optimizing System Performance by Monitoring UNIX Server with SAS Sam Mao,
Jay Zhou		 To optimize system performance and maximize productivity, UNIX performance and resource usage should be monitored. The monitoring is often done manually by sporadic checking using commands such as 'top', 'ps'. A utility has been developed using SAS that monitors and identifies extremely resource-consuming processes, and sends e-mail to notify the owner of the processes. The utility is automatically invoked by another SAS macro in Windows NT. With the regular monitoring in appropriate frequency, performance issues can be detected early, ensuring system efficiency and productivity. This paper presents a number of techniques in interacting SAS with UNIX while going through the major steps of monitoring and reporting of the utility. The techniques include reading from different UNIX command outputs, processing data from UNIX command outputs, taking advantage of information available from UNIX, and sending data dependent e-mail using UNIX mail facility through SAS DATA STEP dataset, macro variable, data step 6
50 Performing Logistic Regression on Survey Data with the New SURVEYLOGISTIC Procedure Anthony B. An Categorical outcomes such as binary, ordinal, and nominal responses occur often in survey research. Logistic regression investigates the relationship be- tween such categorical response variables and a set of explanatory variables. The LOGISTIC procedure can be used to perform a logistic analysis for data from a random sample. However, this approach is not valid if the data come from other sample designs, such as complex survey designs with stratification, clustering, and/or unequal weighting. In these cases, specialized techniques must be applied in order to produce the appropriate estimates and standard er- rors. The SURVEYLOGISTIC procedure, experimental in sample, proc step, proc surveylogistic, logistic distribution, logistic regression, regression analysis, proc logistic, odds ratio, predictor variable, response variable, likelihood function, random sampling, maximum likelihood estimation, main effect, proc surveyreg, sample weight 10
51 Pitfalls and Successes of Automating In-Text Tables for a Clinical Study Report: A Case Study Matthew J. Becker In-text tables within clinical study reports are a vital component for sponsors, but a headache for medical writers. In the past, perhaps presently, medical writers continue to transcribe numbers from tables, listings, and graphs produced with SAS . Novel approaches have been developed and presented to streamline this process. At PharmaNet, rich text format (RTF) output is generated in a Word table so that the medical writer can cut-and-paste the data required into an in-text table. This paper and presentation will outline the process used, first attempt, medical writing needs, and subsequent attempts to provide the best solution csr, clinical trial, sas output, ods rtf, adverse events, macro definition 6
52 Post-processing MPRINT Outputs to Generate Macro-Free Code Jim Edgington Clinical trial programming involves extensive use of SAS macros to facilitate the process and reduce coding time. SAS macro programming creates difficulties for debugging, maintaining, and reviewing the code due to the complex nature of the language, inherent with multifaceted macro routines. Therefore, regulatory authorities prefer to review straightforward and macro-free programs. SAS provides the MPRINT and MFILE (for V7 or above, RESERVEDB1 for V6.12) options with the MPRINT fileref to create an ASCII file only containing the de-macroized source code for the purpose of debugging. But this must be done on an individual file basis with results in a rudimentary format. This paper presents a SAS macro, %Qdecode, that post-processes MPRINT results to create easy-to-read, executable SAS programs with a standard indentation pattern. The macro can process every file in a given directory to produce a macro-free program proc step, proc format, mfile, macro definition, format statement, macro facility, macro variable, dataset, macro call, proc sql 6
53 Preparing the SAS Software Programming Environment for Regulatory Submission Sunil Kumar Gupta For pharmaceutical companies, there is a push to prepare SAS programs for clinical study reporting and analysis. By utilizing a clinical data warehouse design strategy, SAS programs can be set up in advance of the study completion to facilitate the generation of the final study analysis, tables, listings, and graphs. Issues in data warehouse structure and technology and industry standards will be discussed. This paper reviews the process of establishing a standard method to prepare the SAS software programming environment proc step, clinical trial, clinical data, data cleaning, data science, sas programmer, regulatory submission, data warehouse, proc sql, quality assurance, quality control, raw data, view, proc tabulate, macro definition, adverse events, dataset, sample, proc freq 10
54 PROC REPORT and ODS Cindy Stroupe As the need for electronic submissions grows, the need for an easy way to create FDA- acceptable electronic files grows as well. With the release of Version 8, SAS provides a tool, the Output Delivery System (ODS), to convert from traditional SAS output to other file formats. PROC REPORT and ODS together give a user with little ODS knowledge the ability to create RTF, HTML, or PDF files with ease. This paper will give a brief overview of the REPORT procedure and some ODS style options that can be used within PROC REPORT to customize the electronic output proc report, proc step, output delivery system, ods listing, ods rtf 4
55 Punctuation in SAS Programming John Morrill It's the little things that get you! Punctuation is often relegated to less than even a second thought, only to be considered when an error exists. It is our contention that a systematic review of punctuation in SAS programming yields for even experienced users a number of interesting tidbits and valuable efficiencies. In this paper we first briefly identify the common uses of several type of punctuation. Then from the less obvious uses we found during our search, we present the cases most likely to increase efficiency or provide new functionality. Examples of coverage include the when and why of single versus double punctuation marks (e.g., the semi-colon, the period, the question mark, the exclamation mark, the @ sign, and the pipe) including a discussion of macro processing, a review of punctuation usage differences between SQL and traditional SAS code, and some interesting notes on the underscore character proc step, proc sql, put statement, macro variable, pointer control, macro call, data step, proc report, automatic variable, sas programming, macro quoting function, character variable 6
56 Rapid Assessment of Data Set Structures Philip Vecchione When SAS data is received from an external source, such as a sponsor, CRO, or another department, it is important to assess the new data to determine if variables have been added or removed, locate specific variables, and identify missing data in key datasets. Changes in dataset structure can have an impact on existing analysis programs, requiring modifications or creation of new programs, which uses up valuable analysis time. In order to speed up this process and create an efficient way to assess new SAS datasets, a family of macros has been designed to quickly obtain this information. These macros work for individual datasets as well as for a library of datasets. They rapidly provide output that identifies potential problems, allowing the user to communicate them back to their external source quickly; therefore streamlining the process time from data receipt to final analysis proc contents, proc step, macro call, study day, label statement, data science, dictionary tables, sample, missing value 4
57 Recovering Lost Lines: Using the Whole Page with Proc Report Richard R. Allen Since the FDA went to the 12pt font requirement for submissions, it is more important than ever to have complete control of th e total page for designing tables. When using PROC REPORT to desgn these tables SAS automatically outputs a line after all titles and before all footnotes. A simple macro will be presented to remove these blank lines when extra lines are needed to fit a report on a page relative risk, macro call, proc report, proc phreg, hazard ratio, group by, proc step 6
58 Reduce Paper Output When Using SAS PROC COMPARE (A Validator's Dream Come True!) Lila Thome Whether you are validating a data transfer or the data from a table or listing, using SAS PROC COMPARE can lead to organizational nightmares for the validator due to the volume of output. This paper will present a macro that produces a condensed version of the PROC COMPARE output. The program below utilizes SASHELP.VTABLE to compare entire libraries, not just data sets proc step, proc compare, if-then-else, proc sort, macro variable 4
59 Report Macro - A Tool to Generate Flexible Summary Reports Diane E. Foose When creating output, you can not always anticipate the number of columns needed in your summary tabulation or that each page of the report may vary on the number of columns used. This may cause you to program many variations of the same style of report for each slightly different output. This paper will present the concept of using a SAS macro as an easy-to-use solution to this problem. Considerations for how to use this strategy for reporting your data will be discussed. An implementation of this concept, with examples, will also be presented proc report, sample 6
60 SAS Cluster Disk Management: Using SAS to Manage Virtual Disks on a Cluster of File Servers Haftan M. Eckholdt,
Ezra Benun		 Using SAS to manage low cost virtual drives on clusters of file servers is very simple and enables researchers to approach larger more complex problems. We will discuss methods for using SAS for: (1) managing very large virtual drives in file server clusters, (2) random and symmetric file naming, file seeking, file mapping, and management on these virtual drives, (3) visual and quantitative assessments of cluster usage and file distribution, and (4) provide crude versions of RAID backup redundancy. These methods allow research groups to acquire very large storage capacity at extremely low costs dataset, macro loop, proc step, macro definition 6
61 SAS Cluster Gene Matching: Using SAS To Manage Computer Clusters And To Perform Protein Matching Haftan M. Eckholdt In latest efforts to understand genes, it is critical to find proteins of similar sequence, structure, and function both within and between species. This process is extremely CPU and disk intensive and requires a complex approach to produce useable liquidity gap, parallel processing 6
62 SAS Coexistence with FDA 21 CFR Part 11, How Far Can We Get? Gustav Sporon-Fiedler,
Marie Lassen		 FDA 21 CFR part 11 (Ref[3]) presents a number of requirements that challenges the standard usage of SAS in the scientific communities within pharmaceutics. This paper examines SAS techniques and configurations which may help the SAS user community get closer to compliance requirements. The standard consolidated SAS environment is non compliant with FDA 21 CFR part 11. A number of add on facilities and non SAS products are needed to achieve compliance regarding such items as security and version control. However SAS does supply excellent audit trail and log facilities that should be fully exploited, when attempting to comply with the requirements as set forth by FDA 21 CFR Part 11 21 cfr part 11, dataset, audit trail, data cleaning, proc step, data step, proc print, access control 6
63 SAS Macros for Graphical Data Analysis Doug E. Lassman SAS/GRAPH has many useful procedures to aid in graphical data analysis including PROC GPLOT, PROC GCHART and PROC G3D. Combining these with PROC GREPLAY allows for even more powerful displays of data. Sometimes the exact display is not available with the previously mentioned procedures. This is where PROC GANNO and ANNOTATE datasets come to the rescue. Virtually any display imaginable can be created with a combination of dots, lines, bars, polygons and text. After a brief proc step, proc gchart, proc gplot 8
64 SAS Solutions for Addressing 21 CFR Part 11 Compliance - the P21 Biomedical Knowledge Platform Edward D. Helton The heart of 21 CFR Part 11 is less than three pages long - but the impact on regulated industries has been enormous. SAS' P21 Biomedical Knowledge Platform offering has been designed to specifically address the issues associated with 21 CFR Part 11 and the FDA's Guidance for Industry documents. P21 provides these capabilities while offering an enhanced operating environment for managing clinical data, programs, logs, documents and reports. Careful consideration has been given to the intended performance with respect to data warehousing, analysis and reporting, electronic submissions and related e- Signature requirements. The development of P21 adheres to the industry standards for both process and quality management. Application of both process and quality management has assured that the software meets the intended requirements of the system's 21 CFR Part 11 functionality 21 cfr part 11, data cleaning, regression analysis 6
65 SAS Techniques to Get the Most out of Oracle Clinical David Johndrow Many research companies are short on Oracle talent but long on SAS expertise. How can such organizations further benefit from SAS experience within the clinical setting? This presentation has 3 main goals: 1. Expose an alternative approach to the extraction of clinical data into SAS (with a larger role played by SAS programmers). This approach can also potentially have a significant impact on cost. 2 . E x p l o r eh o wS A Sc a nb ee m p l o y e dt ot a pi n t ot h e underpinning tables of the Oracle Clinical system, and how this available treasure trove of data can be used within the SAS setting to generate project reports and track study progress. 3. Expand on the ability to create SAS reports by demonstrating some of the newer output formats available in SAS, via the Output Delivery System (i.e. RTF, HTML). A few simple SAS tricks can turn reports into interactive tools character variable, libname statement 8
66 Save Time Today Using SAS Views Robert S. Ray In this time of exploding data volumes, your old friend the SAS/GE2 data view can be more important than ever as a tool to reduce total job I/O and, thereby, total time. Views can be created in SAS with either SQL or the DATA STEP and used singly or in combination. SAS data views can be used as one might use Unix 'pipes' to chain multiple data manipulation stages together without intervening I/O. Judicious use of this established technology can cut execution real-time significantly and bring your critical jobs back into their time windows. Additionally, this paper will briefly discuss the new "pipes" feature of the multi-process (MP) SAS/CONNECT /Ge2 product proc sql, proc step, order by, cpu time, set statement, mp connect, dataset 4
67 Serology Inventory System Mariya Charny In this paper we present an application used to track serology samples for clinical trials. The system used SAS to interact with a Clinical data management system (Clintrial) and an Oracle database through the use of Base SAS, SAS Macro, SAS/AF, Open Data Base Connectivity (ODBC) and SAS Output Delivery System (ODS). Serology samples are collected from subjects at various time points during the conduct of clinical trials designed to evaluate investigational vaccines and immuneglobulins. Historically the serology inventory system was implemented using Microsoft Access. This required double entry of data in the Clintrial and MS Access databases, and required constant checking to assure they were identical. Furthermore, the Access database lacked an audit trail. Users who needed to access the systems were completely dependent on the programmers to write the necessary queries clinical database, sample, clinical trial 8
68 Simulating Attrition and Visual Models of Type I and Type II Errors Haftan M. Eckholdt Longitudinal studies are plagued by attrition. While statisticians can address these problems, researchers have trouble understanding the impact of attrition. Methods described here are designed to help scientists understand and appreciate the impact of attrition on their hypotheses. Starting with the essential hypotheses, the approach begins by characterizing the sample at each stage of research to date: population of origin, recruitment, baseline, first follow up, etc. [BASE & STAT: PROC FREQ, PROC TABULATE, etc,]. The next step executes 1000's of simulations of attrition through the random as well as systematic deletion of cases [%MACRO]. In the next step, models are run [STAT: PROC GLM, PROC MIXED, etc.] and relevant coefficients are saved [ODS]. Finally, distributions of coefficients (observed, expected, random deletion, and systematic deletion) are compared [GRAPH: PROC GPLOT] sample, proc step, type ii error, cross-sectional data, macro definition 10
69 Some Analysis Methods for Behavioral Toxicology Data Richard R. Allen Five groups of monkeys were tested on a battery of four behavioral tasks to determine if being exposed to a particular drug had any effect on their learning capabilities. They were tested on two pairs of tasks every other weekday for a period of a year and a half. This talk will concentrate on the longitudinal analysis of the accuracy of response for the Conditioned Position Responding Task (CPR), a color and position discrimination task. The data and results of different but similar analyses of the data are discussed. Some of the results are presented in a web report built using the ODS system with drill-down graphics from Version 8 of the SAS System. Generalized linear modeling techniques were used to determine if there were any differences between the groups proc step, regression analysis 8
70 Streamlining SAS Data Listings to Match CRF Data Flow for Use in Limited Internal Audits Michael Acra,
Greg Henkle,
Glaxo Smith Kline		 At Glaxo Smith Kline we use a proprietary electronic data capture (EDC) tool to database clinical study data. Before authorizing a database, an internal audit is performed to compare the eCRF data printout from Microsoft Access tables with the processed data that is extracted from the EDC output into SAS data sets. The converted data sets are uploaded to the VAX for further processing, data checking, and printing. The data set print utility is written using SAS 6.08 on the VAX and is comprised of three macros that work in tandem to produce the printed output in the desired data sequence for a 10% random list of subjects to be sampled. The eCRF order of variables and data sets repeats in the printed output for each subject. This allows for a direct comparison between the original EDC tables and the SAS data sets found on the VAX macro variable, dataset, cdisc 2
71 Table Lookups: From IF-THEN to Key-Indexing Arthur L. Carpenter One of the more commonly needed operations within SAS programming is to determine the value of one variable based on the value of another. A series of techniques and tools have evolved over the years to make the matching of these values go faster, smoother, and easier. A majority of these techniques require operations such as sorting, searching, and comparing. As it turns out, these types of techniques are some of the more computationally intensive, and consequently an understanding of the operations involved and a careful selection of the specific technique can often save the user a substantial amount of computing resources. This presentation will demonstrate, compare, and contrast the following techniques: IF-THEN, IF-THEN-ELSE, SELECT, user defined formats (PUT function), MERGE, SQL joins, merging with two SET statements, using the KEY= option, use of ARRAYS, and Key-Indexing proc step, proc print, proc sql 8
72 Techniques for SAS Enabling Microsoft Office in a Cross-Platform Environment Vincent DelGobbo With the ubiquity of the Microsoft Office suite of products, you often need to use SAS data and analytical results in the Office products EXCEL or Word. But transferring data from SAS to the Office applications can be difficult when the SAS System is installed on a non-Windows platform, such as OpenVMS, UNIX or OS/3790. This paper provides techniques for integrating the analytic processing power of the SAS system, irrespective of platform, with Microsoft EXCEL and Word. The use of the Output Delivery System (ODS) is also discussed excel, sample, ods html, ods rtf, output delivery system, proc step, ods csv 8
73 Techniques for the Successful Management of Laboratory Data for Clinical Research and Development Benjamin H. Frierson During the course of a clinical trial, the results of laboratory tests remain some of the most important indicators of drug safety. It is therefore imperative that a programmer utilizing SAS to m anage these data for safety evaluations command the knowledge of the diverse sources, compilation, formatting, and validation necessary to their reporting in a timely and accurate fashion. For a given protocol, laboratory data may originate from a lab local to a study site or number of sites, from a lab centrally located for processing from a site or number of sites, or from a combination of both. These local or central labs may additionally have assorted means of data transfer, from regularly scheduled shipments of text files or spreadsheets that may be incremental or cumulative, to continuous delivery of lab reports that may be entered into a database proc step, proc sql, proc sort 6
74 The 800-Pound Gorilla: Taking on the Challenge of Globally Standardizing Adverse Event and Medication Reporting Deborah Harper,
Jennifer Tjernagel		 When Pfizer merged with Warner-Lambert/Parke-Davis in 2000, work began to harmonize various processes across all sites. One of these efforts sought to standardize the collection and reporting of clinical data. The principle guiding the process was first to determine what needed to be reported, and then work backwards toward data collection. The initial step was to agree on a set of standard summary and listing tables for each data type defined in the Clinical Data Interchange Standards Consortium (CDISC). This presentation will discuss how we started, the process used, decisions made, resulting tables, and the SAS algorithms necessary to implement the standards for adverse events (AEs) and medications. Issues we addressed include common coding dictionaries, levels of reporting, definition of treatment, imputing dates, and the definition of treatment emergent signs and symptoms (TESS). Our presentation will focus on tables, and algorithms, which aid in the development of SAS code adverse events, clinical data, trial elements, high-level term, clinical trial, concomitant medications, cdisc 4
75 The CDISC Laboratory Data Interchange Standard: Using SAS To Read and Write CDISC Compliant Data Sets and Files Margaret Hung The Clinical Data Interchange Standards Consortium (CDISC) has released for public review and comment its data model for clinical trial laboratory data interchange. The version 1 release of the laboratory data interchange model defines standards for laboratory data, data standard, clinical trial, flat file, character variable, dataset, sample, clinical data, cdisc 8
76 To Err Is Human; To Debug Devine Roger Staum This tutorial describes four types of programming errors and how you can recognize them. It also discusses four approaches to handling errors: deter, detect, deflect, and debug missing value, proc step, error message, proc print, put statement, macro variable, input statement, numeric variable, proc sql, options statement, merge statement, by group processing, character variable, error, sas log, macro facility 12
77 Two Step Approach to Automatically Generate ESTIMATE Statements for Pair-wise Treatment Comparisons Eric Qi,
Guanghan (Frank) Liu		 To do pair-wise treatment comparisons in a clinical study with repeated measures, we often use ESTIMATE statements in the PROC MIXED procedure. However, writing ESTIMATE statements sometimes may be challenging, and that is especially true when some weights are considered for stratification factors in the analysis model. This paper explains how to use a two step proc mixed, proc step, change from baseline, macro loop, clinical trial 4
78 Using SAS Macros to Construct Items 5 and 11 of an Electronic Submission Elizabeth Dennis The Food and Drug Administration (FDA) has set forth guidelines to which companies must adhere when submitting New Drug Applications (NDA) electronically. These guidelines are changing the way companies view the preparation of an NDA, and they should change the way statistical programming teams view their tasks. The team should no longer view their job as merely producing tables and listings, but in producing output that facilitates an electronic submission. Many of the requirements of Items 5 and 11 of an Electronic Submission can be produced using SAS. To meet these requirements, three SAS macros have been developed. These macros can be used individually to generate specific components of the submission. They may also be combined with pre- programming specification documents to further ease the construction of output suitable for Items 5 and 11. Key Words: FDA, electronic submission, SAS macro, item 11, transport file, data definition table statistical programmer, macro definition 4
79 Using SAS ODS to Enhance Clinical Data Summaries: Meeting eSub Guidelines Steven Light,
Paul M. Gilbert		 SAS programmers in the pharmaceuticals industry are aware that the Food and Drug Administration (FDA) has raised the bar for presentation of summary tables and listings used as part of electronic submissions (eSub). Guidelines for font selection, font size, and margins have made traditional text file output somewhat more challenging to work with. Fortunately for us, the new guidelines have coincided with the SAS Institute's implementation of the Output Delivery System (ODS), which provides the ability to both greatly enhance the appearance of tabular output and to meet the FDA's eSub guidelines. This presentation will provide an introduction to use of ODS in generating tabular reports using ODS (to RTF and PDF destinations) with a focus on the eSub guidelines and the advantages of using ODS over listing-mode output pharmaceutical industry, sas programmer 6
80 Using SAS to get more out of Oracle Clinical Jim Johnson Oracle Clinical creates SAS data sets or SAS data views as part of its regular function. This paper will describe ways to get more out of Oracle Clinical than it "wants" to give you. Oracle Clinical provides several types of data (Value Text, Full Value Text, DVG long value, DVG short value, DVG number, and Exception value text just to name a few). Each of these data types serves a specific purpose, and most will at some point become necessary for processing, reporting, and delivering the data to your user or client. Unfortunately, Oracle Clinical only allows a certain number of bytes for each record when generating the SAS data sets or views, and that width is not large enough to recover all the types of data available for data sets with a lot of variables. Only one 'active' extract is available at a time, so you need to know how to get the most out of it. Using SAS, you can use one extract get to all of these data types with little cooperation from the Oracle Clinical product proc step, data cleaning, crf, dataset, proc compare, length statement, view, numeric variable, proc sql, data validation rules 10
81 Using Styles and Templates to Customize SAS ODS Output Sunil Kumar Gupta SAS's new Output Delivery System (ODS) feature enables the creation of various new file types including Rich Text Format (RTF), PostScript and HTML. The standard default style used to control the output is generally well organized and sufficient for most purposes. When the default SAS ODS output report will no longer suffice, SAS users can take advantage of ODS advanced features to customize the output report with styles and templates. By having greater control over the report details such as color, font, size, justification, order and labels, the format and quality of the report will be enhanced. This paper introduces the concepts necessary to understand and apply the advanced features of SAS's ODS. Issues in defining and selecting output destinations, selecting output objects and creating customized output files will be discussed. The focus is on the creation of custom styles and templates with PROC TEMPLATE for visually appealing output output delivery system, proc step 10
82 Various Techniques for Processing CDISC ODM XML Christopher R. Olinger This paper describes various techniques for processing the CDISC Operational Data Model (ODM) using SAS and Java. The CDISC ODM group's stated mission is to "develop a vendor independent standard model for the interchange and archiving of clinical trials data with study metadata." However, this means little if you cannot get the data out of it and process it effectively using your favorite tools (aka SAS). This paper will deal with the issues and techniques surrounding processing ODM XML for use with SAS clinical data, dataset, libname engine, cdisc 34
83 Web-Based Code Generation for Derived Data Using SAS/Intrnet Luai R. Alzoubi,
Kent Nassen,
Subrahmanyam Pilli		 Do you want to know how to generate SAS code using SAS/INTRNET? This paper describes a web-accessible SAS/INTRNET-based application that generates dynamic SAS code that can be run to create derived data sets with computed variables. The system accesses Oracle Clinical SAS views and features include user selection of the clinical study, data sets and variables to use. The user must specify analysis week definitions, which weeks are included in the baseline period and which week will be used to carry forward the last observation. The output includes a SAS command file with the customized code to create two SAS data sets for clinical efficacy data and ODS HTML output to display the code in the web browser dataset, clinical trial, view, proof of concept, proc step, ods html, proc sql 10
84 What's New in the Output Delivery System, Version 9.0 Sandy McNeill Along with every new version of SAS, the ODS developers work hard to bring you new features. So what's new with Version 9? A new destination called MARKUP along with something called markup definitions (aka TAGSETS) will become production. This new destination allows configurability of your markup output. Another new destination (also available as a PROC) that will be production for Version 9 is the DOCUMENT. If you've ever wanted to the ability to run your procedures once and then go back and pick and choose which pieces you want or how you would like them rendered, then this is for you. In addition to these new destinations, RTF has some new features and functionality -- options for Page X of Y, decimal alignment, and cell indention are a few of the highlighted features proc step, ods rtf, output delivery system, proc template, footnote statement, proc print, excel, sas output 10
85 You CAN Save Your Log and View It, Too: An Improved Process for Automatically Saving the Contents of the Log and Output Windows Amy Rhoads Saving an electronic copy of the program log and output when producing analyses of clinical data is a critical step. However, the two usual methods of saving the log and output in PC-SAS version 6.12 have drawbacks. Specifically, if we use PROC PRINTTO to automatically send the log and output to files, we lose the advantage of viewing the color-coded log in the log window. Also, it is more difficult to view and print the output. On the other hand, when manually saving the files, it can be time-consuming to navigate to the desired directory and enter the FILENAME. Additionally, there is the risk of saving items under the wrong FILENAME or type or forgetting to save them at all. We have developed a macro which can be used to automatically save the log and output to files, yet still display the contents in the SAS windows macro variable, sas output, sas log, proc printto, proc step 6

Back to Conf go 2003 go 2001

This site serves as a convenient hosting location and does not work as a publisher. All copyrights belong to their respective owners.

The information on this site is provided for general reference purposes only. Whilst every effort is made to ensure that the site is up to date and accurate, no warrant, nor any responsibility or liability for, the accuracy or completeness of the content or for any loss which may arise from reliance on information contained in this site is provided. Links to other sites are provided for your convenience but no responsibility or liability for the content of those sites or of any external site which links to this site is accepted. References to any specific commercial product, process, service, or corporation are also for information only, and do not constitute endorsement, recommendation or favoring.

SAS is a registered trademark of SAS Institute Inc. SAS, Statistical Analysis System, and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.