HOME
QUICK SEARCH
AUTHORS
KEYWORDS
HELP
BUGSPharmaSUG 2018 Papers - Index
230 papers. Source: PharmaSUG (Pharmaceutical Software Users Group) conference proceedings.
| # | Title× | Authors× | Abstract× | Keywords× | Extras | Pages |
|---|---|---|---|---|---|---|
| 1 | "A tale of a tail" or how to make your graph more informative | Iryna Kotenko | The nature of the data undoubtedly determines its visual shape and quite often you have to look for workarounds to make a better visual representation of the data. This article discusses the features of PROC TEMPLATE & PROC SGRENDER for efficient data scaling and gives number of practical examples |
proc step, confidence interval, proc template | source | 12 |
| 2 | A Collection of Items from a Programmers' Notebook | David W. Franklin, Cecilia Mauldin |
Every programmer should have a notebook with notes and pieces of code that are of interest to them. Some will be categorized as personal, notes only of interest to them, but most are of interest to a broader range of SAS programmers. This paper brings together new and updated tips from the notebooks of two SAS programmers with nearly fifty years of combined experience that will be useful to beginner and experienced SAS programmer alike. Topics include advice for using the SAS Editor; combing SAS data sets; reviewing SAS Logs; creating EXCEL files from SAS data sets; creating SAS data sets from EXCEL files; creating editor macros in Enterprise guide; splitting the info from &sysinfo; keep only the character, numeric or printable characters in a strip of text; creating empty data sets; to that little bit of code that calculates the first, middle or end date of month. Bring along your notebook and take away an idea or two |
excel, proc step, dataset, sas log, proc import, proc sql | source | 18 |
| 3 | A Conceptual Strategy and Macro Approach for Partial Date Handling in Data De-Identification | Amy Caison, Kelsey Reppert |
To address the privacy concerns of individual patient health data, regulatory authorities in both the US and Europe have established benchmark requirements for the protection of patient privacy. This is ever more critical now that the European Medicines Agency Policy 0070 is being phased in which will, when fully implemented, require the public disclosure of anonymized patient-level clinical trial data. Data de- identification is the process of anonymizing personal information while maintaining the scientific value of that data. Dates are notable indirect identifiers which, when combined with other potential identifiers, can be used to identify a subject | clinical data, clinical trial, date variable | source | 8 |
| 4 | A Critique of the Use of the Medical Device SDTM Domains in Therapeutic Area User Guides | Carey G. Smoak | The Clinical Data Interchange Standards Consortium (CDISC) encompasses a variety of standards for medical research. These standards were originally developed with drug development in mind. Medical Device standards have been developed for the Submission Data Tabulation (SDTM) model. Seven SDTM domains have been developed for Medical Devices. Therapeutic Area User Guides (TAUGs) have been a recent focus to provide advice, example and explanations for collecting and submitting data for a specific disease. Interestingly, the use of the seven Medical Device SDTM domains are being used more frequently in the TAUGs than in actual device submissions. This paper will focus on the use of these seven Medical Device SDTM domains in the TAUGs. Twenty-two of the twenty-six TAUGs have examples of device data, but there are differences in these examples. Increased participation by device experts in the development of TAUGs would be beneficial to resolving these differences | sdtm domain, sdtmig, cdisc | source | 10 |
| 5 | A CRO's perspective on successful partnering to deliver SDTM/SEND contributions | Helen Owen, Kirsty Payne |
Is standardisation at the submission level driving complexity at the contributing CRO level? As the number of organisations contributing to a data package increase so does the complexity of the delivery. Here a CRO presents the challenges, successes and recommendations for other CR Os and partners | raw data, sample, data governance, adpc, pk parameter, adpp, cdisc | source | 7 |
| 6 | A Framework for Implementing [Conflicting] FDA Guidance | Todd Case | On July 21, 2004 the US Food and Drug Administration (FDA) announced a format, called the Study Data Tabulation Model (SDTM) that sponsors can use to submit data to the agency. Twelve years later (on December 17, 2016) the FDA began enforcing the requirement of standardized electronic data submissions in SDTM format and now, in addition to SDTM, there are multiple sources (and versions) of data Standards which impact data supporting applications to the FDA: the FDA Data Standards Catalog (primary list and source of standards) AND the Study Data Standardization Plan, the SDTM model (Version 1.4), the SDTM Implementation Guide (SDTMIG - Version 3.2), the Analysis Data Model (ADaM) - Version 2.1, the ADaM Implementation Guide (Version 1.1), the FDA Guidance for Industry (April, 2017¹), the Study Data Technical Conformance Guide (October, 2017²) and the Prescription Drug User Fee Act (PADUFA), Version V for fiscal Years 2013-2017 (and VI for fiscal years 2018 - 2022) | data standard, clinical data, clinical trial, cdisc | source | 6 |
| 7 | A Gentle Introduction to R From A SAS Programmer-s Perspective | Saranya Duraisamy, Nate Mockler |
Just as a screwdriver cannot hammer in a nail, no single programming language can sol ve every single problem you will encounter in your programming career. Like a good craftsman, a skilled statistical programmer should know many techniques for analyzing and visualizing complex clinical data. While SAS has impressive analytic capabilities, it has its own limitations when it comes to complex analytics and visuals. That is why it is prudent to have another tool in your toolbox. R - an open source language, has made enormous strides in the past few years in the areas of data science and graphic s. This talk will serve as an introduction of R's analytic capabilities including basic data manipulation techniques and the grammar of graphics that serves as the basis of the popular "ggplot2" capability. This will also be discussed from the perspective of a SAS programmer wanting to learn R as an additional language | sas programmer, proc step, proc sql | source | 10 |
| 8 | A Linux shell script to automatically compare results of statistical analyses | Chen Wang, Shu Zan, Joshua Conrad |
Double programming is a standard validation method used in statistical programming in the pharmaceutical industry. With this method, production and validation programmers independently generate and compare data sets or statistical analyses outputs to ensure their correctness. While the SAS COMPARE procedure is readily used to compare data sets, there are no standard tools available to compare the analytic results in form of tables, figures, and listings (TFL). Generation of a text file in parallel with each TFL PDF file allows us to compare using programming tools such as the diff command on Linux systems. However, when the number of TFLs increases, it is time consuming to manually write the commands and difficult to track the results |
data cleaning, txt, independent qc, sample | source | 9 |
| 9 | A Macro to Create Program Inventory for Analysis Data Reviewer's Guide | Xianhua Zeng | As per Analysis Data Reviewer's Guide (ADRG) released by the PhUSE CSS Working Group in January, 2015, a program inventory that contains all inputs (such as SDTM domains or other analysis datasets) and macros used should be listed in the section Submission of Programs. It can be laborious and time consuming to extract inputs and macros manually. This paper presents a SAS macro ExtDsnMan, which programmatically extracts all input datasets and macros and produces an XLS report. The code is currently hosted in my Github page: https://raw.githubusercontent.com/XianhuaZeng/PharmaSUG/master/2018/extdsnman.sas | in=, adam dataset, cdisc | source | 6 |
| 10 | A Method for Independent Program Validation utilising SAS, R and Python | Aik Hoe Seah | The possibilities of comparing results of independent program validation utilising SAS, R and Python were explored. This was achieved utilising the SASPy package (that provides Python APIs to the SAS system), the commonly used data science packages native to Python (NumPY, SciPY, pandas, Matplotlib, statsmodels, etc.), and Rpy2 (that enables users to call R from Python) | dataset, data cleaning, proc step, data science, statistical software | source | 13 |
| 11 | A One-Sided Fisher's Exact Test: A Tail of Clinical Worsening | William Coar | In many disease areas, therapeutic intervention is designed to prevent clinical worsening of the disease, thereby increasing the quality and length of life for patients with life threatening diseases. While this research hypothesis may be simple to state in non-statistical terms, there are many endpoints that can be used to provide insight to this question. One of the common approaches to address this question is to simply ask if a patient experiences clinical worsening after some fixed exposure to a new therapy. For diseases that rapidly progress, intervention of a new therapy should delay clinical progression for the drug to be deemed efficacious. This suggests that the proportion of patients who experience clinical worsening while on active study drug should be lower than the proportion of patients experiencing clinical worsening in a control group. There are multiple ways to demonstrate this with a statistical test | alternative hypothesis, exact test, fisher exact test, treatment arm, proc freq, odds ratio, sample, proc step, class variable, response variable | source | 8 |
| 12 | A SAS Macro to Create Validation Summary of Dataset Report | Zemin Zeng | This paper will introduce a short SAS macro developed at work to create summary report of dataset validation. The code o f the SAS macro is included in the paper and can be eas ily modified to fit reader 's need. The m acro is warmly welcomed by the p rogramming group at work and has proven to be very useful and efficient | data cleaning, macro definition, adam dataset, clinical trial, proc step, cdisc | source | 6 |
| 13 | A SAS Macro to Generate Information Rich Kaplan-Meier Plots | Chia-Ling (Ally) Wu | The number of patients at risk at each time point in a Kaplan-Meier (KM) plot is critical information needed to interpret the curve. This paper describes a SAS macro which provides a tabular summa ry of data used to generate the KM plot, including the number of subjects at risk, the number of events and total s ubjects, median survival rate and Cox proportional hazard ratio. This macro also supports a custom ized output header, labels for time interval (X-axis) and survival rate (Y-axis), reference line, and tick marks for time interval (X-axis). In addition, the ability to add a filter to the data to generate KM plots by various subsets is provided. The output destination can be PDF, RTF or html files. This macro applies the SAS PROC LIFETEST procedure to delineate Kaplan-Meier survival plots and computes the survival function by log-rank test and Wilcoxon test |
hazard ratio, proc step, proc phreg, median survival, kaplan-meier plot, macro variable, regression analysis, macro definition | source | 11 |
| 14 | A SAS@ Macro Application on Confidence Intervals for Binominal Proportion | Kaijun Zhang, Sheng Zhang |
Confidence Intervals (CI) are very important to present results in clinical trial. One of the most common methods for categorical data is used in situations in which the outcomes of interest can be categorized into two, i.e. either a success or a failure. Binomial proportion can either be the proportion of successes or the proportion of failures. This topic is relevant to both statisticians and programmers because binomial proportion estimation is so commonly applied in a wide variety of situations. SAS procedure which is usually used to obtain these proportions and CIs is the SAS @ procedure PROC FREQ. The common methods for the computation of CI include Wald, Clopper Pearson and Willson score. But sometimes, the direct computation of CI based on mathematical formula may be better choice depending on data source. The right algorithms of CI is regularly chosen by statisticians |
proc freq, proc step, confidence interval, binomial distribution, sample | source | 13 |
| 15 | A Simple Method for Integrating Analysis Result Metadata into Existing Define.xml 2.0 | Jeff Xia | Analysis Result Metadata (ARM) adds significant value to a regulatory submission. It provides agency reviewers with a means to trace results back to their source documents, such as key pieces of SAS programs used for analysis, datasets, data selection criteria, and even specific pages in a statistical analysis plan. However, the define.xml for many completed studies does not have ARM included due to earlier technical constraints. It is of great interest to add ARM features to these existing define.xml if there is a business need to include them in future submissions. The Pinnacle 21 (P21) Community version becomes increasingly popular for generating define.xml 2.0 for submissions by the pharmaceutical industry. The latest version of P21 (version 2.20) still lacks the capability to generate define.xml with the feature of ARM. This paper presents a simple method for integrating ARM in the existing define.xml version 2.0 | excel, data science, cdisc | source | 9 |
| 16 | A Statistician's Guide to Leadership and Success | Heather Murphy | It has long been presumed that statisticians tend to be more introverted and numbers focused. While this may be true, it does not need to be the primary way in which we are described. Statisticians can be both leaders and educators in our fields of work. This paper will describe how to communicate effectively with those who are not statisticians, how to maximize the value of those on your team, and how to reinforce positive behaviors in order to shape a stronger, more cohesive team. Using the skills and behaviors that will be described in this paper will aid in building a more solid team in which you are all equally valued and respected. In order to produce quality results, we need to work together and that begins with an understanding of each other's roles and responsibilities. By spending the extra time in the beginning to foster this environment, we will be better able to grow as a team and prevent issues down the road | adam dataset, data science | source | 5 |
| 17 | A time saving approach to track data issues | Aishhwaryapriya Elamathivadivambigai | Data issues in raw datasets are a menace in statistical programming. The earlier they are identified and fixed, the smoother the programming process. Edit checks performed by the Data Management team often do not evaluate the data for compliance to CDISC standards. For instance, case s like missing values in a required variable or incorrect visit values according to the protocol could be missed by the DM's screening process. Many such issues are detected during the TFL generation process , thus affecting the quality of the outputs, and not to mention the time consumed to program around them. Such data errors that elude DM's initial edit checks need to be detected at an earlier stage to produce accurate and precise outputs. This paper proposes a way to detect such data issues upstream and list them in a n EXCEL document for investigation by the programmer, thus saving a significant amount of time for programmers and allowing production of higher quality outputs |
sdtm domain, adverse events, data validation rules, missing value, adam dataset, date variable, excel, cdisc | source | 8 |
| 18 | Accessing the Metadata from Define-XML | Lex Jansen | Many papers have been written about creating Define-XML documents. This paper is about what we can do when we receive a Define-XML Document. We can view the Define-XML document in a browser by using a stylesheet, but how do we access the metadata in the Define-XML document? Once we have access to the metadata we can use it to drive automation. This paper will discuss various technologies to extract the metadata from a Define-XML document, like SAS XML Mapper, SAS PROC GROOVY, SAS PROC XSLT and the SAS Clinical Standards Toolkit. This paper assumes that the reader is familiar with some basic XML concepts, and Define- XML version 2. An earlier paper by the author (see references Jansen (2012) [1]) contains both a brief overview of the XML needed to understand this paper, and an overview of the structure of a define.xml file based on Define-XML version 1.0.0 |
proc groovy, proc step, dataset, macro variable, libname engine, format statement, character variable, controlled terminology, foreign key, data cleaning, macro definition, cdisc | source | 44 |
| 19 | Accredited, Bona fide, Certified, Diploma'ed, and Edumacated: The ABCDEs of Automating the Validation and Monitoring of Professional Requirements for Employees and Job Candidates Through Dynamic, Data-Driven SAS Reporting | Troy Martin Hughes | Job postings typically have stated requirements such as education, training, certifications, and other criteria. Some requirements must be met before a candidate is hired while others can be fulfilled after employment during a "grace period" of a specified duration. Because many certifications require periodic refresh or renewal, validation of professional requirements often continues after candidates are hired and throughout their careers. While validation and monitoring professional requirements for a small team might be accomplished in minutes, the effort increases with the number of employees, the diversity of employee roles, and the diversity of requirements that must be tracked. This text demonstrates a flexible, scalable, reusable macro (COMPLIANCE) that validates and monitors education, training, and professional certifications for job candidates or employees | data cleaning, data validation rules, sample, excel, xls, proc import, sas base, txt, dataset, data analyst, sas programmer | source | 20 |
| 20 | ADaM data structure layout: common issues and ways to avoid | Jianfei Jiang | From reviewing many studies' ADaM validation reports, which are the outputs from running Pinnacle 21 Enterprise (P21E), some common issues draw my attentions. The major problem was that some variable's values were not extended correctly or sufficiently due to complex study design - a design with multiple treatment periods, the consequence was getting over 10,000 warnings or errors from one data set only. Although the issues could be explained in ADRG, but they still exist, data quality and ADaM compliance were compromised. It is desirable to get the cleanest conformance check reports, and we should bear this idea in mind during ADaM specification development. Here I list a few points for good ADaM layout practice that may get not only cleaner conformance check report, but also better data traceability and less burden on explaining the issues in ADRG | treatment period, error message, cdisc | source | 5 |
| 21 | ADaM Define.xml v2.0 Validation - The Perl Way in SAS | Yuxin (Ellen) Jiang | In the latest version of the Study Data Technical Conformance Guide, the FDA indicated that Define-XML version 2.0 is the preferred version of the Define-XML format, which serves as a metadata guide to the data submitted for a clinical trial. There are various methods that can be used to validate the define.xml, such as Pinnacle 21 Community and SAS Clinical Standard Toolkit. However, these tools do not cover all the checks for validating the contents of the define file. Additional SAS based methods have been proposed which involve a step to get define.xml contents using XMLMap software and some manipulation steps to transform the define file into SAS datasets in a desired format. The paper discusses using Perl Regular Expressions pattern matching functions to directly read the content from Define.xml to SAS macros or SAS datasets without using XMLMap, and then performing validation checks in SAS | data cleaning, dataset, sample, variable list, cdisc | source | 8 |
| 22 | ADaM for Medical Devices: Extending the Current ADaM Structures | Julia Yang, Sandra Minjoe |
The current ADaM structures were designed and developed for common drug and biologic studies, which are focused on the subject (USUBJID). In fact, variable USUBJID is a requirement in all standard ADaM dataset structures. Additionally, the ADSL structure of one-record-per-subject is a requirement for submission of drug and biologic studies. Within medical device studies, subject may or may not be of interest, and USUBJID may not even be collected. Instead, the unique device identifier (SPDEVID) is the primary identifier of interest. This means that the current ADaM structures are usually not a good fit for the analysis needs of medical device studies. This presentation will highlight how ADaM can be extended to work with medical devices data. It focuses on content in an ADaM document, in development at the time of this writing, that is designed specifically to address medical device analysis needs | adam dataset, bds, clinical trial, controlled terminology, adverse events, cdisc | source | 7 |
| 23 | ADaM Integration for Summary of Clinical Safety: The 'Unique Patient' Paradox | Aakar Shah, Tracy Sherman |
There are various ways to structure ADaM data for a Summary of Clinical Safety (SCS) analysis that integrates multiple studies, and the methods chosen depend on a number of factors. One of these is the issue of whether the ADaM data sets can or should be created at the study-level and just stacked together, or if there is a need to create new integration records within each patient. How should you handle patients at the study-level when you also need to integrate patients that were enrolled in multiple studies? That's right: accommodating both types of analysis ('patient experience' and the 'unique patient') is a requirement when structuring analysis data sets to support the TLFs (table, listing and figures) in a | laboratory abnormality, laboratory shift table, adverse events, shift table, data science, cdisc | source | 12 |
| 24 | ADaM Structures for Integration: A Preview | Deborah J. Bauer, Kimberly Minkalis, Wayne Zhong |
Integration and analysis of data across all studies in a submission is a vital part of applications for regulatory approval in the pharma industry. The existing ADaM classes (ADSL, BDS, and OCCDS) already support some simple cases of integration analysis. However, there has been a need for an integration standard that supports the more complex cases. To address this need, the ADaM Integration sub-team is developing the upcoming ADaM Integration standards document. This paper introduces the new IADSL, IBDS, and IOCCDS classes found in this document. IADSL allows for multiple records per subject. IBDS and IOCCDS work effectively with the new IADSL class. This paper also discusses the analysis needs that necessitated the creation of the new classes, and provides examples in the form of usage scenarios, data, and metadata. With them, no future integration will prove too complex | sample, adam dataset, raw data, cdisc | source | 14 |
| 25 | Adding another dimension to oncology graphs: 3-D Waterfall Plots in SAS | Liz Thomas, Mark Woodruff |
Waterfall plots and swimmer's plots are almost ubiquitous for displaying the results of oncology trials. Both present compact patient-level summaries that can be quickly absorbed and offer more detail than a traditional table of best response, but they are typically presented as separate outputs with differing sort orders. This separation makes it difficult to match a patient's change in tumor burden with their duration of therapy. One proposed visualization displays the traditional waterfall plot on the facing X -Z plane and adds a swimmer's plot of duration of therapy on the orthogonal X-Y plane. This 3-D waterfall plot has been most commonly implemented in R, using a variety of 3-D projection libraries. We created a set of macros to dynamically normalize data to a unit cube, transform 3-D coordinates to a customizable 2-D projection, and used these projected coordinates with vector and polygon statements in PROC SGPLOT to construct a 3-D waterfall plot |
waterfall chart, proc sgplot, proc step | source | 10 |
| 26 | ADINTDT and ADTTE for Survival Sweep in Oncology Studies | Christine S Teng, Wenyu Hu |
Many oncology studies have either a primary or a secondary endpoint in Overall Survival; therefore, the survival data is critical to study success. The event of interest is death from any cause, and subjects without any event are censored to the latest follow-up date or the last known alive date. When the Interim-Analysis database lock for a certain data cut-off date is approaching, the sites must enter the most up-to-date data for those subjects who are still alive. Of interest are those subjects who are either (1) on study treatment but not yet have the next visit planned, or (2) off treatment but on follow-up phase and have yet to reach the planned visit, or (3) discontinued from the study but agreed to be (though have not yet been) contacted by phone for survival status. This paper utilizes the oncology ADaM datasets, ADINTDT and ADTTE, to create a list of censored subjects | interim analysis, overall survival, adam dataset, imputation, randomization, survival analysis, cdisc | source | 6 |
| 27 | Advanced Analytics approach to explore CENSUS Data and know the population in order to determine the best opportunities to open the new Hospital | Krystian Matusz | Healthcare industry is always very demanding and opening a new Hospital is absolutely always a big challenge. There is a lot of different people, illnesses and things that you as the decision maker, planner, enabler or even GP hav e to face on. There is also multitude scenarios that you need to consider. Each of them can be critical for making the final, rational decision about the right location of the new hospital that will save peoples life. You need to precisely know how to find the best place and you know that the margin of error in current demanding and difficult market is absolutely minimal. It's about people's life. This paper will present usage of public data to support making the critical business decision on the new Hospit al location. The described framework is intended to help make smart(er) decisions about planning and network development. It is also intended to help better understand the demographics and local population data | proc step, macro definition, proc freq, proc sql, correlation | source | 12 |
| 28 | Advanced Clinical Graphs using Axis Tables | Sanjay N. Matange | A key feature of graphs used for the analysis data or for clinical research is the inclusion of textual data in the graph, usually aligned with the x or y axis. The Axis Table statements available with the SGPLOT procedure make it easy to add such data to the graphs. Axis tables can also be used for creating custom axes and multiple axes. This presentation will describe how to use Axis Tables to create complex graphs |
kaplan-meier plot, forest plot, proc sgplot, proc sgpanel, clinical research | source | 11 |
| 29 | Advanced Visualization using TIBCO Spotfire and SAS | Ajay Gupta | In Pharmaceuticals/CRO industries, you may receive requests from stakeholders for quick access to clinical data to explore the data interactively and to gain a deeper understanding. TIBCO Spotfire s an analytics and business intelligence platform which enables data visualization in an interactive mode. Users can further integrate TIBCO Spotfire with SAS (used for data programming) and create visualizations with powerful functionality e.g. data filters, data flags. These visualizations can help the user to self-review the data in multiple ways and will save a significant amount of time. This paper will demonstrate some advanced visualization from Preclarus Patient Data Dashboard (Preclarus PDD) within PPD created using TIBCO Spotfire and SAS (for the SDTM database) and share our experiences and challenges while creating this visualization | adverse events, bar chart, box plot, clinical trial, laboratory data, scatter plot, class variable, vital signs, data wrangling, cdisc | source | 14 |
| 30 | AGING MAN_AGEMENT or millennial TECH_AGEMENT - road ahead for organizations | Charan Kumar K J | Industrial Revolution, Workplace 4.0 - words sending tremors to our heart and mind. The question - will MAN_AGEMENT be put to rest leading to TECH_AGEMENT. "Techager", a role which might analyze behavioral patterns of machines or employees leading to precision decision making. An insight into the analyses of employee mindset would be innovative task in hand and challenging too. Machines replacing humans in certain tasks would pose a different challenge. Avalanche of advances in Information Technology/Electronics - IoT, Artificial Intelligence leading to smartness in devices, applications coupled with developments in biotechnology - CRISPR technology and more; all these would change the way we work - workspace, employees, strategy, management models. The objective of this abstract would be to decipher the role of new age 'manager' or techager who would be responsible for managing employees working in a semi-human, semi-robotic workplace 4.0 | clinical trial, artificial intelligence, pharmaceutical industry, clinical research | source | 12 |
| 31 | An alternate way to create the standard SDTM domains | Sunil Kumar Pusarla | Sponsors who initiate clinical trials after 2016-12-17 are required to submit data from the studies in CDISC Study Data Tabulation Model (SDTM) format in their FDA submission. SDTM Implementation Guides provide the rules, assumptions, and examples necessary to create the standard domains. It is common practice to use software tools to assess compliance with CDISC standards before transmitting the domains to a regulatory agency. By pre-building the standard domain templates directly from the CDISC Share data stores, compliance with metadata (type, storage, and internal documentation) is guaranteed. These version-specific templates are then used as the scaffolding to which study-specific data are added. This paper outlines how this top-down approach minimizes coding issues as well as errors/warnings raised by metadata checking tools. A macro with dual functionality will be presented | macro variable, sdtm domain, sas library, cdisc | source | 8 |
| 32 | An Automated Macro to Compare Data Transfers | Michael Stackhouse, Franklin E. Menius |
Routine data transfers throughout the course of an ongoing study are a necessity for any analysis, but for statistical programmers, updates and modifications to the database can plague programs and break code. Sponsors can also utilize multiple vendors for one study, which can make ensuring clear communication a problem between data management and statistical programming. Problems and changes in the database might not be found until a program breaks somewhere in SDTM, ADaM, tables, listings, or figures. Therefore, it is imperative to review your data transfers and compare back to previous cuts of data. This paper will explore a macro tool used to compare two cuts of data. The tool can identify added or dropped datasets, significant changes in variable metadata, and examines datasets, looking for potentially dropped records. This macro additionally minimizes user input, and only requires the two directory paths containing your cuts of data | extract, sample, cdisc | source | 12 |
| 33 | An Efficient Tool for Clinical Data Check | Chao Su | High-quality data in clinical trials is essential for compliance with Good Clinical Practice (GCP) and regulatory requirements. If data issues are observed during statistical analysis, the interpretation of study results is impacted . As such, statistical programmers may be asked to perform data validity checks and detect potential data issues across multiple datasets during the analysis and reporting (A&R) p rocesses. The focus of this paper is NOT on finding data issues prospectively, but on how to accurately and efficiently track findings during the course of statistical analysis and programming processes. In this paper, some routinely perform ed data checks are discussed briefly. The results of the data checks are generated and summarized through a macro consisting of two sub- macros. The macro combines the results into one EXCEL file and compares the current check results with the ones in the previous versio n automatically |
excel, data science, dataset, data validation rules, randomization | source | 8 |
| 34 | An eSubmission Tracking Tool for both Components and Timelines | Todd Case | There are at least as many ways to create an eSubmission as there are companies that do it. Given all the different components required for an eSubmission, this paper will demonstrate a way to manage, organize and create a timeline for all the deliverables, including all of the specific eSubmission components | database lock, data cleaning | source | 5 |
| 35 | An Exploratory Way to Draw a Flow Diagram of Two-Arm Disposition Using SAS 9.4 M3 | Huijuan He | Many clinical trials are designed as two-arm randomized trials. Statistical programmers and biostatisticians work together to provide a flow diagram describing the disposition of subjects in the study . From the diagram, people may obtain information about the progress and status of the study visually, such as the number of subjects who were screened, randomized, dosed, discontinued or were ongoing in each arm. The CONSORT (Consolidated Standards of Reporting Trials) statement 2010 provides a flow diagram template to be filled out, available in PDF and DOC format. Users can download the template, modify as needed and manually input the numbers. Alternatively, the diagram can also be prepared using software such as Microsoft Visio although manual entry is still required. These manual approaches can lead to data entry errors | proc sgplot, macro variable, proc step | source | 12 |
| 36 | An Introduction to the Analysis of Rare Events | Nathaniel Derby | Analyzing rare events like disease incidents, natural disasters, or component failures requires specialized statistical techniques since common methods like linear regression ( PROC REG ) are inappropriate. In this paper, we'll first explain what it means to use a statistical model, then explain why the most common one (linear regression) is inappropriate for rare events. Then we'll introduce the most basic statistical model for rare events: Poisson regression (using PROC GENMOD or PROC COUNTREG ). KEYWORDS: SAS, Poisson regression, PROC COUNTREG . The graphical output in this paper is from SAS 9.3 TS1M0. All data sets and SAS code used in this paper are downloadable from www.stakana.com/RareEvents |
regression analysis, proc reg, proc step, poisson distribution, likelihood function, poisson regression, goodness of fit, box plot, proc genmod, akaike information criterion, histogram, response variable | source | 19 |
| 37 | An Investigation of Distribution Distance Measures | Charity del Sol, James E. Blum |
Draft FDA guidance proposes utilizing Earth Mover's Distance (EMD) as a means of assessing similarity of distributions of parameters such as globule size. For a univariate distribution, EMD can be easily calculated from the empirical cumulative distribution function (CDF), making it readily comparable to the Cramer-von Mises and Kolmogorov-Smirnov metrics. This paper will explore the sensitivity of each of these metrics under a variety of simulated scenarios for equal and unequal distributions, and the ef fects on subsequent population bioequivalence tests | sample, cumulative frequency | source | 12 |
| 38 | Analysis of Oncology Studies for Programmers and Statisticians | Kevin Lee | Compared to other therapeutic studies, oncology studies are generally complex and difficult for programmers and statisticians. There is more to understand and to know such as different clinical study types, specific data collection points and analysis. In this paper, programmers and statisticians will learn oncology specific knowledge in clinical studies and will understand a holistic view of oncology studies from data collection, CDISC datasets, and analysis. Programmers and statisticians will also find out what makes oncology studies unique and learn how to lead the programming and statistical teams. The paper will cover three different sub types and their response criteria guidelines. The first sub type, Solid Tumor study, usually follows RECIST (Response Evaluation Criteria in Solid Tumor) or irRECIST (immune-related RECIST). The second sub type, Lymphoma study, usually follows Cheson | clinical trial, complete response, randomization, cdisc | source | 7 |
| 39 | Analyzing Adverse Events of Special Interest using Lab Tests and Toxicity Grades | Vamsi Krishna Medarametla, Liz Thomas, Gokul Vasist |
Reporting and analyzing adverse events is one of the most important programming activities in any clinical trial. All adverse events and their severities are captured in the AE dataset. Determination of clinical re levancy and subsequent qualification as an adverse event is at the discretion of the reporting investigator. This allows for a potential bias that can lead to underreporting adverse events. Incorporating additional data sources may enhance the understanding of an agents' emerging risk profile and combat this bias. When an AE of Special Interest (AESI) has an associated quantifiable laboratory finding, such as thrombocytopenia and low platelet c ount, laboratory (LB) datasets can be used to complement the adverse events (AE) datasets. Starting with assigning a severity in the form of standardized AE toxici ty grades to the LB data, one can define an imputed AE event from the lab data | adverse events, ctcae grade, laboratory data, trial elements, clinical trial, sample | source | 6 |
| 40 | Animated Multi-dimensional Scatter Plot Visualization for Longitudinal Clinical Trial Data Reporting | Tao Shu, Jianfei Jiang |
Clinical trial data are often in longitudinal format with repeated measurements, when the aim is to analyze the changes occurring over time. One of the biggest challenges is how to summarize and present large clinical trial data in a simple and easy -to-understand way to give business intelligence, and sponsors of trials insightful information in order to make important decisions. Graphic visualizations promote effective communication of complex data to a varied audience in a more readily understandable format. The effectiveness of many temporal visualizations, such as line and bar, however, is compromised when dealing large clinical data. In this paper, we demonstrate how to generate animated scatter plot to depict the impact of different treatments on the endpoint measurements and/or clinical response across multiple time periods, and to explore relationships amongst variables. Using visual attributes such as color, size and shape, we can add up to 6 dimensions to a scatter plot | bubble plot, clinical endpoint, proc sgpanel, scatter plot | source | 7 |
| 41 | Application of Survival Analysis in Multiple Events Using SAS | Jane Lu, David Shen |
Cox proportional hazard model with time to first event as the outcome variable is commonly used to estimate the survival probability after adjusting for both baseline hazard and explanatory variables. This model counts subjects at risk at the time of first event. After that subjects are no longer considered to be at risk. There is a growing interest in analyzing recurrent or repeated events data in clinical studies. It has been proposed to extend survival models based on Cox proportional hazard approach to handle recurrent multiple events data. This paper introduces several different models in analyzing multiple events, compares and evaluates these models, and discusses their pros and cons. SAS codes are also provided for implementing these models | treatment effect, proc calis, within-subject correlation, correlation, proc phreg, hazard function | source | 9 |
| 42 | Are You Ready for It? Preparing for Your Next Technical Interview | Christine McNichol | Whether you are interviewing for a position as an intern, programmer, statistician, standards expert, or technical manager, you are likely to come across… the technical interview. Even if you have experience and know how to do your job inside and out, do you really need to spend time preparing for an interview? Yes! Though you may feel it is just going to be questions about what you do every day, there are some points about preparation and presentation that should be considered. Whether you are about to graduate college and go on your first interview, you have not interviewed since before smart phones, or even if you interview every few years, it does not hurt to be a little more prepared for an interview that may determine how and where you spend your days Monday through Friday. We will cover topics and considerations from preparing for an interview through participating in and following up after a technical interview | sdtm domain, complete response, cdisc | source | 6 |
| 43 | Automate Clinical Trial Data Issue Checking and Tracking | Krishna Avula, Dale LeSueur |
Well organized and properly cleaned data are fundamental for clinical trial data analysis. Prog rammatic data issue checks play an important role in data cleaning. Our tool is a set of SAS macro programs which automate data checking and issue tracking outside of data capture systems | excel, clinical trial, clinical data, clinical database, unique key | source | 5 |
| 44 | Automated Validation of Complex Clinical Trials Made Easy | Richann J. Watson, Joshua M. Horstman |
Validation of analysis datasets and statistical outputs (tables, listings, and figures) for clinical trials is frequently performed by double programming. Part of the validation process involves comparing the results of the two programming efforts. COMPARE procedure output must be carefully reviewed for various problems, some of which can be fairly subtle. In addition, the program logs must be scanned for various errors, warnings, notes, and other information that might render the results suspect. All o f this must be performed repeatedly each time the data is refreshed or a specification is changed. In this paper, we describe a complete, end-to-end, automated approach to the entire process that can improve both efficiency and effectiveness |
proc compare, proc step, dataset, macro variable, data cleaning, sas output, sample, where statement, independent qc, proc report, macro parameter, macro call, macro definition | source | 35 |
| 45 | Automatically Configure SDTM Specifications Using SAS and VBA | Xingxing Wu | SDTM is becoming the standard for pharmaceutical companies to submit data to FDA. The SDTM specifications are the foundation to create SDTM domains. Usually a standard study specification template (SST) is available and the study team needs to configure it to meet the specific requirements of the study. Currently, the SDTM specifications are configured manually. This paper proposes an automatic and efficient approach to configure certain parts of SDTM specifications using SAS and Visual Basic for Applications (VBA) while maintaining the original specification formats or converting the formats to the desired ones. This approach can also be directly applied to other areas to dynamically update the EXCEL files and satisfy the format requirements |
excel, dataset, sdtm domain, length statement, annotated crf, cdisc | source | 15 |
| 46 | Automating and Abstracting SAS Format Creation: Leveraging the CNTLIN Option To Build Dynamic SAS Formats That Clean, Convert, and Categorize Data | Troy Martin Hughes | The SAS FORMAT procedure "creates user-specified formats and informats for variables."i In other words, FORMAT defines data models that transform (and sometimes bin) prescribed values (or value ranges, in the case of numeric data) into new values. SAS formats facilitate multiple objectives of data governance, including data cleaning, the identification of outliers or new values, entity resolution, and data visualization, and can even be used to query or join lookup tables. SAS formats are often hardcoded into SAS software, but where data models are fluid, formats are best defined within control files outside of software. This modularity-the separation of data models from the programs that utilize them-allows SAS developers to build and maintain SAS software independently while domain subject matter experts (SMEs) separately build and maintain the underlying data models |
proc format, format statement, data standard, proc report, dataset, exception handling, proc step, macro parameter, macro variable, master data management, data governance, proc freq, data transformation, macro definition | source | 21 |
| 47 | Automation of STDM dataset integration and ADaM dataset formation | Rinki Jajoo, William Wei |
Ongoing Aggregate Safety Evaluation (OASE) is a key business imperative in enabling the clinical safety model. Integrated data specification and derivation are key parts of the necessary data pooling strategy. Currently, data integration at the compound level faces many challenges. Due to a drug's development history, variable attributes, such as name, type, and length, frequently are inconsistent across studies. Thus, when datasets are pooled for an aggregate analysis, programmers need to check the datasets and make conversions manually so that the attributes are consistent. Differences in variable name and type can cause inaccuracies, such as missing values, and differences in length can potentially cause truncation of the variable value if it is not checked thoroughly. These tasks are very time consuming and tedious, and there is the possibility of human error. The key to enable the OASE is to efficiently integrate the trial-level data to compound level in an automated fashion | adam dataset, macro variable, length statement, character variable, data integration, in=, out=, numeric variable, macro loop, proc sql, sdtm domain, macro definition, %if-%then-%else, cdisc | source | 11 |
| 48 | Avoid chasing one's Tail - Challenges and Solutions for managing changes in SDTM Standards Development | Erica Davis, Veena Nataraj |
Change is the way of life that applies to data standards. Internal organizational changes such as acquisitions may lead to the support of a new Therapeutic area (TA). External factors such as changes in regulatory requirements, new or revised standards released by Standards Development Organizations such as CDISC also need to be taken into account when building companies standards. This presentation discusses our experience, various challenges encountered and solutions taken at Shire on SDTM standards, with regards to the harmonization of legacy company standards and the development processes to ensure a faster, more reliable and more consistent implementation of SDTM standards | data standard, data cleaning, cdisc | source | 7 |
| 49 | Avoiding Sinkholes: Common Mistakes During ADaM Data Set Implementation | Richann J. Watson, Karl Miller |
The ADaM Implementation Guide was created in order to help maintain a consistency for the development of analysis data sets in the pharmaceutical industry. However, since its inception we have seen issues with guideline non-conformance which can impede this development process and carry impacts that are felt down-stream in subsequent processes. When working with ADaM data sets, non - compliance and other related issues are likely the number one source for numerous hours of re-work; not only creating unnecessary additional work for the data sets themselves, but also for reports, compliance checks, the Analysis Data Reviewers Guide (ADRG), etc. all the way down to the ISS/ISE processes. Considering this breadth of impact, one can see how devastating these sinkh oles can be. Like any sinkhole, there is a way out of it but it is a long, tedious process that will consume a lot of resources and it is always better to avoid the sinkhole entirely | sdtm domain, imputation, raw data, change from baseline, crossover study, per protocol, interim analysis, treatment period, adverse events, cdisc | source | 18 |
| 50 | Be a "Gyro", Turn your SAS Report into a Sandwich | Michael Stout | The title of this abstract may make you hungry. Bread is an essential part of any sandwich; just as titles and footnotes are an essential part of any report. Titles and footnotes provide context (the bread) and support for tables, figures and listings (TFLs). There are many methods for assigni ng titles and footnotes. This paper will show how standardization of titles and footnotes using a SAS MACRO can improve the appearance and quality of deliverables. Key data points that should appear on every report will be reviewed. Lastly, the paper will present various techniques for controlling the format of titles and footnotes. Get ready for the best report sandwich ever! | proc report, macro variable, proc step, sample | source | 6 |
| 51 | BEACH: an open platform for building interactive and automatic analysis powered by R/Shiny | Danni Yu, Michael Man |
dataset, clinical trial, data cleaning, data science, excel | source | 19 | |
| 52 | Best Practice for Explaining Validation Results in the Study Data Reviewer's Guide | Kristin Kelly | In addition to providing the datasets, define.xml and annotated CRF (aCRF) in a clinical SDTM package, the Technical Conformance Guide (TCG) recommends to also include a Study Data Reviewer's Guide (csdrg.pdf) to convey any further information about the study data that cannot be described in the aCRF or the define.xml that would aid the reviewer in understanding the data. An important section of this document contains results of automated validation checks for conformance to SDTM and regulatory business rules that cannot be resolved due to data oddities. Each issue that remains should be listed with an explanation for why the specific validation check cannot be rectified. Providing a vague, incorrect response or no explanation at all can affect reviewability of study data. This paper will focus on examples and best practices for providing comprehensive explanations of issues in the cSDRG | data cleaning, label statement, data validation rules, cdisc | source | 6 |
| 53 | Best Practices in Data Standards Governance | Melissa R. Martinez | Most organizations working in the pharmaceutical and biotechnology industries have adopted CDISC submission data standards by now, but the challenge of effective governance and compliance within an organization remains high. CDISC data standards are notoriously open for interpretation and the assumptions and understanding of the data standards can vary widely among users. Adopting CDISC standards is more than just making CDISC-style submission datasets and using tools to help with the programming and compliance checks. Each organization needs to define its own interpretat ion of CDISC standards to ensure consistency among its studies, put in place workflows and processes to facilitate the governance process, determine what it means to be compliant with the CDISC data standards, and find tools to help with the governance process and compliance determination | data standard, life sciences, raw data, controlled terminology, cdisc | source | 4 |
| 54 | Beyond MERGE for Combining Datasets | David W. Franklin | Merging two datasets is one of the most common actions that a SAS programmer does when manipulating data to bring it into a form for either storage or analysis. The most common way data is merged is with the use of the MERGE statement inside a DATA STEP but there are others. This paper will introduce methods that use PROC SQL, HASH tables, POINT and KEY options in a SET statement and PROC FORMAT; looking at one-to-one, one-to-many and many-to-many situations |
proc sql, proc step, merge statement, hash table, set statement | source | 4 |
| 55 | Bring Unix into SAS | Jaiying Guo | While SAS is a great tool for clinical programming, it has some limitations. Most programs will work on one single file, and cannot change the folder structure or apply similar changes to a group of files under one directory. Thus, automated programs must be supplemented with manual window actions. With repeated tasks, these manual window actions become time consuming. For example, if you have 50+ programs to move from one location to another, the program path needs to be update d for all programs. Manually updating each file is time consuming. SAS has the ability to communicate with external software such as EXCEL, Word and UNIX. UNIX has the power to execute commands to edit folder structure and file content. This paper will describe step-by-step instructions of how to utilize Unix commands within the SAS environment |
macro quoting function, macro definition | source | 4 |
| 56 | Build Models without Code with the new SAS Visual Interface | Jim Box | SAS has made it even easier to build statistical and machine learning models with a new visual interface. See how to partition data, build and compare multiple models without writing any code. Learn how to export model and scoring code for future use | proc hpsplit, logistic regression, machine learning, regression analysis, logistic distribution, data cleaning, gradient boosting, bar chart, feature selection, response variable, training set | source | 12 |
| 57 | Building Simulated CRF Data for Study Mock-Ups | Ashley Kesler, Hisham Madi, Jonathan W. Duggins, James E. Blum |
Simulation of mock trial data has wide-ranging utility in analysis plan development, data mapping and TLF programming, quality control for clinical trials, and development of employee training activities. Simulation strategies and data use will be covered in a variety of examples | sample, clinical trial, raw data, vital signs, adverse events, integer, proc step, proc sql, randomization, macro variable | source | 10 |
| 58 | Calculating Restricted Mean of Survival Time | Jiuzhou Wang, Tony Qi |
The commonly used descriptive statistics for survival analysis include median survival time, survival rate at given time point(s), all derived from Kaplan-Meier estimate. Mean survival time (MST), however, has received less attention in the field of clinical research, partly because it is often subject to underestimation due to the largest observation being censored. Restricted mean survival time (RMST) can provide additional insight to the survival distribution and is not subject to underestimation by definition. In this paper, we discuss the calculation of RMST at any given time point (up to the largest observation) using PROC LIFETEST. A SAS macro is also developed to streamline the calculation when inference of RMST is required at multiple time points |
survival analysis, proc step, proc calis | source | 7 |
| 59 | Can you easily create your own interactive dashboards in SAS? Is SAS 4GL all you need? | Piotr Karasiewicz | Easily finding the right information is key to smart and timely business decisions. In Clinical Trials, it can even mean the difference between life and death! There are huge numbers of visualization tools available, but are they best? When you understand your data and user needs fully, are you better off building something from scratch within SAS? Using dummy Clinical Trial data, we'll compare our in-house web-based serverless solution, which uses PROC JSON and the Open Source D3 library for graphing and drill-down. We will also review SAS techniques helpful in building in-house platform: PROC GKPI, PROC STREAM and custom html+javascript codes run through SAS. As an alternative we look at paid- for solutions like Microsoft Power BI Desktop. Finally, we'll consider if the breadth of options offered by tools like Power BI help or hinder, compared to targeted reports defined by experts |
proc step, data step, proc json, txt, excel | source | 18 |
| 60 | Challenges & Strategies in PKPD Programming | Jing Su, Jiannan Kang |
PKPD Programming specializes in handling Pharmacokinetic (PK) data and creating data sets for PK analysis and PKPD modeling. The PK data creation process faces many difficulties including lack of industry standards on PK data sets, disparate data sources, evolving data collection standards over t ime, source data accuracy and completion, and complicated programming logic due to the additional requirements of the specialized software for PKPD modeling. These difficulties make it even more challenging to achieve quick turn-around with good quality and traceability. In this paper, these challenges will be explained in detail and the strategies to overcome them will be discussed. This paper will also highlight the differences between statistical clinical programming and PKPD programming, and emphasize the importance of the collaboration between these two programming groups on consistent usage of clinical data | imputation, sample, clinical data, raw data, key, statistical programmer, data cleaning, response variable, flag variable, adam dataset, concomitant medications, cdisc | source | 6 |
| 61 | Challenges for Implementing Study Data Standardization Plan (SDSP) in Legacy Data Submission | Shengfeng Ho | The development of a Legacy Data Conversion Plan (LDCP) is identified as a need in recent released Study Data Technical Conformance Guide in August 2017 which associated with FDA Guidance on Providing Regulatory Submissions in Electronic Format - Standardized Study Data (Final FDA Guidance December 2014). Unlike Study Data Standardization Plan, LDCP serves for legacy study and assists regulatory reviewers in understanding your data conversion process and possible traceability issue. This paper presents the challenges for implementing LDCP in legacy study for ISS/ISE submission. The case study and examples shown in this paper are our working experiences while leading a group of programmers in the effort to convert data for 9 sub-studies from legacy format to industry-standard | adam dataset, data transformation, cdisc | source | 9 |
| 62 | Challenges in Implementing Device Related ADaM and SDTM Standards in a Drug-Device Study | Wenying Tian, Maureen Maunsell, Karin LaPann |
In this case study, a Medical Device was used to deliver an investigational study drug. The analysis needed to monitor the device functionality at the subject level. Of interest were the following events and outcomes: the occurrence of any device malfunctions, whether or not they were resolved and whether a device was removed or another intervention occurred. In order to accurately represent the device life cycle during the study, we implemented the seven published SDTM device domains in addition to the foundational SDTM for the creation of the analysis datasets (ADaM). The challenge occurred when it was difficult to paint a picture of the device lifecycle with existing SDTM data. The derivations in ADaM were becoming more and more complex. Several data issues were identified, and when traced to the source, we found better ways of collecting the data, simplifying the ADaM process | sdtm programming, adam dataset, sdtm domain, crf, clinical trial, raw data, cdisc | source | 9 |
| 63 | Combining Electronic Diary Seizure Data with Visit-Based Seizure Data in the MONEAD Study | Ryan May, Julia Skinner |
Electronic diary (eDiary) mobile applications can be a useful tool to collect subject - reported data outside of scheduled study visits. However, integrating eDiary data with data stored in a traditional electronic data capture (EDC) system can present uniq ue challenges. eDiary applications sometimes function under different business rules than traditional EDC systems. Collecting data in different software systems makes linkage of records more difficult, even if a common key exists. This paper describes me thods used in the Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) study to combine daily subject - reported seizure data recorded in an eDiary application with clinically assessed seizure information collected in the study ED C system . I | edc, missing value | source | 8 |
| 64 | Coming soon: ADNCA and the PK submission | Luke Reinbolt, Liz MacDonald |
This talk will give a preview of the soon to come non-compartmental pharmacokinetic analysis CDISC ADaM standard (ADNCA). ADNCA will be the submission data set that is sent to the PK scientist for calculation of non-compartmental PK parameters. We will present the proposed guidance and give an overview of the steps involved in developing this standard for analysis and submission to the FDA. The ADNCA data set will adhere closely to the BDS structure. The dataset itself will contain a combination of PK concentration data, its corresponding dosing data, and any additional data necessary to complete the non-compartmental analysis. The inclusion of this extra data is necessary, in part, because of the need to present timing variables related to the concentration event and dosing time. We will discuss the rationale for those timing variables as well as several other variables that are commonly of interest to the PK scientist | sample, key, pk parameter, compartmental analysis, adpp, cdisc | source | 6 |
| 65 | Common Data Related Review Issues and Prevention: A Statistical Reviewer's Thoughts | Huanyu Chen | adam dataset, quality control, raw data, ae grade, analysis set, data cleaning, efficacy summary, cdisc | source | 18 | |
| 66 | Compare Pharmacokinetic Data Submission Processes in CDISC Environment | Xiaopeng Li, Shallabh Mehta, Edward Elam |
Pharmacokinetic (PK) analysis is an essential part of early clinical research studies. In early clinical research, the turnaround time of PK analysis is one of the key factors. Shorten turnaround time of PK analysis can reduce the overall drug development time. FDA recommends generating Study Data Tabulation Model (SDTM) PC and PP and Analysis Data Model (ADaM) ADPC and ADPP domains, respectively, as part of PK analysis. There are different processes that can be followed for PK Analysis apart from the process mentioned in PhUSE White paper titled "Analyses and Displays Associated to Non-Compartmental Pharmacokinetics - with a focus on clinical trials". This paper introduces different PK analysis processes and compares it with PhUSE and other alternative processes | key, pk parameter, clinical data, clinical research, clinical trial, cdisc | source | 7 |
| 67 | Complexity in collection of PGx data and challenges in mapping to SDTM | Rama Kudaravalli | Pharmacogenomics is the study of how genes affect a person's response to drugs. It is the combination of pharmacology and genomics to develop effective, safe medications and doses that will be tailored to a person's genetic makeup. This new field is still in its infancy and its use is currently quite limited, but new approaches are emerging in clinical trials. The pharmacogenomics research is entering the era of "Big data"- a term that refers to the explosion of available information. This vast amount of data brings both opportunities and new challenges in data conversion, standardization and analysis. It is a complex process to convert and standardize the pharmacogenomics data collected from clinical laboratory, as it is difficult to translate the results into actionable prescribing decisions for the affected drugs. This paper will introduce the challenges faced in collection of pharmacogenomics data, curation and mapping it to SDTM domains | sample, clinical trial, sdtm domain, cdisc | source | 9 |
| 68 | Complying with the ADaM Compliance Rules | Alyssa M. Wittle, Steven R. Kirby |
Is it hard to determine sometimes if we are living in a deluge of compliance rules for ADaM or a drought? There are multiple sources available: company specific, Pinnacle 21, PMDA, and CDISC; but what happens when they do not agree? What about when the compliance rules do not fully agree with the ADaM Implementation Guide? Or when there are not rules available for the latest Implementation Guide version? Lucky for us there is only one "official" version of the compliance rules available and are created by the CDISC ADaM team. While these rules are available, they have not always been easy to understand and implement - perhaps they even have some issues or false negatives. As a co-lead of the ADaM Compliance Rules sub-team, I have seen these challenges in depth and in detail | adam dataset, data standard, data cleaning, cdisc | source | 4 |
| 69 | Confusing Data Validation Rules Explained | Michael Beers | An important step in the submission of tabulation and analysis data to regulatory agencies is the validation of the data according to rules developed by regulatory agencies and standards development organizations (SDO). Pinnacle 21 tool is commonly used throughout the industry to produce these validation results. Errors and warnings in study data reported by the Validator must be corrected. If the issues won't be corrected, then they must be explained in the Reviewer's Guide. This dispositioning of validation findings can be a challenge, as some business rules and validation messages are more confusing than others. Unclear understanding of validation results can lead to important data issues not being corrected, or incorrect explanations in the Reviewer's Guide. This presentation will walk through some of the more confusing of the validation rules and explain what the findings would mean | data cleaning, data validation rules, controlled terminology, sdtm domain, concomitant medications, cdisc | source | 6 |
| 70 | CONSORT Diagrams with SG Procedures | Prashant Hebbar, Sanjay N. Matange |
In Clinical trials, Consolidated Standards of Reporting Trials (CONSORT) flow diagrams are an important part of the randomized trials report. These diagrams present a bird's eye view of the flow of patients through the different stages of the trial. The SG Procedures do not support a statement for drawing these diagrams. But with some data processing steps and tricks, we can draw CONSORT diagrams using SG Procedures. In this paper, we show how to harness power of SG Procedures to create these diagrams | proc sgplot, hash table, proc step, series plot | source | 9 |
| 71 | Core and Extension studies - Challenges and solutions towards SDTM submission | Murali Marneni | It is becoming common practice in clinical trials t o assess the safety, tolerability and efficacy of d rugs using Core and Extension studies. Most of the time , subjects that were enrolled and completed the mai n study will be enrolled into the Extension studies w ith subject consent. Data for these studies are oft en collected in the same EDC system with just one eCRF and have two protocols for both studies but the client asks us to treat them as different trials. T his often causes some confusion and challenges in creating the SDTM mapping for these two studies. Th e intention of this paper to produce high quality SDTM data for Core and Extension studies that are s ubmitted as two separate studies | adverse events, medical history, concomitant medications, cdisc | source | 6 |
| 72 | Creating a Publication Quality Graph Embedded in another Graph | John Henry King, Warren F. Kuhfeld |
Several recent papers in the New England Journal of Medicine (NEJM) display graphs inside of graphs (Marso et al. 2016b, a; Zinman et al. 2015; Holman et al. 2017). Typically, both graphs are Kaplan-Meier plots. The larger graph has a Y axis that ranges from 0 to 100. The failure curves occupy a relatively small part of the graph and extend perhaps no more than 10%-20% of the way up the Y axis. Inside this graph, in the open white space, you can place another graph in which the Y axis extends from 0 to slightly above the maximum value. The outer graph shows the failure probability in the full 0%-100% context, whereas the inner graph zooms in to the results and provides additional text and annotation, such as curve labels, the hazard ratio, and p-values. The outer graph also displays the number of subjects at risk. The format for the NEJM is quite specific. This paper illustrates how to use ODS Graphics to create a graph inside a graph that is suitable for publication in the NEJM |
proc step, sg annotation | source | 20 |
| 73 | Creating an ADaM Data Set for Correlation Analyses | Chad W. Melson | The purpose of a correlation analysis is to evaluate relationships between two variables, likely from two different types of measurements (e.g., blood pressure and lab results). As a result, correlation analyses usually require data from multiple source data sets. In this paper, I describe how an ADaM data set using the OTHER class structure (one of the four classes of ADaM data sets: ADSL, BDS, and OCCDS are the others) was constructed from multiple ADaM data sets to produce outputs displaying correlations of actual and change from baseline values at individual visits. Pros and cons of the non -normalized data set structure are also presented, which may give programmers ideas on how to create an ADaM data set for similar types of analyses | correlation, change from baseline, adam dataset, forest plot, raw data, cdisc | source | 11 |
| 74 | Creating and Customizing Graphics using Graph Template Language | Saihua Liu, Yanmei Zhang, Titania Dumas-Roberson |
Graph Template Language (GTL) is a powerful language in SAS 9.4. PROC TEMPLATE in GTL can create customized graphs. The Graphics Template Language can also be used in conjunction with special DATA STEP features to produce graphs independently. This paper introduces the GTL with clinical trial examples. In addition, it provides some methods to solve problems we often meet when we create graphs. Such as, label is too long to fit in one line, and legend is truncated, and so on. Creating complicated graphs by using an annotate data set is discussed in the paper, too |
bar chart, proc step, proc template, clinical trial | source | 9 |
| 75 | Creating Data Shells with the DS2 procedure | Matthew J. Wiedel | When one creates a standardized database, each dataset has a specific title, structure, and collection of variables. They are most likely created from one or more raw datasets. Variables within each dataset will most likely have specific lengths, formats, and labels. As studies become more complex, the requirements placed upon the information become more nuanced and complicated. So far, the DATA STEP and the SQL procedure are quite sufficient. However, there may come a time when a different approach to data creation will be needed. The DS2 procedure is a completely different way to work with a SAS dataset than that of the good ol' DATA STEP. Object oriented programming is a different way of thinking, with its threads, methods, packages, declaration statements… It also provides an extended array of formatting options. PROC DS2's declare statement is akin to the DATA STEP's attribute statement as it assigns formats and labels to variables |
data step, proc step, macro variable, proc sql, sas programmer | source | 7 |
| 76 | Customized Project Tracking with SAS and Jira | Nancy Brucken | As programmers and statisticians, most of us are far better at programming tasks than we are at project management. Jira is a powerful and inexpensive commercially-available application designed to help programming teams track their progress on projects. It is built on top of a PostgreSQL database, which can be queried from SAS to generate a variety of custom reports | data cleaning, sample | source | 8 |
| 77 | Data Management and CDISC Formatting for Transdermal Patches | Lois Lynn | Transdermal patch use is an alternative delivery method for medication that has many advantages to the user. One advantage of transdermal medication delivery is that it bypasses the gastrointestinal (GI) tract and goes directly to the location where it is needed; patches therefore tend to have fewer GI side effects than oral medication. Clinical trials to evaluate the safety and efficacy of transdermal products are a specialty area of investigation included in the Food and Drug Administration's Guidance's for Devices. The Food and Drug Administration (FDA) regulations and guidance's mention study designs, objectives and measurement scales to evaluate the various aspects of transdermal patch performance. Some are the Fitzpatrick Skin Type, Patch Application Site Evaluation, Study Drug Administration, Adhesion, Discomfort, Pain, Irritation, and Adhesive Residue | clinical trial, study day, adverse events, data governance, randomization, cdisc | source | 11 |
| 78 | Data Standards for Population Pharmacokinetic (POPPK) and Non-Compartmental Analysis (NCA) | Neelima Thanneer | Pharmacometrics (Pm) Programming specializes in creating modeling-ready datasets for Pm analyses. Challenges to create these datasets include the familiar list of accounting for deficiencies in dosing information, standard rules for handling source data issues, and variability in the specifications. Given these challenges the datasets are prepared and documented inconsistently across the companies and sometimes within the companies, making it difficult to review and exchange the datasets. To address this, International society of pharmacometrics (ISoP) has come up with an initiative to standardize the POPPK datasets. This paper will go over the challenges associated with programming POPPK datasets and describes the best practices to handle them. In addition, an update will be provided on the ISoP Data Standards Initiative and ADaM NCA effort from the CDISC PK working group and how that can augment and impact the POPPK and NCA dataset creation process | imputation, compartmental analysis, population pk, sample, raw data, data standard, proc freq, cdisc | source | 7 |
| 79 | define.xml (noun): Fear Inducing Task for SAS Programmers | Kjersten Offenbecker, Antonio Cardozo, Kirsty R. Lauderdale |
Faced with creating the define.xml, whether for SDTM or ADaM, programmers tend to shudder with fear. Why is that? Maybe because it tends to be tedious, manual and is not 'defined' very well. What needs to be included, what should it look like and how on earth do you create it? We all know it's a vital piece of any submission and most sponsor companies now require it, even if they have no idea what it is or why they need it. Many companies have a 'home-grown' system (a.k.a. macro) which they use to generate this, but did you know that there is a better way. In our paper, we will show you how to take the fear out of the define.xml creation process | proc sort, proc step, cdisc | source | 17 |
| 80 | Derivations of Response Status from SDTM Domains using RECIST 1.1 | Christine S Teng, Lei Pang |
The Response Evaluation Criteria in Solid Tumors (RECIST) provide standardized rules for solid tumor response assessments. RECIST defines categories of response, which include Complete Response (CR), Partial Response (PR), stable disease (SD), and progressive (worsening) disease (PD), after the start of treatment. Endpoints based on tumor response as determined by RECIST 1.1 can be the basis for regulatory approval by both the Food and Drug Administration (FDA) and European Medicines Agency (EMEA). This standardization of rules provides a framework for reproducible analysis of results in many oncology clinical trials. This paper utilizes the CDISC SDTM oncology domains to illustrate by examples the use of SAS code to derive tumor response based on RECIST 1.1 to validate site and central response status | best response, complete response, data governance, cdisc | source | 10 |
| 81 | Designing Flexible Data Standards Models | Melissa R. Martinez | When we think of CDISC submission data standards, we often focus on the tables and variables that the standards describe. We build products and business processes around creating these tables , but isn't there more to the story than just the submission data sets? This paper explores ways to design data standard models with data processing in mind. It also shows how to add to industry data standard models to define and standardize additional attributes, such as de-identification algorithms, variable calculations, and even blocks of SAS code. Finally, this paper describes features available in the forthcoming SAS Life Science Analytics Framework version 5.1 that facilitate these concepts | data standard, life sciences, raw data, clinical trial, excel, cdisc | source | 4 |
| 82 | Developing Large SAS Macro Applications | John Ingersoll | The SAS macro language is used every day by statistical programmers for repetitive code and substitutions in programming data sets and displays. But it can also be used to develop larger applications more similar to a traditional software package than what many think of when they hear "SAS macro." However, most companies don't have a "SAS Macro Application Development Department" and many talented and experienced SAS macro programmers don't have experience in a traditional software development environment. This paper presents the development of a validated application writte n solely in the SAS macro language, containing 124 macros. It discusses the challenges, considerations, and solutions to issues experienced in the development process, with attention to the fuzzy distinction between a SAS macro and a software application or system | data cleaning, macro definition, macro variable, sas programmer, clinical programming, macro flow control, stored macro, macro call, sas programming, autocall macro, macro catalog, macro facility, error, statistical programmer, proc step, error message, macro loop, %if-%then-%else | source | 13 |
| 83 | Diagnostics of technical errors in define.xml file | Sergiy Sirichenko, Max Kanevsky |
Study metadata in standardized format plays critical role in automated processes. Wrong or missing information and technically invalid representation of define.xml file may result in incorrect processing of submission data at FDA and PMDA. Today there is a lack of understanding in what study metadata is critical for automated processes and how to ensure its correct implementation. Invalid technical implementation of Define- XML standard and additional regulatory requirements may lead to major errors in metadata structural consistency and missing information. Pinnacle 21 software helps identify these errors, but some validation messages may be tricky and hard to interpret by less experienced users. This paper will identify the most important study metadata utilized by automated processes and show examples of errors caused by missing or incorrectly implemented define.xml file. This paper will also provide guidance on how to perform diagnostics of technical errors in define.xml file | data cleaning, cdisc controlled terminology, controlled terminology, flag variable, data validation rules, data standard, verbatim term, cdisc | source | 9 |
| 84 | Discover the Deeper DATA Step | Timothy J. Harrington | All SAS Programmers will be familiar with the SAS DATA STEP, however, there many SAS users who may be unfamiliar with or even unaware of facilities in the DATA STEP which, if exploited, would result in easier programming, code that is easier to understand, better run-time performance, and easier QC and verification. This paper lists and describes the most common of these DATA STEP 'extras', describes how and when to use them, using examples of SAS code, and compares with alternative methods of achieving the same result |
macro variable, by group processing, proc step, missing value, order by, character variable, proc sort, proc sql, put statement, set statement, retain, numeric variable, output, symbol table, automatic variable | source | 14 |
| 85 | Diving Deep into SAS ODS Graphics Styles | Dan Heath | Creating an effective style for your graphics can make the difference between clearly conveying your message to your audience and hiding your message in a sea of lines, markers, and text. There have been a number of books written on the concepts of effective graphics, but you need an understanding of how styles work in your environment to correctly apply those principles. The goal of this paper is to give you an in-depth discussion of how styles are applied to ODS graphics, from the ODS style level all the way down to the graph syntax. This discussion will include information about differences in grouped versus non-grouped plots, precedence order of style application, using style |
source | 19 | |
| 86 | Doctor-s 'Prescription' to Reengineer Process of Pinnacle21 Community Version Friendly ADaM Developm | Aakar Shah, Tracy Sherman |
With development and availability of Pinnacle 21 community version software, a lot of power has been given back into the hands of SAS programmer at any skill level to generate Define.xml 2.0 and to validate ADaM used for submission. The new prescriptive approach of generating Define.xml upfront creates an opportunity to not just modify but re-engineer the ADaM development process by either using Define specifications as programming specifications, ADaM attribute assignments and/or data validation. It also improves communication within the programming team and with stakeholders such as the study statistician for reviewing ADaM specification throughout the study life cycle. This paper will outline the re-engineered process of ADaM development including step-by-step instructions to complete Define.xml specifications used by Pinnacle 21 Community software | adam dataset, data cleaning, cdisc | source | 9 |
| 87 | Dosing In NONMEM Datasets an Enigma | Sree Harsha Sreerama Reddy, Vishak Subramoney |
Dosing data plays an integral part in Non-Linear Mi xed Effects Modelling(NONMEM ) or Modeling and Simulation(M&S) data sets. Representing accurate dosing information and history in these data sets leads to an efficient and robust Pharmacokinetic/Pharmaco dynamic(PK/PD) analysis. Dosing source data collected for a Phase II/Phase III QD (once daily dose) or BID (twice daily dose) study (e.g.: Oncology) can often be complex and would require complicated transformations before it is included in a NONMEM data set. The transformations may involve expansion of d osing history, multiple imputations, dose time imputations etc. There are various methods that cou ld be used to program the Dosing module in a NONMEM data set building exercise. Choosing the best appr oach for a study will depend on a range of factors like - it results in a compact NONMEM data set, do you want a simple and direct approach , is incomplete dosing source data collected for the study and various other factors | sample, raw data, dose interruption, sdtm domain, cdisc | source | 11 |
| 88 | Easy CSR In-Text Table Automation, Oh My | Janet E. Stuelpner | Your medical writers are about to embark on creating the narrative for the clinical study report (CSR). They are tired of asking the programmers to re -run the in-text tables and then cutting and pasting them into the report. And each time that the report needs to be updated, they need to copy and paste the in -text tables again. Isn't there a better way to do this in a more automated fashion? This paper will show you that using your CSR template along with SAS Enterprise Guide with stored processes and SAS Add-in for Microsoft Office, this becomes an easy task for the medical writers | csr, clinical trial, sas enterprise guide, excel, proc freq | source | 6 |
| 89 | Electronic Data Submission and Utilization in Japan | Yuki Ando | data cleaning, clinical research, data validation rules, clinical data, clinical trial, data standard, csv | source | 25 | |
| 90 | eSubmission - Are you really Compliant? | Majdoub Haloui, Suhas R. Sanjee |
Pinnacle 21 Enterprise is a valuable tool for SDTM, ADaM and define.xml validation. However, the tool is still evolving and has some limitations, so additional quality checks need to be created to supplement the tool's assessments to ensure accur acy, validity and completeness of SDTM/ADaM datasets as well as the define.xml. Adding manual quality checks can be cumbersome and time consuming. In this paper, we will discuss the extra quality checks that are implemented beyond what is available in the Pinnacle 21 Enterprise tool. The paper focuses on how these checks are created and an efficient approach for converting define.xml into SAS datasets so they can be validated against the actual SDTM/ADaM datasets | dataset, adam dataset, cdisc | source | 5 |
| 91 | Expand Your Skills from SAS to R with No Complications | Andrii Korchak | Learning a new language is rarely a quick and simple process. Learning word by word from the very beginning requires a great deal of energy and time. Is there a way not to start all over again? Is it possible to use the knowledge of your native language, and just pull up a little to begin to communicate? In this article, we will outline the main similarities and differences between SAS and R, and show how to start programming in R relying on your SAS skills. In particular, we will program DATA steps and procedure calls to perform data analysis in both languages, and present results by creating a table and graphs. We will learn how to recreate some popular SAS operations in R and make use of a few R - specific ones. Use your SAS dictionary to break through a language barrier! | by group processing, proc sort | source | 12 |
| 92 | Experience of electronic data submission via Gateway to PMDA | Kumiko Kimura, Iori Sakakibara, Laurence Carpenter |
PMDA started accepting the submission of electronic study data of clinical trials for new drug applications (NDA) on October 1st, 2016. In this poster, our preparation for the electronic data submission and experience of PMDA consultations for ADaM, SDTM and Define.XML, including how to explain the errors in Appendix 8, will be shown. In addition, the experience of submission via Gateway system will also be shown. In Gateway system, it is difficult for the sponsor to input descriptions for each item of study data by clicking every single file. To avoid that, a TSV file is provided from Gateway system to edit and update information outside the system. However, this TSV file can only be provided after the advance notice of the application (1- 5 weeks before NDA) to PMDA. To prepare in advance, creating the TSV file using a SAS program will also be shown | data cleaning, error message, key, excel, sas programming, cdisc | source | 7 |
| 93 | Exploring techniques of using multiple SET statements in a DATA step | Yizhong Huang | SET statement is one of the most used statements in SAS programs. Yet, some powerful features of SET statement have been widely ignored. Usually, PROC SQL or MERGE statement in a DATA STEP will be used to handle "One-to-One" or "Many-to-One" merge. When there is "Many to Many" merge, PROC SQL procedure seems to be the only choice for SAS programmers. This is NOT true! Using multiple SET statements in a DATA STEP, this paper will explore implementations of "Many-to-Many" merge compared to PROC SQL joins, as well as other techniques that give you more flexibility and make your programs more efficient |
set statement, proc step, proc sql, inner join, full join, do loop, master data management, right join, treatment arm, by group processing, proc freq, proc sort, put statement, merge statement | source | 10 |
| 94 | Exploring use of R for Clinical trials | Kalpesh Prajapati, Parvee Kumar |
SAS is leader for data analysis in health care industry being accepted by regulatory bodies worldwide. R is one of the open source software used in academic and other research firms. As regulatory bodies never endorse any particular software, R can revolutionize the analysis part of clinical trials | clinical trial, data science | source | 7 |
| 95 | Family of PARAM***: PARAM, PARAMCD, PARAMN, PARCATy(N), PARAMTYPE | Kamleshkumar Patel | Creating compliant analysis parameters (and related variables) in Basic Data Structure (BDS) is an important key for successful design of ADaM compliant datasets. Programmers take different approaches in creating analysis parameters in BDS. For ADaM specification writers, there are certain guiding tips to create PARAM and related variables like PARAMCD, PARAMN, PARCATy(N), PARAMTYP for various datasets. Many specification writers struggle to understand and implement concept of PARAM related variables to make 'analysis-ready' datasets in ADaM. Furthermore, there are many challenges in handling PARAM and PARAMCD variables. For instance, creating PARAM and PARAMCD in ADLB datasets from LBTEST and LBTESTCD is tricky. Hence, it can create complexities if those variables are not handled carefully. In addition, it is very important to design other variables like PARCATy(N), PARAMTYP for BDS datasets | mock shell, sample, proc step, controlled terminology, overall survival, standard deviation, adam dataset, cdisc | source | 12 |
| 96 | Fifteen Functions to Supercharge Your SAS Code | Joshua M. Horstman | The number of functions included in SAS software has exploded in recent versions, but many of the most amazing and useful functions remain relatively unknown. This paper will discuss such functions and provide examples of their use. Both new and experienced SAS programmers should find something new to add to their toolboxes | anyalpha, if-then-else, character functions, tranwrd function, notalpha, sas programmer, translate, whichn, missing value | source | 8 |
| 97 | Framework for Critical Thinking - Introducing Lean Philosophy to Statistical Programmers | Wenyun Ji | 'Lean', coined by Jim Womack's team at MIT in the late 1980s, describes the Toyota Production System that aims to deliver goods and services of the highest quality at the lowest cost in the shortest lead time. We're seeing businesses in all industries including healthcare are using the 'lean' principles to guide their best practices. Lean is based on creating a culture on how to respond to a problem rather than what tool to use. Lean approach creates a problem solving mindset. It has four main principles: customer first, trusting and respecting each other, continuous process improvement (CPI), and elimination of waste. This paper will introduce and discuss the lean principles in the context of statistical programming in the pharmaceutical industry. Customer first is to understand what is of the highest value to the customers. Trusting and respecting people is to build the best team and create a learning-centered environment | statistical programmer, clinical programming | source | 6 |
| 98 | From FREQing Slow to FREQing Fast: Facilitating a Four-Times-Faster FREQ with Divide-and-Conquer Parallel Processing | Troy Martin Hughes | With great fanfare, the release of SAS 9 delivered multithreaded processing to a single -threaded SAS world. Procedures such as SORT, SQL, and MEANS could now run faster by taking advantage more fully of system resources through parallel processing paradigms. Multithreading commonly implements divide-and-conquer methodologies in which data sets or data streams are decomposed into subsets and processed in parallel rather than in series . Multithreaded solutions are faster (but typically not more efficient ) than their single-threaded counterparts because execution time (but not system resource utilization) is decreased. As the costs of memory and processing power have continued to decrease, however, there remains no excuse for not implementing multithreaded processing wherever possible. To this end, and because SAS unfortunately abandoned some hapless procedures in single-threaded Sheol, this text aims to reunite the single-threaded FREQ procedure with its multithreaded bedfellows |
proc freq, parallel processing, sample, macro variable, character variable, frequency analysis, proc step, sas base, repeated measures, macro definition, stress testing, correlation, numeric variable, macro loop, key performance indicator, sysrc, stepwise selection, regression analysis, proc sql, %if-%then-%else | source | 23 |
| 99 | Generating Define.xml from Pinnacle 21 Validator | Pinky Anandani | Define.xml is an XML document that describes the structure and contents (metadata definitions) of the data collected during the clinical trial process1. In December 2011, the CDER Common Data Standards Issues Document2 stated that "a properly functioning define.xml file is an important part of the submission of standardized electronic datasets and should not be considered optional." While the metadata definitions of a clinical study may not be the most difficult deliverable to create, transforming those specifications to an xml file can be a daunting experience for many SAS programmers who lack skills in other programming languages. This paper details a simpler process that can be followed to generate the Define.xml document for a study without the need of any additional programming skills. Use of this procedure eliminates the additional time needed to transform study metadata into Define.xml at the end of the study | where statement, adam dataset, excel, sdtm domain, label statement, clinical trial, cdisc | source | 15 |
| 100 | Generating Receiver Operating Characteristic (ROC) curve using SAS Macros | Jiaying Guo, Isabella Wang |
Early identification of responsiveness to treatments has significant clinical and economic implications. The Receiver Operating Characteristic (ROC) curve is a key tool for diagnostic test and has been used in identification of early clinical responses that could predict long-term outcomes. [1,2] Youden's Index is often used in conjunction with ROC analysis[3], and the maximum value of Youden's index may be used as a criterion for selecting the optimum cut-off point[4]. Based on the reported method[1,2], this paper illustrates how to generate ROC curves and Youden's index using SAS Macros and simulated data. This visual method could help research scientist to estimate the predictive accuracy easily | roc curve, logistic regression, logistic distribution, regression analysis | source | 8 |
| 101 | Get Your Sort On: Macro Driven Method for Automated Unique Sorting | Kevin Miller | With an ever growing desire to automate processes for increasingly more tasks, pressure falls to programmers to make their programming more adaptable and efficient. One such task is uniquely sorting data; this paper will present a method, a SAS macro, to perform this task using minimal oversight | macro variable, by group processing, macro definition, proc sort, proc step, proc sql | source | 3 |
| 102 | Global checklist to QC SDTM Lab data | Murali Marneni, Sekhar Badam |
Laboratory data is one of the primary datasets that we usually see in every submission in clinical tri als. It also is the primary dataset that most of the progra mmers usually encounter with a lot of problems. Due to greater complexity, as it requires more data manipu lations to create the SDTM lab data, most programmers completely rely on the Pinnacle report to make sure the SDTM lab data is accurate and can be used for downstream tasks. It is important to r ealize that passing the Pinnacle issues checks does not mean everything is good. So, it is important to do additional checks to catch the data issues ahead of time since reporting the data issues later costs more mo ney and resources, and sometimes leads to inaccurate results in analysis. This paper talks ab out effective and useful techniques that help to QC your SDTM Lab data, and will help as a checklist for lab data on your future projects | laboratory data, sample, proc sort, cdisc | source | 10 |
| 103 | Graph Display of Patient Profile in SAS | Yanwei Han | We usually create patient profile listings to present clinical study subject individual information. If we present most the information in one graph using SAS, it would be easier for reviewers to read them visually | study day, adverse events, serious adverse event, concomitant medications | source | 7 |
| 104 | Graphic Visualization for Treatment Switch Pattern | Anni Weng, Huanxue Zhou |
Visual report is a commonly used tool to present complex data by drafting understandable and digestible graphs. A well-designed visual report not only extracts key business insights from vast amounts of information, but also provides instant visual feedback to real-time changes. In this paper, we will introduce two elegant and easily-generated visual reports for treatment switch pattern. First is the web-based interactive tree diagram, which uses circles to represent treatments and length of links between two circles to represent time length of each treatment switch. The biggest advantage of this diagram is that it offers users complete control over displayed information as each circle can be collapsed or expanded. The second visual report is the web-based Sankey diagram, which adopts nodes as treatments and width of flows between two nodes as quantities switched from one treatment to another. This diagram enables users to easily perceive the transition trend over time | proc step, proc json | source | 9 |
| 105 | Great Time to Learn GTL | Richann J. Watson, Kriss Harris |
It's a Great Time to Learn GTL! Do you want to be more confident when producing GTL graphs? Do you want to know how to layer your graphs using the OVERLAY layout and build upon your graphs using multiple LAYOUT statement? This paper guides you through the GTL fundamentals! | bar chart, macro variable, proc template, proc step | source | 12 |
| 106 | Hiring the best SAS programmer among today's high competitive market | Jian Hua (Daniel) Huang, Li Joy Wang, Mary Evanchick |
Hiring a new SAS programmer could be time consumi ng and challenging, espe cially in today's high competitive job market. Being a programmer manager for 4.5 years, I have led a fast-growing team starting from 5 people to 20 people globally, and I would like to share my successful experience on how to hire top SAS programmers | sas programmer, sample | source | 4 |
| 107 | How PK and PD analyses drive dataset designs | David Radtke, Brian A. Moser |
Historically, the format of pharmacokinetic analysis datasets has been driven by the nature of time-based evaluations and the pharmacokinetic analysis programs used for these analyses. The format of these analysis datasets differs, sometimes remarkably, from those used for common statistical analyses of clinical data. Perhaps not surprisingly, attempts to standardize data collection and reporting have focused on meeting the clinical safety and efficacy analysis needs of statisticians, and have largely overlooked the corresponding structure and format of data for pharmacokinetic, pharmacodynamic, and pharmacometric (PK, PD, and PMx) evaluations. Given the increasing importance of PK, PD, and PMx analyses for dose selection and optimization en route to product registration, the need to develop an understanding of data collection, compilation, and reporting for PK, PD, and PMx analyses has never been greater | adam dataset, pd, clinical research, data science, data standard, cdisc | source | 7 |
| 108 | How to automate validation with Pinnacle 21 command line interface and SAS | Aleksey Vinokurov | Pinnacle 21 software is widely used by the industry for helping prepare data in a standardized format that is required for regulatory submissions. Pinnacle 21 Command Line Interface (CLI) allows automation of validation jobs resulting in higher reliability of results and better overall performance of data preparation process. This presentation will review new capabilities and syntax of the recently updated version of Pinnacle 21 CLI and provide examples for automating validation using SAS programming language | data cleaning, sysrc | source | 9 |
| 109 | How to Be a Great (Stats or Programming) Manager: Advice from an Underling | Mary Grovesteen | Programmers and statisticians aren't well-known for being social butterflies. Most of us will tell you we prefer working with computers or numbers rather than people. In the contract research organization (CRO) world, however, we often find ourselves having to work with many different people, not just within our own company but also with clients and vendors. This environment can be challenging for introverts who are just starting a new job but even more so for introverts who find themselves in a managerial position. A manager may be a great programmer or statistician, but this doesn't mean he or she will also be a great manager. Additional skills are needed to bring a department or team together and help everyone, from the lowest ranks on up, to succeed | data science, clinical trial | source | 6 |
| 110 | How to Continue to Use SAS System Viewer in SAS 94 | Jeff Xia | SAS System Viewer 9.1 is an application that SAS programmers use to check the contents of SAS related documents. Compared with later versions of similar applications, such as System Universal Viewer or SAS Enterprise Guide (EG), SAS System Viewer can open a dataset in a much faster speed due to its lightweight design. However, SAS has stopped official support for SAS System Viewer since SAS 9.2 was released. SAS programmers can still use SAS System Viewer to open SAS datasets in SAS 9.2 and 9.3, but not any more in SAS 9.4. Many SAS programmers asked in different internet forums how to continue to use SAS System Viewer in latest version of SAS. So far no satisfactory answers have been provided. This paper presents a work around solution for SAS programmers to continue using the lightweight SAS System Viewer even in SAS 9.4. It provides a much more efficient tool to help SAS programmers in their daily life at work | sas programmer, dataset, sas enterprise guide | source | 5 |
| 111 | IF there is a better way than IF-THEN | Bob Tian, Anni Weng |
In this paper, the author compares different methods for implementing piecewise constant functions (step functions) in SAS. The author uses a simulated approach to measure the efficiencies of different | proc sql, histogram, in=, proc step | source | 8 |
| 112 | Implementation of ADaM Basic Data Structure on Genetic Variation Data for Pharmacogenomics Studies | Linghui Zhang | Genetic variations have essential impacts on drug exposure and response. Pharmacogenomics (PGx), deciphering the drug-gene relationship, have been widely applied in drug design, discovery, development, and labeling. To facilitate scientific progress in the field of PGx and the use of PGx data in drug development, FDA officially issued a few guidelines to assist pharmaceutical sponsors engaged in evaluating the role of genetic variations on drug response and to encourage voluntary PGx data submission. To that end, the Study Data Tabulation Model Implementation Guide: Pharmacogenomics/Genetics (SDTMIG-PGx) was published in 2015 to provide guidance on the implementation of SDTM PGx domains. However, there is no official guidance on implementation of Analysis Data Module (ADaM) on PGx data. There are even fewer discussions on ADaM implementation of PGx data. Due to the complicated nature of PGx data, it is not only technical but also scientific challenge to fit PGx data in ADaM structure | data science, adam dataset, clinical trial, sdtm domain, sample, efficacy summary, bds, proc princomp, sample size, scree plot, cdisc | source | 10 |
| 113 | Implementation of Data Cut Off in Analysis of Clinical Trials | Mei Dey | Interim analysis can result in key decisions on early stopping for futility, continuation of the trial or early declaration of trial success. Data from an interim analysis may form the basis for a regulatory submission in the case of early declaration of trial success. Data cut off (DCO) is primarily utilized in interim analysis. In oncology trials, overall survival (OS) and progression free survival (PFS) are key efficacy endpoints. Where patient survival is paramount, the correct application of DCO methodology has a major impact on trial interpretation. Not applying or incorrectly applying DCO methods can negatively affect the trial result and has the potential to turn a positive trial into a negative one. In this paper, we explore the application of DCO principles as illustrated in its specific application in oncology trials. We will look at the operational considerations and how to manage operational and programmin g challenges when applying DCO for an on-going trial | raw data, adverse events, ctcae grade, adam dataset, sdtm domain, overall survival, sample, concomitant medications, data cutoff, interim analysis, clinical trial, data science, cdisc | source | 15 |
| 114 | Implementing the Rank-Preserving Structural Failure Time Model in SAS and R | Bradford J. Danner, Indrani Sarkar |
In randomized two-arm trials where survival is an endpoint of interest, patients may be required to stay on the assigned treatment until loss of clinical benefit or death. However, crossover from a reference to an active treatment is often allowed in oncology trials, after disease progression. Traditional intent -to-treat analysis can lead to underestimation of the treatment effect on overall survival when crossover is not taken into consideration. One possible method to minimize this estimation bias is the application of the rank-preserving structural failure time (RPSFT) model. A key feature of RPSFT model is to use a multiplicative factor to estimate the gain/loss in survival time from switching treatments contributed by the active treatment. While SAS PROC PHREG may be commonly used to perform a conventional Cox regression to estimate the hazard ratio between two treatment arms, an option for the RPSFT is not built - in |
treatment arm, proc phreg, overall survival, regression analysis, hazard ratio, randomization, treatment effect, proc step | source | 8 |
| 115 | Improved Transparency in Key Operational Decisions in Real World Evidence | Rebecca Levin, Irene Cosmatos |
The joint International Society for Pharmacoepidemiology and the International Society for Pharmacoepidemiology and Outcomes Research ( ISPE-ISPOR) Special Task Force on Real World Evidence in Health Care Decision Making recently published a paper that emphasized the need to improve the reporting of study details in research conducted using longitudinal healthcare databases.1 Lack of transparency in the operational details of published studies has led to "controversies over apparent discrepancies in results, and reduced confidence in evidence generated from healthcare databases." Transparency in large healthcare databases is dependent on several factors, including clarity regarding (1) pre-processing of source data, (2) operational decisions to create an analytic dataset, and (3) analytic choices. If independent investigators apply identical operational choices to the identical source data as the original researchers, similar if not identical results should be reproduced | sensitivity analysis, adverse events, sample size, sample | source | 11 |
| 116 | Improving Listing Generation with ODS EXCEL | Stephanie Ann Sanchez | The ODS destination for EXCEL simplifies data listing generation while enabling attractive features like sheet protection and AutoFilter to simplify team reviews in a format which is familiar and accessible to non-statistical reviewers. With SAS 9.4M4, now you can create native EXCEL files without any post- processing steps, display full titles and footnotes including multiple justifications, and print page numbers and titles on every page, all while reducing the amount of programming required to fit the table to the contents. This paper describes how to utilize the features of the ODS destination for EXCEL to create submission-ready data listings and will provide you with simple workarounds for common compatibility issues |
excel, footnote statement | source | 4 |
| 117 | Improving Metadata Compliance and Assessing Quality Metrics with a Standards Library | Veena Nataraj, Erica Davis |
Establishing internal Data Standards helps companies ensure a consistent interpretation of CDISC standards in submissions to health authorities around the globe. This presentation discusses our experience, various methodologies and solutions ta ken to improve the quality of SDTM metadata standards | controlled terminology, define-xml, sdtm domain, trial arms, data cleaning, cdisc | source | 9 |
| 118 | Incorporating Pinnacle 21 with LSAF | Aleksey Vinokurov, Sandeep Juneja, Sonali Garg |
Clinical studies data need to adhere to CDISC standards for submission to Regulatory agencies and to check if data is compliant to these standards, Pinnacle21 (P21) checks seem to be the de facto standard widely used by companies. Life Science Analytics Framework (LSAF) provides ability to have 21 CFR part 11 compliant web based Statistical Computing Environment (SCE) where studies data can be stored securely with versioning and audit history capabilities. However, since data had to be downloaded and moved manually to P21 site to execute compliance checks, traceability of the data was at risk. This paper talks about steps taken to automate integration between LSAF and P21, thereby eliminating manual data movement process. Using P21 client, P21 validator can be invoked from LSAF directly which handles the data movement and execution of the checks and results report of the checks are checked into LSAF Environment thereby maintaining full traceability and integrity of the data | data cleaning, controlled terminology, life sciences, excel, cdisc | source | 6 |
| 119 | Infrastructure Designed to Maximize Workflow | Paul D. Hamilton | All clinical statistical programmers work within a common set of constraints: we must support the production of Clinical Study Reports after a trial is complete; we must support filings to various regulatory agencies; we must operate under the aegis of CFR 21 Part 11; we must contend with frequent changes in specifications and contracting timelines; and most of us use SAS software in our work. This paper presents an architecture purporting to maximize programmer workflow, minimize manual steps, and optimize programmer efficiency. Guiding principles include: single source of metadata, stored locally; all processes driven by metadata and automated through software; hands-off production of finished publishable products; and minimizing human-only tasks wherever possible | proc step, dataset, macro call, work library, data cleaning, statistical programmer, database lock, macro variable | source | 9 |
| 120 | Infrastructure for Healthcare Analytics: A Foundational Approach to Development | Jennifer R. Popovic | Healthcare-related organizations and the data they generate are notoriously siloed, with varied systems, structures, data standards, and governance processes and policies. Research teams and organizations analyzing these data can be similarly siloed and challenged in their ability to create robust analytic infrastructure that would enable the sharing of their most valuable analytic assets: data and the expertise, methods, approaches and tools to analyze them. These challenges often lead to inefficient analy tic development based on one-off approaches that do not prioritize or promote the ideas of reusability, flexibility, scalability or reproducibility. This paper discusses the concept of analytic infrastructure and will discuss strategies that research teams and organizations can use in their approaches to develop infrastructure | health insurance, insurance claims | source | 11 |
| 121 | Insights to Meticulous Clean Patient Tracking via Analytics | Siddartha Kondapally, Kishore Pothuri, Bhavin Busa |
The Conventional ways of tracking clean patients by Clinical Data Management is by extracting data from multiple sources or generating reports manually from EDC system into EXCEL spread sheets. Even though this was proven to be successful in the past, still it is a tedious and laborious method along with a high risk of making an error while handling the data. The modern and most efficient way of tracking clean patients is by using Statistic al Analysis Software (SAS) in combination with Visual Analytics. This approach helps not only the Clinical Data Management, but also the Medical Monitors and Clinical Operations to overcome the challenges in data cleaning, tracking and reviewing the patients. Additional features and functionality can be added through customizations |
excel, data visualization | source | 4 |
| 122 | Instant Gratification: Calculating Resource Needs Based on Operational Data using an Interactive Tool | Melissa Hill, Baker Bill |
In SAS Visual Analytics, we leveraged existing operational metrics and a few cleverly placed user - supplied parameters to create a custom, interactive and user-friendly calculator for our clinical operations team, which extrapolates required monitoring days based on current electronic data capture (EDC) system status | data visualization, sample | source | 9 |
| 123 | Integrations Made Easier | Kirsty R. Lauderdale | Too often we are found to have an integration of a number of studies that potentially adds a few gray hairs for each of us. The documentation for data sources is missing, every study is a different standard, controlled terminology doesn't line up; the studies were so far in the past no one is quite sure what was done for the original, figuring out who actually was a roll-over subject, etc. The list can go on and on. Do not despair! There may be some tips that will help you to hold off on some of those grey hairs a bit longer! This paper will present some tested and true tips for an integration with fewer headaches. Some of these tips consist of starting out with an integrated SDTM which combines study level SDTMs, particularly when dealing with lots of studies. If prepared well early, study level ADaMs could also be combined as another useful option | controlled terminology, do over, cdisc | source | 6 |
| 124 | Interactive Programming Using Task in SAS Studio | Suwen Li | SAS Studio is a web browser-based application with visual point-and-click user interface tasks that generate SAS code. Users can create their own tasks using the common task model and the Velocity Template Language. This paper illustrates how to conduct interactive programming by creating customized tasks in SAS Studio | kaplan-meier plot, hazard ratio, swimmer plot, proc phreg | source | 7 |
| 125 | Introducing Dynamic Data Quality Control Methods: Abstracting Data Validation through External Data | Troy Martin Hughes | Data quality is a cri tical component of data governance and describes the accuracy, validity, completeness , and consistency of data. Data accuracy can be difficult to assess, as it requires a comparison of data to the real -world constructs being abstracted. But other characteristics of data quality can be readily assessed when provided a clear expectation of data elements, records, fields, tables, and their respective relationships. Data dictionaries represent a common method to enumerate these expectations and help answer the question What should my data look like? Too often, however, data dictionaries are conceptualized as static artifacts that only describe data. This text introduces dynamic data dictionaries that instead prescribe business rules against which SAS data sets are automatically assessed, and from which dynamic, data-driven, color-coded exception reports are automatically generated | data dictionary, quality control, data validation rules, data cleaning, variable, excel, if-then-else, xls, dataset, macro definition, proc import, proc format, proc step, macro variable | source | 20 |
| 126 | Introducing Interactive Data Step Debugger: What you can do with SAS Data Step Debugger (SAS Enterprise Guide 7.13) | Kevin Kim | Fall of 2016, the Interactive DATA STEP debugger has been released (SAS Enterprise Guide 7.13). Debugging is the process of removing logic errors from a program. Unlike syntax errors, logic errors do not stop a program from running. Instead, they cause the program to produce unexpected results. While the SAS log can help you identify data errors, the DATA STEP debugger offers you an easier, interactive way to identify logic errors, and sometimes data errors, in DATA steps. You want to have this debugging tool |
sas enterprise guide, proc step | source | 9 |
| 127 | It's all about getting the Source and Codelist Implementation right for ADaM Define.xml v2.0 | Supriya Davuluri | There are some obvious challenges in generating A nalysis Data Model define.xml v2.0 with confusion prevailing over populating appropriate source for ADaM v ariables and implementation of c odelists, which often causes lack of structural consistency. The ADaM M etadata team proposed conventions for populating attribute Type of element source in a consistent manner to distinguish between Predecessor, Assigned and Derived along with some guidelines for implementation of Control Terminology (CT) i.e. NCI/CDISC or User specified CT for ADaM Variables [3]. This paper summarizes the best practices to be followed while populating the variable source, reflecting its occurrence in the given ADaM datasets along with the CT implementation considerations such as usage of Enumerated Lists vs Codelists, CT naming conventions, dealing with Subsets of CTs and specifying CT for a v ariable to set up the metadata correctly when generating an ADaM define.xml v2.0 | adam dataset, controlled terminology, cdisc controlled terminology, char., cdisc | source | 11 |
| 128 | Know Thy Data: Techniques for Data Exploration | Charu Shankar | Get to know the #1 rule for data specialists: Know thy data. What are the common keys for joins? Are there data type conflicts? How to locate changed variable names? How to reorder variables in the dataset without physically typing in all the names but instead using metadata to perform this action. In this session, you will learn to employ powerful PROC SQL's dictionary tables to easily explore aspects of your metadata |
dictionary tables, proc step, proc sql, proc print, proc contents, cpu time, variable list | source | 9 |
| 129 | Know your PET: From the Scans to SDTM Datasets | Balavenkata Pitchuka, Bhargav Koduru |
PET with the glucose analog FDG as a tracer is a mature and increasingly available clinical imaging technique used for diagnosing, staging, and detecting the recurrence of many cancers. US FDA and EMEA have approved FDG and authorized it as a diagnostic radiopharmaceutical in the diagnosis of infection. FDG-PET has rapidly gained importance in Lymphoma Clinical Trials with the 2014 Lugano Classification and more recent LYRIC. irRECIST and iRECIST, the latest response criteria for immunotherapies in solid tumors are also becoming more reliant on FDG-PET for assessing inflammation. The role of FDG-PET has grown from a qualitative positive or negative response to a quantitative response with the use of Deauville five-point scale (2009) in assessment of treatment response in Hodgkin Lymphoma and certain types of non-Hodgkin lymphomas. Deauville score calculation involves consideration of various individual scores: SUVmax target lesion, SUVmeanbackground | sdtm domain, complete response, cdisc | source | 8 |
| 130 | Let's Get FREQy with our Statistics: Data-Driven Approach to Determining Appropriate Test Statistic | Lynn Mullins, Richann J. Watson |
As programmers, we are often asked to program statistical analysis procedures to run against the data. Sometimes the specifications we are given by the statisticians outline which statistical procedures to run. But other times, the statistical procedures to use need to be data dependent. To run these procedures based on the results of previous procedures' output requires a little more preplanning and programming. We present a macro that dynamically determines which statistical procedure to run based on previous procedure output. The user can specify parameters (for example, fshchi, plttwo, catrnd, bimain, and bicomp), and the macro returns counts, percents, and the appropriate p-value for Chi-Square versus Fisher Exact, and the p-value for Trend and Binomial CI, if applicable | macro parameter, proc freq, binomial distribution, proc step, sample, sas output, class variable | source | 14 |
| 131 | Let's get to the Source and streamline it to the End | Jennifer Price | Esource is simply the use of the original source data as collected, without requiring transcription. Using source data provides many advantages, but also challenges. I will review successful implementations, challenges, TransCelerate, SCDM and CDISC activities and best practices | raw data, edc, clinical trial, cdisc | source | 6 |
| 132 | Let's Start Something New! A Beginner's Guide to Programming in the SAS CAS Environment | Kevin Russell, Gerry Nelson |
Starting something new (whether it is a home project or a work project) is not always the easiest thing to do. However, more often than not, starting something new can be very rewarding. So while you might be a little hesitant to start programming with SAS Cloud Analytic Services (CAS), you will likely find it extremely beneficial! CAS is a server that provides a cloud-based, run-time environment for analytics and data management. The server can run as a distributed server on multiple machi nes with multithreaded processing, which means that it can perform extremely high-performance analytics against very large tables. The purpose of this paper is to help you take that first step toward programming with CAS. This paper provides a complete example of programming in the CAS environment, from showing you how to establish a CAS session to running one of the many high-performance procedures that are available in CAS | proc step, sas studio, sas viya, by group processing, cas action, cas, libname statement, proc freq, frequency analysis, sas programmer, put statement, excel, regression analysis | source | 15 |
| 133 | Leveraging Standards for Effective Visualization of Early Efficacy in Clinical Trial Oncology Studies | Kelci J. Miclaus, Lili Li |
Solid tumor oncology studies have become a major therapeutic focus area and present unique challenges in both quantifying and categorizing efficacy. The related domains (TR/TU/RS) implemented by CDISC SDTM provide a rich, sophisticated data structure to accommodate the complexities of multiple lesion response analysis. The published rules by Response Evaluation Criteria in Solid Tumors (RECIST) to define patient overall response further enable standard programming and analysis in such clinical trials. In this presentation we will demonstrate how these industry standards can be leveraged to create report dashboards for progression-free survival analysis (PFS) and target lesion response analysis using effective visualization to target early efficacy endpoints. In performing lesion response analyses, we will focus on application of graphical techniques common to this therapeutic area; including Swimmer Plots, Waterfall Plots, Forest Plots and Spider Plots | waterfall chart, swimmer plot, clinical trial, disease response, change from baseline, forest plot, tumor results, percent change from baseline, sample, best response, sample size, objective response rate, hazard ratio, survival analysis, bar chart, clinical research, data standard, progression-free survival, pd, adverse events, cdisc | source | 12 |
| 134 | Levista - An Integrated Programming Environment with Visualization | Priya Lakshmi Saradha | Today's industry process to generate ADaM datasets is tediously manual dependent heavily on spreadsheets for ADaM specifications and manually written SAS programs. Enormous amount of time is spent in the creation of ADaM specifications as the author switch between various meta -data tabs and documents. As often times, addition of new variables, new codelists, and value -level metadata, during the course of a study is error prone, and unfortunately these errors are identified close to database lock. To address these and various other issues, at LEVS TAT, we utilized a single -screen interface to implement data visualization and standardization. Our platform enables user to visualize the input data, codelist and value level metadata on the same screen | metadata management, data cleaning, cdisc | source | 4 |
| 135 | Link Up Sync Up: Calculation of concordance and discordance rates between Independent Review Facility (IRF) and Investigator site data using SAS | Bhargav Koduru, Girish Kankipati |
Efficacy data , one of the important components of clinical trial s, determines the effectiveness of investigational drug. In some phase II and III clinical trials response data is collected from both Independent Review Facility (IRF) and investigative sites. IRF verifies the endpoints in clinical trials by independent review of imaging data . Radiological assessment of images at the sites are often subjective, resulting in high interobserver variability among radiologists viewing the same images. To overcome this variability the images are also reviewed by an IRF. Central review by an IRF allows an auditable, blinded, and uniform process of evaluation. This provides greater consistency across sites, patients, and time points, achieving higher confidence levels in the data. Often IRF and site response data are assessed in clinical trials, and it is a best practice to check the consistency between these two sources of response | confidence interval, clinical trial | source | 10 |
| 136 | Machine Learning - Why we should know and How it works | Kevin Lee | The most popular buzz word nowadays in the technology world is "Machine Learning (ML)." Most economists and business experts foresee Machine Learning changing every aspect of our lives in the next 10 years through automating and optimizing processes such as: self-driving vehicles; online recommendation on Netflix and Amazon; fraud detection in banks; image and video recognition; natural language processing; question answering machines (e.g., IBM Watson); and many more. This is leading many organizations to seek experts who can implement Machine Learning into their businesses. Statistical programmers and statisticians in the pharmaceutical industry are in very interesting positions. We have very similar backgrounds as Machine Learning experts, such as programming, statistics, and data expertise, thus embodying the essential technical skill sets needed. This similarity leads many individuals to ask us about Machine Learning | machine learning, pharmaceutical industry, neural network, regression analysis, data science, statistical programmer, supervised learning, support vector machine, unsupervised learning, proc reg, logistic regression, sample, logistic distribution | source | 8 |
| 137 | Macro-Supported Metadata-Driven Process for Mapping SDTM VISIT and VISITNUM | Pragathi Dayananda, Eric Crockett |
Programming SDTM VISIT and VISITNUM for a specific domain is often a simple task. Programming SDTM VISIT and VISITNUM consistently across all domains and multiple input data sources is a much bigger task. While it is tempting to generate visit mapping code on the fly, time spent designing a generalized visit-mapping process will save time and money, and better guarantee a conformant, accurate result. Companies seeking to establish a general solution for SDTM visit mapping often base the process on metadata that associates raw visit content with SDTM VISIT and VISITNUM. When properly designed, visit mapping metadata provides a single input that can be used to assign visits for all raw data sources. Where things can get tricky is that the merge strategy to assign visits using the mapping metadata often varies across raw data sets. Two macros will be shared, including examples and code, which facilitate the use of visit mapping metadata to assign SDTM visits | raw data, laboratory data, cdisc | source | 6 |
| 138 | Making Processes Accessible | Arthur Collins | Establishing clear, well defined processes is critical to maintaining quality, compliance, and efficiency in the work we do in Biometrics. Our formal processes often consist of only high-level SOPs which lack broad buy-in, are disseminated by the read and understand method, and are difficult to access for reference when a process is being performed. This paper will present an approach to developing robust processes and documenting, organizing, presenting, and maintaining comprehensive documentation of those processes, including all relevant forms, templates, and training material, in a single easy to access portal. The portal is structured as an end-to-end roadmap, visually representing the spectrum of activities carried out by each function with links to detailed process maps for specific activities which are owned by functional SMEs. In our organization, this approach has led to greater buy-in and compliance and has created a culture of continuous improvement | source | 5 | |
| 139 | Making the most of SAS Jobs in LSAF | Gregory T. Weber, Sonali Garg |
SAS Life Science Analytics Framework (LSAF) provides the abilit y to have a 21 CFR Part 11 compliant web-based Statistical Computing Environment (SCE) where studies data can be stored securely with versioning and audit history capabilities. Jobs are a critical part of an LSAF environment as programs must be run via jobs. Starting in version 4.7.1, the Run and Populat e feature is a new and improved way of creating and running jobs. This paper will present our experien ces with creating jobs prior to 4.7.1, the issues we encountered and how Run and Populate resolves these c oncerns and helps us write efficient jobs reducing the time and resources needed to get our results | 21 cfr part 11, proc contents | source | 5 |
| 140 | Managing Table Preparation and Related Coding Task in Multilingual Settings by Using SAS Unicode Version | Kai Koo | To apply the approval of drugs and medical devices outside the US, it is common that all the supporting documents, including statistical analysis tables and listings of trial results should use or be translated to the official languages of specific countries. Although SAS provides localized versions to support statistical analyses and table outputs in different languages, the SAS codes containing specific character encodings may display correctly in one SAS language version, but fail to work normally in other localized versions. To simplify the code maintenance and output table management in multilingual settings, SAS Unicode version, which can handle most of the writing system, is a favorable solution to produce similar tables and listings with different languages in single code. Furthermore, the SAS codes written in different localized versions can be translated from different character encoding systems to Unicode programmatically through the SAS system option - ENCODING | options statement, txt, sample | source | 6 |
| 141 | Manipulating Statistical and Other Procedure Output to Get the Results That You Need | Vincent DelGobbo | Many scientific and academic journals require that statistical tables be in a specific format (for example, the American Psychological Association [APA] style). This paper shows you how to change the output of any SAS procedure to conform to any journal or corporate style. You'll learn how to save data from any SAS procedure, change the data format using the DATA STEP, and dynamically create a format based on your data. You'll also use the SAS Output Delivery System (ODS) inline formatting functions and style overrides, and produce several output formats, including HTML, RTF (for Microsoft Word), PDF, and Microsoft EXCEL files. Finally, you'll learn how to create and deliver the output on-demand using SAS server technology. This paper is appropriate for all SAS skill levels |
dataset, proc step, least squares mean, proc report, proc mixed, macro variable, sas output, excel, proc means, repeated measures, ods output, output delivery system, treatment effect, repeated measures model, character variable, output, sample, proc format, ods excel | source | 37 |
| 142 | Merge with Caution: How to Avoid Common Problems when Combining SAS Datasets | Joshua M. Horstman | Although merging is one of the most frequently performed operations when manipulating SAS datasets, there are many problems which can occur, some of which can be rather subtle. This paper examines several common issues, provides examples to illustrate what can go wrong and why, and discusses best practices to avoid unintended consequences when merging | merge statement, by group processing, in=, out=, set statement, dataset, if-then-else, subsetting if, sas log, pdv, proc sql, length statement | source | 16 |
| 143 | Mind Over Matter: Building an Army of Passionate Soldiers Dedicated to Saving Patients' Lives | Shefalica Chand | The clinical research industry is one of the fastest paced industries, where sponsors are continuously introducing newer and better products to improve patients' lives. SAS Programmers are expected to deliver speedy analysis results, with high standards of quality and accuracy. In this extremely demanding and competitive culture, managers are responsible for developing and maintaining a positive mindset and attitude within their teams, towards a common goal of benefitting patients. The demand and expectations of our profession can be comparable to that of the soldiery, due to the importance given to saving human lives. Passion, selflessness, integrity, professionalism, seamless quality, efficiency, timeliness, expert insight, and teamwork are just a few common traits | regulatory submission, clinical research | source | 10 |
| 144 | Mining Mass Fragments using SAS and Python for Metabolite Identification of Antibody-Drug Conjugates | Hao Sun, Kristen Cardinal |
Data analysis, visualization and reporting of preclinical and clinical drug metabolism assays were automated with a SAS-based application, AIR Binder (PharmaSUG 2017, MWSUG 2017). Data mining using SAS in combination with Python was also explored for a specific dataset type in drug metabolism, the mass spectrometry data for metabolite identification and characterization. We developed an application, LEO, to generate solutions for relatively challenging metabolite characterization of peptides and antibody-drug conjugates, which was built on an algorithm to search accurate mass change using three main components: linearization for cleavage reactions, extension for conjugation reactions, and oxidation for oxidation-reduction reactions (SASGF 2018). We further extended applications of LEO to a wider chemical space, for metabolite identification of both small and non-small molecules | data science, data mining | source | 5 |
| 145 | Multicollinearity: What Is It and What Can We Do About It? | Deanna Naomi Schreiber-Gregory | Multicollinearity can be briefly described as the phenomenon in which two or more identified predic tor variables in a multiple regression model are highly correlated. The presence of this phenomenon can have a negative impact on the analysis as a whole and can severely limit the conclusions of the researc h study. This paper reviews and provides examples of the different ways in which multicollinearity can affect a research project, and tells how to detect multicollinearity and how to reduce it once it is found. In order to demonstrate the effects of multicollinearity and how to combat it, this paper explores the proposed techniques by using the Youth Risk Behavior Surveillance System data set. This paper is intended for any level of SAS user. This paper is also written to an audience with a bac kground in behavioral science or statistics | regression analysis, predictor variable, correlation, ridge regression, estimate, proc step, response variable, proc princomp, condition index, proc reg, sample, multiple regression, proc corr, proc freq | source | 13 |
| 146 | No News Is Good News: A Macro to Reveal the Untold Normal Lab Results | Ming Yan | In clinical trials, specific lab microscopic UA, RBC morphology, and special WBC subordinate tests are reported to sponsors by laboratory ONLY if an abnormality is observed. The normal results which are not explicitly reported, however, need to be in place in order to compute a percentage of abnormalities for each lab test in the subject population. This macro starts from SDTM LB domain that only stores observed abnormal lab tests, distinguishes the unreported normal test results from tests that are not performed, and creates ADaM dataset with all the normal results filled in for all patients at all time -point that the lab tests were performed. By doing the imputation at ADaM level, this macro avoids repetitive data processing in table programming and makes table generation simpler and faster | clinical trial, adam dataset, cdisc | source | 7 |
| 147 | ODS/RTF Pagination Revisit | Ya Huang, Bryan Callahan |
ODS/RTF combined with PROC REPORT has been used to generate publication quality TFL by many people, and yet one problem has been persistently bothering them for years: Pagination of listings. When a listing has columns that hold long text (concatenated AE terms, Lab comments etc.), depending on the width of the columns and the length of the text, one line could wrap into multiple lines. Where the line wraps is decided by Microsoft Word rather than SAS. This leads to a nightmare of page break control. This paper presents an easy way to detect the number of lines needed when line wrapping occurs |
proc report, txt, proc step, protocol deviations | source | 8 |
| 148 | One Level PDF Bookmark Created by ODS DOCUMENT and PROC DOCUMENT | Yanmei Zhang, Titania Dumas-Roberson |
There are several publications on how to create PDF files with multiple levels of bookmarks. Sometimes it is not convenient for reviewers to review contents with multiple levels of bookmarks. Very often there is a need to create PDF files with only one level of bookmarks which are clear and easy for reviewers to review documents. This paper presents how to meet this need using PROC REPORT without a BY statement or how to create one level of bookmarks using ODS DOCUMENT and PROC DOCUMENT when there is BY statement in PROC REPORT |
proc report, by group processing, proc step | source | 6 |
| 149 | One Project, Two Teams: The Unblind Leading the Blind Part 2: Options | Kristen Reece Harrington | In the pharmaceutical world, there are instances where two independent programming teams exist to ensure blinded treatments and data are maintained by the appropriate parties. For the purposes of this discussion, blinded corresponds to scrambled data and unblinded corresponds to actual data. Within these projects, blinded programmers use scrambled data to create programs which produce both blinded and unblinded outputs. To ensure the blind is maintained and ethics are upheld, only the blinded programming team produces and modifies the programs. While robust programming is key, there are multiple techniques that can be used to ensure that the blinded programs remain unedited during the unblinded runs. We will discuss the pros and cons of two techniques: utilizing %INCLUDE statements and making files read-only programmatically |
macro variable, do loop | source | 6 |
| 150 | Parallel Processing Your Way to Faster Software and a Big Fat Bonus: Demonstrations in Base SAS | Troy Martin Hughes | SAS software and especially e xtract-transform-load (ETL) systems commonly include components that must be serialized due to real process dependencies. For example, a transform module often cannot begin until the data extraction completes, and a corresponding load module cannot begin until the data transformation completes. While process dependencies such as these cannot be avoided in many cases and necessitate serialized software design, in other cases, programs or data can be distributed across two or more SAS sessions to be processed in parallel , facilitating significantly faster software . This text introduces the concept of false dependencies, in which software is serialized by design rather than necessity , thus needlessly increasing execution time and deprecating performance | sas base, macro variable, parallel processing, proc freq, proc corr | source | 8 |
| 151 | Patient-Level Longitudinal Analysis Plots Showing Adverse Event Timelines and Dose Titration Levels | Keith J. Brown, John R. Gerlach |
The analyses in a clinical trial often include subject-specific longitudinal plots that depict, for example, a lab measurement taken at various times during a study period. Because the study involves dose titration levels, the plots might include vertical lines indicating dose levels, which can increase and decrease during the study period. Perhaps the analysis includes another component: the timelines of adverse events, if any. Both the vertical reference lines and horizontal time lines are superimposed on the longitudinal plot. Obviously, the task becomes more intricate since each plot is subject-specific, having different measurements, dose titration levels, and adverse events. This paper explains how to produce this intricate graphical analysis | adverse events, study day, dose, laboratory data, proc sgplot, proc step, proc sql, macro variable, ordinal variable, proc format, where statement, proc gplot | source | 15 |
| 152 | Pinching Off Your SAS Log: Adapting from Loquacious to Laconic Logs To Facilitate Near-Real Time Log Parsing, Performance Analysis, and Dynamic Data-Driven Design and Optimization | Troy Martin Hughes | Too often in SAS literature, the role of the program log is narrowly conceptualized as a static, post hoc quality control review that validates program success through the detection of program failure or the lack thereof . Especially when software development occurs outside of a formalized software development life cycle (SDLC), as is often the case with non-production software and within end-user development environments, SAS practitioners must painfully parse logs in search of notes, warnings, runtime errors, and other often elusive indications of functional failure. A substantial body of SAS literature advances antediluvian manual log review through the automation of log parsing and analysis and subsequent communication of program success (or failure) to stakeholders. To a lesser extent, SAS literature depicts how log parsing can be utilized to extract, analyze, and ultimately improve software performance metrics | key performance indicator, sas log, proc sort, cpu time, stress testing, macro variable, if-then-else, repeated measures, repeated measures anova, txt, macro definition, sample, proc sql, character variable, data cleaning, exception handling, sysrc, numeric variable, macro loop, parallel processing, proc printto, quality control, options statement, proc step | source | 26 |
| 153 | PK Data Explained | Timothy J. Harrington | Pharmacokinetics, or PK, is the monitoring of the concentration level over time of an analyte within a human (or animal) body. The analyte is typically an active ingredient of a drug, or an element such as sodium or potassium, or a chemical compound which occurs naturally in blood or other body fluids, such as creatinine or bilirubin. PK data collection for an active drug compound is performed routinely in all clinical trials since it is a critical part of measuring the safety and efficacy of an experimental treatment and in determining the treatment dose amount, frequency, and time- release profile | sample, key, pk parameter, character variable, adverse events, dose, imputation, time variable, proc freq, clinical trial | source | 9 |
| 154 | Preparing the Office of Scientific Investigations (OSI) Requests for Submissions to FDA | Zihui Lin, Wei Cui, Yaling Teng, Ran Li |
The FDA Office of Scientific Investigations (OSI) requests that three items (listed below) be provided for all major trials that support the safety and efficacy in a marketing application. This information is used by the FDA OSI to support the identification of clinical sites to be inspected. I. General Study-Related Information and Comprehensive Clinical Investigator Information II. Subject Level Data Listings by Site III. Site Level Dataset (also referred to as bioresearch monitoring (BIMO) dataset) This paper covers the authors' detailed understanding of these OSI requests and current available guidance from the FDA. Experiences and challenges from recent submissions for planning and creating the Item I and II listings and the BIMO data package will be shared. In addition, programming macros developed for improved efficiency and consistency of the Item II listings will be described | clinical trial, clinical programming, concomitant medications, per protocol, phase ii, adverse events | source | 9 |
| 155 | Preparing to Meet FDA Requirements for Submission of Standardized Data and Documentation | Steven R. Kirby, Mario Widel |
PDUFA V gave the FDA the authority to require electronic submission of study data in standard format. That authority was confirmed by PDUFA VI. We will provide a practical overview of FDA expectations for submission of standardized study data and associated documentation for statisticians and statistical programmers and share how those expectations can be met | character variable, controlled terminology, raw data, cdisc | source | 11 |
| 156 | Preventing premature unblinding in PK/PD related studies | Debbie Sisco, Alice Zong |
Blinding is commonly used in clinical research study designs and is one of the critical methodologic features intended to reduce the risk of bias allowing a realistic statistical comparison of the study outcomes. Premature unblinding carries the risk of invalidating the entire clinical study which would be a costly mistake. Unblinding brings special challenges for performing Pharmacokinetic(PK) and Pharmacodynamic(PD) related studies. During the various stages of a blinded PK/PD study, there are inherit high risks of accidental unblinding that may occur during study conduct, sample processing and bioanalysis, data analysis and reporting, or modeling and simulation. This presentation is designed to create an awareness of the challenges and identify operational procedures, processes and programming tips intended to prevent PK/PD related studies from premature unblinding | sample, clinical research, data wrangling, clinical trial, proc standard, clinical database, database lock, pk parameter, sdtm programming, data science, response variable, population pk, cdisc | source | 11 |
| 157 | Proactive Role of a Lead Programmer Could Help Prevent Potential Delay in Submission | Charan Vem, Ranjith Kalleda |
In general you see data issues as data is collected after database go live, but if you can think proactively about the potential data issues that could happen based on the Study design, CRF design, and the way DM collects the raw data, you could avoid major data issues which could delay submission. Also as a lead programmer if you get familiarize yourself with eSub components and their preparation, and your company specific process and SOPs you will save time | per protocol, raw data | source | 4 |
| 158 | Processing and Cleaning Streaming Data in SAS | Mehmet Kocak | The streaming data has become part of everyone's daily life with the common use of smart devices. It is also part of almost every clinical research protocol with the increasing use of smart monitoring systems such as e-scales, wearable blood pressure or heart rate monitors, etc. Although capturing and storing such data is a growing challenge, making such data available for statistical analysis is another challenge for statisticians and computer programmers. In this research, we propose an algorithm aiming to capturing the true profile in a given streaming data with a specific application to BodyTrace weight data captured in the Fit Blue study. We show that our approach is highly sensitive in identifying the correct profile and highly specific in cleaning the correct profile from other mixing profiles. We have implemented the proposed approach in a SAS macro program named %TPF (TrueProfileFinder) | data science, regression analysis | source | 10 |
| 159 | Programmer's Guide for OSI Deliverables - Creation of Site Level Summary Dataset and Automation of BIMO Listins Generation | Charanjit Kahlon, Kristie Kooken, Bhavin Busa, Dharmendra Tirumalasetti |
The FDA conducts site inspections as part of the regulatory review process to ensure clinical trial is conducted according to the FDA regulations. For the drug related application (i.e. NDA), site inspection is managed by Office of Scientific Investigations (OSI) within CDER office of compliance. The OSI requests that the Sponsor should submit data that describes the characteristics and outcomes of clinical investigations at the site level as part of the NDA. The part II of the OSI request, which is mandatory, is to submit subject-level data listings by site ("BIMO listings") for all major trials used to support safety and efficacy in the application. The requested BIMO listings for the most part is standard from one application to the other. The part III of the OSI request, which is optional, is to submit summary level clinical site dataset ('clinsite'). There is an FDA draft guidance and specifications document for preparing and submitting the dataset | adverse events, clinical trial, proc step, csr, adam dataset, concomitant medications, primary endpoint, cdisc | source | 10 |
| 160 | Programming Support for Anti-Drug Antibody Pharmacokinetics in Therapeutic Protein Drug Development | Linghui Zhang, Jiannan Kang |
Therapeutic Proteins (TP) are engineered proteins in laboratory for pharmaceutical use. In the last several decades, benefiting from recent advances in biotechnology, TP products have become an important class of medicines treating a wide spectrum of clinical indications. However, due to their high antigenic potential, TP drugs may trigger host immune responses resulting in varying clinical consequences for efficacy and safety, ranging from no apparent effect to life -threatening reactions. The anti-drug antibodies (ADA) generated in unwanted immune response may sustain, clear, or neutralize the TP drug. As the result, ADA can significantly influence pharmacokinetics (PK), pharmacodynamics (PD), bioavailability, and efficacy. Thus, it is critical to understand the essential role of ADA on PK/PD in drug development. Immunogenicity assessments and validations are also regulatory requirements to ensure the safety and efficacy of TP products | sample, key, clinical trial, clinical research, time variable, box plot, response variable | source | 11 |
| 161 | Purrfectly Fabulous Feline Functions | Louise S. Hadden | Explore the fabulous feline functions and calls available in SAS 9.1 and later. Using CAT functions and CAT CALLs gives you an easier way to streamline your SAS code and facilitate concatenation of character strings. So, leave verbose coding, myriad functions, and the vertical bar concatenation operators behind! SAS 9.2 (and beyond) enhancements will also be demonstrated | character functions, proc contents, sas log, cpu time, proc step, length statement, catq, character variable | source | 15 |
| 162 | Quickly create your own format with hundreds of values and other tips | Sergey Sian | Writing SAS code which automatically defines format(s) with hundreds of values, and using defined format for report is not new. However, this is sometimes a necessary and certainly useful tool to use | proc format, proc step | source | 8 |
| 163 | R 101: The base Package | Arthur X. Li | Statistical computing is employed within a diverse range of industries. In recent decades, an open source project, R, has emerged as the preeminent statistical computing platform. With its unsurpassed library of freely available packages, R is capable of addressing almost every statistical inference and data management problem. In this paper, you will learn the most commonly-used functions and operators from the R base Package, which serves as the fundamental tools for performing data management | missing value, nominal variable, txt, binary variable | source | 22 |
| 164 | Real World Patients: The intersection of Real World Evidence and Episode of Care Analytics | David Olaleye, Youngjin Park |
SAS Institute recently released two products, SAS Episode Analytics and SAS Real World Evidence, which enable healthcare organizations and pharmaceutical companies to leverage the power of real-world data and episode analytics to solve real-world health care questions. Insights gained from constructed episodes can be used to measure cost and quality of care, evaluate care pathways, and monitor efficiency of healthcare resources utilization. Further analyses of the episodes can be undertaken by exporting the episodes data into SAS Real World Evidence. This paper examines two business use cases that highlight the episode analytic capabilities of SAS Episode Analytics and SAS Real World Evidence. We also develop two risk adjustment models - an episode cost prediction model and a readmission likelihood model for patients in the diabetes and COPD episode cohorts | risk adjustment, likelihood function | source | 15 |
| 165 | Reflections on a Career in Clinical/Statistical/SAS Programming (35+ Years in the Making) | Alan Meier | Since 1982, technology, standards, submissions and environments have evolved. I've gone through mergers, job changes and re-organizations. Worked as a programmer, moved into management and returned to programming. Along the way, the 2 lessons that I learned were to be flexible and have a plan | pharmaceutical industry, data cleaning, sas programming | source | 6 |
| 166 | Regulatory Compliance in the digital age - A cloud-based statistical computing environment (SCE) to the rescue | Sridhar Vijendra, Tyagrajan Swaminathan |
The present-day processes in statistical programming for clinical trials are going through a major change, where geographically distributed teams located in multiple time zones across the globe work collaboratively under increasingly stringent data quality, integrity, security and privacy regulations that call for full traceability, data lineage, audit trails and adherence to data standards. To address these, our organization has recently deployed a modern web-based cloud enabled Statistical Computing Environment (SCE). This paper will describe the experience of our organization in migrating to a cloud-based SCE. Migrating from a conventional in-house system to a cloud-based SCE was a big step but its many long-term benefits far out-weighed the initial efforts in setup and training. At an enterprise level, a cloud-based SCE establishes a shared environment with a single step login where sponsors and collaborators can work together | statistical programmer, audit trail, access control, raw data, excel | source | 9 |
| 167 | Report Compatibility: System to evaluate your code against volatile sources. | Joseph Cooney, Randy Santiano, Craig Brelage |
Developing programmed reports and queries via SQL, SAS, or R Code may take several cycles of development and resources. Although there is value with the final product, additional return on investment is obtained when all or portions of the report's code can be reused. However, variation in how data is collected across the clinical trial industry can reduce your organization's ability to reuse programmed reports and queries. This issue can be addressed by applying five basic principles to ensure that your organization has the ability to evaluate programmed report/query compatibility. This will allow for a proactive evaluation for the report code's reusability, and it will also support an environment where the users and developers are given transparency towards new specification development, allow for better workload management, and foster the ability for more accurate report/query catalog planning | data cleaning, crf, adverse events, clinical trial, medical history, proc sql | source | 12 |
| 168 | Review programs used for submission and their input to the define-XML 2.0 | Ari Siggaard Knoph, Morten Hasselstrm Jensen, Bo Sølgaard Andersen |
In recent years the regulatory authorities have been requiring programs for generation of ADaM data and programs for main endpoint analysis as part of the submission package. Generally in the pharmaceutical industry, analysis programs used for clinical trial reports make use of internal sponsor-defined metadata repositories and highly complex macros to facilitate e.g. the generation of several outputs. Therefore, it has become gradually clear, that the authorities demands high readability, more transparency and more traceability in analysis programs for submission as well as in the related analysis results metadata (ARM) | sample, data science, adam dataset, dataset, where statement, clinical trial, excel, cdisc | source | 11 |
| 169 | Risk-based Validation in Clinical Trial Reporting: Focus on What Matters Most | Amber Randall, William Coar |
Quality control is fundamental to ensuring both correct results and sound interpretation of clinical trial data. Most validation procedures are a function of regulatory requirements, industry standards, and corporate philosophies. However, in the current environment of lean work forces and increased workload, the traditional approach of 100% independent programming may no longer be feasible. Further, we would argue that full double programming may not be the best approach for maintaining high quality. Instead, we propose the implementation of a risk-based validation system in which the validation method is chosen individually for each piece of a project. Risk assessment is based on matrix of factors and the outcome guides the assignment of a testing methodology appropriate for that particular risk level. This ensures that higher effort and rigor is focused in areas that matter most | data cleaning, clinical trial, independent qc, insurance risk score, adverse events, concomitant medications | source | 10 |
| 170 | SAS automation techniques - specification driven programming for Lab CTC grade derivation and more | Yuqin Li | SAS programmers often face challenges on SAS coding, for example, Lab CTC grade derivation per lab test, putting variable's attribute into data sets - this can be very time consuming if we do this kind of work manually on coding, since the process repeats many times by test or data sets. Here I introduce a way for SAS coding in that SAS programming techniques take care of these repeated work and achieve automation, as long as a consistent pattern can be identified in the coding logic | sas programming, sas enterprise guide | source | 6 |
| 171 | SAS Macros of Performing Look-Ahead and Look-Back Reads | Yanhong Liu | When working with the time series data, it is generally required use of previous observation(s) (Look - Back) or next observation(s) (Look-Ahead) for certain variable, in order to do some comparison or calculation. The SAS Lag function can be used to get the Look-Back reads, and we can use the Double SET statement to perform the Look-Ahead reads. If you have SAS/ETS installed, you can use PROC EXPAND as an alternative. In this paper, a few of macros are introduced to obtain one or more Look -back or Look-ahead observations with by-groups by implementing the above techniques. Additionally, this paper briefly compares the cost of the DATA STEP Macros to the PROC EXPAND Macros |
proc step, proc sort, sample, lag function, set statement, time series analysis | source | 10 |
| 172 | SASsy Data Checks | Amie Bissonett | Every study has data issues and those issues often make their way to the data sets that the statistical programmers use. It is a part of our job to do data checks, and those same types of tools are invaluable when validating data sets. Following are a few useful tips and tricks to make data checking a breeze | proc step, proc compare, proc sql | source | 4 |
| 173 | Satisfied customers leave, but raving fans stay... | Hrideep Antony, Aman Bahl |
In today's world of fierce competition, your ability to deliver extraordinary customer service will be key to setting your business apart from your competitors. It is no longer enough to "merely" satisfy your customers. "Merely satisfied" customers will eventually leave, and take their business elsewhere. To stay ahead of the game, you must turn your "merely satisfied" customers into "raving fans". Your "raving fans" will be the life-line of your business. Your "raving fans" will ultimately be your loyal customer base, and will stay with you, regardless of one poor experience, or better options, or a couple of business mistakes. It is vital that the primary focus of your business be on customer loyalty, and not just satisfaction. This paper will explain how to turn your "merely satisfied" customers into "raving fans". It will also show the positive impact your "raving fans" will have on the ultimate success of your business in today's competitive world | source | 3 | |
| 174 | SDTM EX and EC: Considerations When Submitting Exposure Data | Jerry Salyers, Kristin Kelly |
Per the SDTMIG v3.2, the Exposure (EX) domain should contain only unblinded data for study treatment that was administered to subjects presented in protocol-specified units. Because of this, the Exposure as Collected (EC) domain can be used to collect blinded data during the study for all administrations given, not taken or missed. After the study is unblinded, the EX domain can be derived from EC. The SDTMIG recommends submitting both EX and EC if necessary, though including EC is not required. Sometimes it can be difficult to determine when both domains are needed to provide the full story of how subjects were exposed to treatment. This paper will focus on use cases where submitting EX is sufficient as well as instances when a reviewer would benefit from receiving both EX and EC | dose, clinical data, cdisc | source | 8 |
| 175 | SDTMs in a Medical Device Trial - A First Attempt | Phil Hall | While clinical trial data of pharmaceutical and medical devices have a lot in common, the differences can be difficult to deal with in SDTMs. This paper acknowledges the similarities, highlights the differences and serves as a case- study of how the Transcatheter Heart Valve (THV) Statistical Programming group of Edwards Lifesciences has decided to map their raw data into Study Data Tabulation Model (SDTM) domains. This is not intended to be the definitive guide of how to map medical device trial data to SDTM standards but to serve as the opening statement of a conversation | clinical trial, adverse events, clinical data, sdtmig, clinical programming, cdisc | source | 6 |
| 176 | Second Primary Malignancy (SPM) Analyses in a Disease Registry Study | Jane Lu, Shankar Srinivasan, Robert Knight |
In hematology clinical trials, monitoring for SPMs plays an important role in the assessment of safety for new agents in drug development. The correlation of exposure to the experimental drug to the development of an SPM has become an important analysis in the reporting of clinical trials. This presentation reports the SPM analyses performed in a US registry trial in which study physicians chose the treatment regimen administered to patients with newly diagnosed multiple myeloma. SPMs were monitored carefully, and a specific SPM case report form was filled out at each study visit. SPMs were categorized as hematologic, solid tumor, or non-melanoma skin cancers. Patients are followed for at least 5 years. Analyses included incidence, incidence rates (SPMs per person-years), time-to-event, and risk factor assessments. We used the first SPM event start date to calculate drug exposure | clinical trial, multivariate analysis, clinical database, univariate analysis | source | 6 |
| 177 | Seeing the Forest for the Trees: Part Deux of Defensive Coding by Example | Donna E. Levy, Nancy Brucken |
As statisticians and programmers, SAS is part of our daily life. Through assessing patterns, data quality, programming datasets, analysis displays or developing simulations, we need to determine the best ways to conduct our daily work, allowing us to see the forest for the trees. This paper provides guidance on quality defensive programming, efficient coding as well as good programming concepts. Programming no no's will also be discussed. The concepts discussed will allow us to navigate through the trees --- that is, seeing the trees for the forest. We may have been programming in SAS for weeks, months, years or decades. Regardless, we should continue to expand our skills and continue learning and updating our techniques. With this paper, we will provide reminders for paths lost in the past, as well as new tips to help us clear the brush from the trail | proc step, in=, keep statement, proc genmod, proc sql, out=, sas programmer, temporary dataset, cpu time, raw data | source | 10 |
| 178 | Sense and Censorability: Learn censoring techniques with ADTTE for your survival | Shilpakala Vasudevan | In many studies in clinical trials, the time to event is the primary endpoint, with an event being either an adverse event, death or recurrence of a disease. Time to event is calculated as the time from a defined starting point to the time of occurrence of the event of interest. In some cases, the event might not occur within the duration of study. An interesting feature of survival data is that at the end of the observation period, the event of interest may not have occurred for all subjects, or the exact event time may be unknown. In these cases, censoring rules need to be applied to determine the time to event. Depending on the nature of the study, ADTTE can accommodate single events or composite events, with binary or multiple censoring. This paper will present how the endpoints of these subjects are accounted for, by applying censoring rules | overall survival, hazard ratio, survival analysis, clinical trial, primary endpoint, swimmer plot, right censoring, confidence interval, interval censoring, randomization, data science, regression analysis, adverse events | source | 12 |
| 179 | Set-up and usage ways of Jupyter Notebook for SAS | Alona Bulana | The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text. It has support for over 40 programming languages, including SAS. The interactivity of the notebook implemented via so-called kernels. SAS Software released the SAS kernel for Jupyter in September 2016 which is open source. The Notebook itself looks like a chain of cells which may contain either code or Markdown. Code cells can be evaluated and SAS Output or Log will be stored immediately after cell as an inline result. Cells with Markdown are desired for explanations or any other kind documentation. You may ask what are advantages of using a Jupyter Notebook instead of just a SAS code with comments? The answer is that the notebook stores computed results in it which can be recomputed if needed and makes the results completely reproducible | sas studio, sas output, sas programming | source | 11 |
| 180 | Shiny Happy People: Using RShiny and SDTM Data to generate a Quick Interactive Dashboard | Saranya Duraisamy, Nate Mockler |
This workshop will show how to use the "shiny" package in R and SDTM data to create an open -source, interactive and modular dashboard that can take your SDTM data and create insights that can saved time and much less work. A basic understanding of R and SDTM data is recommended | cdisc | source | 12 |
| 181 | Some Common Programming Errors and Possible Solutions That Could Impact a Successful NDA/BLA | Amos Shu | Common errors from the statistical programming function are formatting, traceability, uniformity, completeness of datasets, missing variables, inconsistent data cutoff dates, missed or poor definitions of variables, non-functioning SAS statistical programs, un-executable software, inadequate and/or incorrect annotations within the annotated CRFs, miscalculation, misconducts, etc. All types of errors could not only delay the NDA/BLA approval, even possibly destroy the entire trial, but also hurt financially by take additional time and efforts. This paper illustrates several types of errors from the process of SDTM, ADaM, and tables generation that happened in recent years to help us rethink about the programming strategies and ensure data quality for our NDAs/BLAs | adam dataset, clinical trial, statistical programmer, data governance, data science, clinical programming, data cleaning, cdisc | source | 5 |
| 182 | Speed Up SDTM/ADaM Dataset Development with If-Less Programming | Lei Zhang, John Lu |
One of major tasks in building FDA submission packages for a clinical study is to create SDTM/ADaM datasets from CDISC compliant specifications. Most of programmers will find this task tedious, time-consuming, and prone to error because the variables in the specifications often have to be derived from different datasets with complex if-else statements, and updated, and tested with ever evolving datasets. In order to alleviate this dreary process, this paper introduces the if-less programming technique that helps programmers to create SDTM/ADaM variables with fewer or no if-else statements. By applying this technique, a programmer can not only write pieces of derivation code shorter and quicker, but also make them easier to understand, modify, and reuse. What's more, this technique will greatly reduces the programming time to create ASDTM/ADaM datasets and at the same improves the quality of SDTM/ADaM datasets generated | if-then-else, adam dataset, when, data step, imputation, sample, clinical trial, cdisc | source | 17 |
| 183 | Square Peg, Square Hole-Getting Tables to Fit on Slides in the ODS Destination for PowerPoint | Jane Eslinger | An output table is a square. A slide in Microsoft PowerPoint is a square. The table, being the smaller square, should fit in the bigger square slide. Right? Well, not always. Despite the programmer's expectations, some tables will not fit on the slide created by the ODS destination for PowerPoint. It depends on the table. For instance, tables with, say, more than 10 rows or more than 6 columns might end up spanning multiple slides. But, just as with the popular children's toy, by twisting, turning, or approaching the hole from a different angle, you can get the peg in the hole |
proc step, proc report, proc freq, proc tabulate, proc genmod, proc glm, class variable, proc print | source | 18 |
| 184 | Standardization Efforts at Medtronic, its benefits and Lessons Learned | Firdouse Fathima | A standard data model is defined as a data model that is widely applied in some industry and shared amongst competitors per Wikipedia. Data Standards in our industry can be defined as a framework including structures, vocabularies, and business processes that enables the efficient conduct of a clinical trial. In the context of this paper, standardization refers to creation and build of standard CRF's, database per CDISC standards up until the creation of SDTM Datasets. Standardization helps organizations by increasing efficiency, quality, generates savings and eventually helps us submit our data in compliance with regulatory agencies requirements such as US FDA. It is becoming commonplace that regulatory agencies in United States, Europe, and Japan are requ iring organizations to submit their data in a CDISC compliant format. This is especially true for Pharmaceutical industries and similar requests are expected to reach medical device industry soon | crf, sdtm domain, cdisc | source | 4 |
| 185 | Statistical Analysis and Visualization of Microbiome data in Clinical Trials | Athira Sudhakaran, Limna Salim |
Radical changes across Microbiome research in Clinical Trials , has materialized with the widespread application of next-generation sequencing technologies. This is empowered by high-throughput profiling of the genetic contents of microbial communities. With the advent of big data technologies and high throughput computing resources, large complex microbiome data sets can be easily analyzed. Research is ongoing to identify new methods and models suitable for analyzing microbiome data. Some of the areas that need attention includes a. Analysis of microbiome data and tools to explore its composition b. Longitudinal data analysis c. Causation Analysis Since data is extremely complex, it remains a challenge for statisticians to develop tools required for such analysis. This paper explores the possibilities of developing a dynamic software framework using Angular JS, SAS, R, Python and Rasa NLU for easy and effective analysis of microbiome data sets | sample, cluster analysis, proc princomp | source | 8 |
| 186 | Statistical programmer application framework with R/Shiny | Ashok Gunuganti | This paper presents a basic framework for developing R/Shiny applications to accomplish various tasks as a better alternative than doing those tasks in SAS. The primary advantage of Apps developed in R/shiny over base SAS is that the applications can be published to the Web. They are also more dynamic, interactive, and are able to incorporate the versatile R packages into Shiny which makes them very powerful. Since Shiny applications can be deployed to the Web, the end user can use them with zero knowledge of R | dataset, by group processing | source | 9 |
| 187 | Statistical Programming: Revolution, Evolution, or Status Quo? | Jeffrey M. Abolafia, Frank C. DiIorio |
Opinions are plentiful about the radical transformation of statistical programming in the pharmaceutical industry in the past 20 years. They usually contend that programmers must become conversant in languages other than SAS, acquire data science skills, and improve communication and other "soft" skills. But is all that really necessary? If not, what are the implications for programmers, managers, and their organizations? This paper reviews the evolution of both the pharma environment and the role of the stat programmer, focusing on the skill sets required for submission deliverables. It also summarizes interviews with managers in pharmaceutical companies and CROs, focusing on how their technology and deliverables have changed | statistical programmer, data cleaning, adam dataset, pharmaceutical industry, clinical programming, data science, sas enterprise guide, raw data, sas grid, cdisc | source | 9 |
| 188 | Statistician in my Soul. Statistical Questions from SPA | Nadiia Hryhorenko | Many times, statistical programmers cannot implement deep analyses of clinical data without advanced statistical knowledge, and answers to their questions may be too simple to bother the statisticians for help. In the article we address a few of the statistical issues programmers often face. Mastering these methods allows deeper understanding of the clinical data and may even help in detecting inconsistencies in the data analysis | confidence interval, sample, normal distribution, proc step, response variable, null hypothesis, clinical trial, parametric test, anova, binomial distribution, confidence level, survival analysis, kruskal-wallis test, proc freq, normality test, clinical data, efficacy summary, kaplan-meier estimator, proc ttest, heteroscedasticity, proc anova | source | 12 |
| 189 | Stay ahead of the game: Know the SDTM compliance check reject messages before creating your SDTM datasets | Lixiang Larry Liu, Yuqin Li |
Beginning October 2016, Pharmaceuticals and Medical Devices Agency (PMDA) has started accepting electronic study data submissions with a 3.5 year transitional period. Even though PMDA's requirements are very similar to those of U.S. Food and Drug Administration (FDA), there are some differences between these two regulatory agencies. One of the key differences is the severity level for both the SDTM and ADaM validation rules. While FDA only has the error and warning messages, PMDA also has the reject severity level, and has clearly stated that "Rules which, if violated, will cause the review to be suspended until corrections are made." Because the drug development is done globally, and generally, the same submission package will go to both FDA and PMDA, the study team should avoid all the reject messages in both SDTM and ADaM submission packages. This paper provides a comprehensive review of the nine reject SDTM validation rules per PMDA severity level | data validation rules, data cleaning, sdtm domain, clinical trial, controlled terminology, cdisc | source | 6 |
| 190 | Summary Table for Displaying Results of a Logistic Regression Analysis | Lori S. Parsons | When performing a logistic regression analysis (LR) for a study with the LOGISTIC procedure, analysts often want to summarize the results of the analysis in a compact table. A combination of SAS DATA STEP, SAS macro, and SAS Output Delivery System code is presented here as a method to create and display a summary LR table. The summary table includes the odds ratio estimate (OR), 95% confidence interval (CI), and P value for each covariate included in the LR model. The reference group (Ref) of each covariate included in the model is indicated in the table. The total number of observations (N) for each covariate level is also included in the table. Additionally, a table footnote is generated, by macro variable code, to describe the total number of observations read and the total number of observations used in the LR model |
logistic distribution, p-value, macro variable, logistic regression, binary variable, medical history, proc step, regression analysis, proc logistic, proc report | source | 9 |
| 191 | TECHNIQUES FOR GLOBAL, ADAPTABLE, AND EFFICIENT PROGRAMMING | Sunil Kumar Pusarla | SAS programmers in the pharmaceutical industry have to perform many common tasks across different study programs. Re-usability of the code we produce goes a long way to make our lives easier. Many situations arise when we need to modify our own code. In such times, rather than inspecting each line of a program for statements which need modification, we can constrain our attention to one easily identified location, which affects all subsequent code automatically. There are many simple yet effective techniques that will increase coding productivity and produce highly robust programs. This paper addresses some of those techniques: legacy procedures like FORMAT; new enhancements to procedures like PROC SUMMARY; obscure functions and options; as well as more recent resources such as PROC FCMP |
proc step, proc format, proc fcmp, proc freq, time variable | source | 8 |
| 192 | Telling Stories with Jupyter Notebook | Kevin Lee | The Jupyter Notebook is an open-source web application that allows programmers and data scientists to create and share documents that contain live code, visualizations and narrative text. Jupyter Notebook is one of most popular tool for data visualization and machine learning, and it is the perfect tool for story telling tool for data scientist. First, the paper will start with the introduction of Jupyter Notebook and why it is the most popular tool for data scientist to show, share and visualize the data and analysis. The paper will show how data scientist uses Python programming language in Jupyter Notebook. The paper will show how data scientists import data into Jupyter Notebook using Pandas. The paper will introduce Python data visualization library, matplotlib, and show how data scientists use matplotlib to easily create scatter plot, line, histograms, Kaplan Meier curves and many more | data science, machine learning, data visualization, dataset, adam dataset, histogram, cdisc | source | 14 |
| 193 | The Applications of Tolerance Intervals: Make it Easy | Marina Komaroff | It is known that Confidence Intervals (CI) cover a certain population parameter with a stated confide nce; however, there is a lso a way to cover a fixed proportion of the population for a characteristic of interest with a specified confidence and such interval is named Tolerance Interval (TI). The application of T I(s) in clinical trials has been used for defining drug response with biomarker(s). Transdermal medicine that requires evaluation of stability and drug release to adhesive transfer for the evaluation quality of transdermal patches relies on calculation of TIs . SAS Institute provides a way to calculate TIs with PROC CAPABILITY but requires the SAS/QC package. However, most users do not need nor can afford this entire package for industrial engineering when the need for tolerance intervals can be rare and unpredictable. That means that researchers need to write their own SAS code |
normal distribution, standard deviation, macro definition | source | 9 |
| 194 | The Changing Clinical Data Analytics Landscape - Visualization & Beyond | Siddartha Kondapally, Kishore Pothuri, Bhavin Busa |
"A picture is worth a thousand words" is a commonly used phrase in every industry where there is data involved! In the clinical industry, plotting data and presenting statistical information using graphics is not new. Until most recently, the traditional approach was to "look" at the static output without giving end users an ability to explore and drill-down further. With the rapidly increasing availability of data visualization and analytic tools, the industry landscape is shifting. End users now want to 'see' their data more interactively, identify trends, visualize the patient profiles and review results at a high-level while still being able to drill -down to get a complete picture. In this paper, we will talk about how we have used the power of SAS and TIBCO Spotfire, to build our "out-of-the-box" dashboards to meet these demands. We will discuss our experiences, processes, case scenarios, and challenges towards implementing such a solution | data review, data visualization, data science, clinical trial, raw data | source | 6 |
| 195 | The Fundamentals of Macro Quoting Functions | Arthur X. Li | The Macro quoting functions is one of the most difficult concepts to grasp when one learns to write a macro program. One of the reasons for its complexity is closely related to understanding the macro processing, such as macro compilation and execution phases. Furthermore, the situations for utilizing macro quoting functions are not obvious for programmers. In this talk, in addition to reviewing the macro processing, we will examine the most useful macro quoting functions via various examples | macro variable, sas log, macro quoting function, character functions, macro function, macro call, macro put, substr function, execution phase, if-then-else, macro facility, &variable, missing value, %if-%then-%else | source | 10 |
| 196 | The Next Generation "Smart Program Repository" | Hrideep Antony, Aman Bahl |
The term repository sounds very "routine", but what if it can do the job of both a "project manager" and a "SAS programmer"! That is right! Introducing the smart repository that will create initial SAS Programs, generate status updates, track project progress, create audit trails, and check program logs, all with the click of a few buttons! While this utility provides a structured approach to programming and increases the programming efficiency, it also ensures the compliance of each program, making them audit ready. Combined features of EXCEL VBA and macros are used to build this utility |
excel, data cleaning | source | 8 |
| 197 | The Statistical Programming Summit | Todd Case | On September 14, 2017 eighteen Statistical Programming leaders from twelve Bio/Pharma companies in the Boston area met and discussed topics endemic to work we are all involved with, including: CDISC standards, Quality Control (QC) and Quality Assurance (QA) of Statistical Programing deliverables, Metadata-driven solutions (e.g. specifications to generate data and define-xml) and Submission Experience | clinical programming, quality assurance, cdisc | source | 4 |
| 198 | Tips and Fixes for Cross-Environment Batch Transfer of SAS Data | Yun (Julie) Zhuo | Contract research organizations (CROs) frequently receive data transfers from pharmaceutical companies or other clinical organizations. A batch process is commonly used as we need to transfer a number of files multiple times. A smooth transfer is the first key step to a successful project. This paper cautions against some common errors during the batch process, and it presents tips on resolving these issues. The tips include selecting the correct conversion method, avoiding data truncation or data loss, converting data set encodings, and building a quality process. Code samples are provided. The scope of this paper is limited to the conversion of SAS data sets and/or transport files at the receiving end of the transfer process. The primary audience includes programmers, statisticians, and data managers responsible for receiving and converting data | raw data, error message, proc step, character variable, proc copy, proc cport, data cleaning, proc contents, adverse events, proc cimport | source | 10 |
| 199 | Tips and Tricks for Producing Time-Series Cohort Data | Nathaniel Derby | Time-Series cohort data are needed for many applications where we're tracking performance over time. We show some simple coding techniques for effectively and efficiently producing time-series cohort data with SAS | macro variable, macro function, macro definition | source | 6 |
| 200 | Tips and Tricks to Enhance RTF Output | Rohini Rao | PROC REPORT has long been a powerful and flexible tool in clinical programming for producing tables and listings. The ODS TAGSETS.RTF destination was introduced in SAS version 9.2, enabling programmers to produce higher quality output with less effort. This paper demonstrates a few tips and tricks to enhance the æsthetics of the reports. Topics covered include: the use of decimal tabs to properly align numerical output the usage of Unicode to insert special characters the deletion of excess blank lines at the top and the bottom of a page the insertion of page-specific footnotes * the repetition of a header row when the body of the section spans multiple pages. SAS version 9.2 on the Windows platform is used in this paper. It is targeted for an audience with an intermediate level of SAS and ODS knowledge |
adverse events, clinical trial, proc step | source | 8 |
| 201 | To Error is Human: An Overview of Human Errors in SAS Programming and How to Mitigate Them | Nagadip Rao | Until a few years back, Clinical SAS programming, used to be done by a team of programmers who would write programs based on the specifications provided by study statistician or lead programmer. However, in recent years, this has changed, programming task now involves coordination between multiple teams of programmers belonging to different organizations, spread across the globe. The complexity of programming tasks, along with work being done by multiple outsourced service providers, increases the risk of human errors especially, if the pharmaceutical client is a large organization | clinical trial, adverse events, laboratory data, clinical sas, service provider, adam dataset, data cleaning, adverse event severity, data science, clinical programming, cdisc | source | 9 |
| 202 | Traceability: Some Thoughts and Examples for ADaM Needs | Richann J. Watson, Kent Letourneau, Wayne (Weizhen) Zhong, Sandra Minjoe |
One of the fundamental principles of ADaM is that datasets and associated metadata must include traceability as a link between analysis results, ADaM datasets, and SDTM datasets. The existing ADaM documents contain some examples of simple traceability, such as variable derivations and inclusion of the SDTM sequence number, but what about more complex examples? An ADaM sub-team is currently developing a Traceability Examples Document, showing how traceability can be employed in a wide variety of practical scenarios. Some of these examples contain content from other CDISC documents, modified to focus on the traceability aspects. Others are being developed specifically for the Traceability Examples Document. As members of the Traceability Examples ADaM sub-team, we are including in this PharmaSUG paper and presentation a selection of examples that demonstrate the power of traceability in complex analyses | adam dataset, sdtm domain, data science, adverse events, cdisc | source | 18 |
| 203 | Trial Summary: The Golden Gate towards a Successful Submission | Bhargav Koduru, Girish Kankipati |
Trial Summary (TS) has gained importance in the recent submissions due to specific guidelines introduced in the recent versions of FDA's Study Data Technical Conformance Guide (latest version was released in Oct 2017). As of December 2016, it is required to use UNII, NDF-RT and SNOMED CT in NDAs, ANDAs and certain BLAs (FDA Data Standards Catalog v4.10 updated in Oct 2017), while entering the input for certain TSPARMCDs like TRT, CURTRT, COMPTRT, and PCLAS etc. In addition, failure to include TS dataset in the submission package would result in a technical rejection with a high severity. With the limited description provided for the TSPARMCDs in the CDISC Controlled Terminology (CT), it is often not clear how to derive certain CT. This can lead to improper derivation, which jeopardizes the entire submission. For example, NDF-RT houses the values for all the following TSPARMCDs PCLAS, TRT and CURTRT | trial summary, clinical trial, cdisc | source | 9 |
| 204 | Triangulating Multiple Sources of Contradictory Prescription Data - a Real-World Case Study | Vidhya Parameswaran | Effective clinical decision -making relies on summarizing information from reliable, evidence- based research. With the increasing need for real-world evidence generation, scientists emphasize on developing a comprehensive understanding of their data to ensure generalizability of the observed results. Data triangulation is an efficient information integration technique through which multiple sources of overlapping data are combined to validate and strengthen the quality of the datasets. Merging data from mult iple sources representing the same set of variables may give rise to contradictory values, and this may present challenges in the interpretation of the results. In this paper, I use SAS to study the benefits of triangulating three different sources of prescription information in a retrospective, observational study setting and develop an algorithm to identify and resolve data inconsistencies | sample, clinical trial | source | 13 |
| 205 | Tumor Dashboard | Sudhir Singh | Strategic management of oncology trials requires clinical study team to synthesize range of patient and disease information from multiple sources. Monitoring tumor progression is very important in oncology trials. Traditionally we used to generate several reports for scientist to monitor the data. Our team decided to leverage data analytics tools to create interactive dashboard that enables our researcher to get quick insight of patient tumor time course data. The interactive dashboard available on intranet displays collected data about patient enrollment, PFS/OS events, non target and target lesion, swimmer plot with an aim to facilitate both data exploration and data quality check by summarizing tumor information for each patient or group of patients. It displays the Sum of Longest Diameters (SLD), the Target Lesions (TL) and RECIST criteria, such as Overall Response (OR), Best Overall Response (BOR) and the appearance of new lesions | swimmer plot, clinical trial | source | 6 |
| 206 | Useful Adverse Events (AE) data Diagnostics and Summarization | Abhinav Srivastva | Safety remains a critical concern in all Clinical trials conducted by the sponsors. Among many aspects to look at for the safety of the patients such as Vital Signs (VS), Laboratory samples (LB) , etc, Adverse events (including Serious Adverse events) play a key role in addressing the safety of the drug or biologic under investigation. A typical AE Case Report Form (CRF) c ollects AE verbatim name, start and stop dates, toxicity, relatedness, action taken, seriousness etc. The paper focusses on data checks that can be done on the AE data and present some useful summarization that can be run as part of routine monitoring reports or submissions. All the results will be reported graphically for easy display using SAS Enterprise Guide 7.1 | adverse events, sample, sas enterprise guide | source | 6 |
| 207 | User-defined Functions for Processing Lab Data | Richard R. Allen | Lab data is often messy with multiple units given for the same parameter. Converting these to standard units can be challenging. Usually macros are used for these conversions. This presentation shows a method for converting labs using a user-defined function written using PROC FCMP. The procedure builds functions using DATA STEP code that can be stored and reused for multiple studies. A method for grading the toxicity of labs using a function will also be presented |
proc fcmp, proc step, laboratory data, excel, sample, if-then-else, sas function, adverse events | source | 13 |
| 208 | Using Agile Analytics for Data Discovery in the Pharma Industry | Bob Matsey | Submission Title: Integrating Your Analytics In Database with SAS, Hadoop and the Teradata EDW with cover how cutting edge companies are using Agile Analytics and Data Labs to combine and analyze many different data types in Teradata EDW | data science, excel | source | 4 |
| 209 | Using JMP Scripting Language (JSL) to create an application interface for automated analysis of Lab Data in Clinical Trial | Dongsun Cao | Laboratory data is the key element of the safety profile in drug development. Due to its complexity, it often represents a great challenge to conduct an efficient and effective review. JMP has been proven to be a powerful interactive visualization tool in analyzing clinical data. This paper describes the development of JMP Add-in Lab Analyzer using JSL, which takes ADaM datasets as input and allows interactive visualization and customized analysis of Laboratory data. The analysis approaches are based on the FDA reviewer guidance on lab data, including measures of central tendency, outliers or shifts from normal to abnormal, and marked outliers as well as exploratory lab data analysis such as Hy's Law | clinical trial, laboratory data, change from baseline, treatment period, outlier, scatter plot, shift table, box plot, central tendency, clinical research, line chart, adverse events, cdisc | source | 11 |
| 210 | Using ODS LAYOUT, GTL, and ODSTEXT to Generate a Compact Graphical Codebook | Shane Rosanbalm | In this paper we present a SAS macro that will produce a compact graphical PDF summary of all of the variables in a SAS dataset. The report is considered compact because each variable summary is confined to a 1" tall strip. The report is considered graphical because, in addition to presenting metadata and statistics for each variable, the report also includes a small graph of the distribution for each variable: a histogram for continuous variables and a bar chart for categorical variables. Not to be mistaken for a data dictionary, this compact graphical report is designed to allow the user to quickly gain familiarity with the variables in a dataset | data dictionary, nominal variable, interval variable | source | 4 |
| 211 | Using SAS GTL to Visualize Data When there is Too Much of It to Visualize | Nathaniel Derby, Perry Watts |
In many of the SAS Institute publications about ODS statistical graphics, there is an introductory sta tement that defines an "effective" graph as one that reveals "p atterns, differences and uncertainty that are not r eadily apparent in tabular output" (Kuhfeld, 2010 ; Rodriguez, 2008 ; Rodriguez and Cartier, 2009 ). Good graphs are also said to "provoke questions that stimulate deeper investigation, and ... add visual clarity and rich content to reports and presentations." Developing a good graph becomes a challenge, howeve r, when input data map to crowded displays with ove rlapping points or lines. Such is the case with the Framingh am Heart Study of 5209 subjects captured in the sas help.heart data set and a series of 100 cumulative booking cur ves for the airline industry. In addition, interlea ving time series plots can be difficult to interpret, and patterns c an be missed when lattice plot panels are charted o ut-of-order |
proc step, scatter plot, histogram, series plot, time series analysis, proc template, proc freq, contour plot, order by, missing value, proc sgpanel, graph template language, proc univariate, proc means, nominal variable, cumulative frequency, line chart, numeric variable, proc sgplot | source | 22 |
| 212 | Using Simulation to Create Enrollment Plans | Jim Box | Developing a plan to reach a Study Enrollment goal can be a tricky business. Choosing countries and sites with their associated rates and costs complicates things. Simulation is an excellent way to understand sources of variability and comparing options to meet the enrollment goals | clinical trial, sample | source | 7 |
| 213 | Using the SAS ODS Excel Destination Options to Enhance Your Excel Output | William E. Benjamin Jr. | The new SAS ODS EXCEL Destination is a feature rich addition to the SAS ODS tool set. In this paper I will show you how to output multiple spreadsheets, add some style options and place your data where you want it. I will use the ID option, the STYLE option, and the START _AT sub-option of the SAS ODS EXCEL Destination. The ID option allows you to define different output EXCEL Workbooks that use the same SAS Code but use different output options when writing an EXCEL Workbook. I will show you how to create three workbooks at once using this ODS option. The STYLE option can enhance your output workbooks, I will show you how to use to the STYLE option to enhance the visual display of your output data. Finally I will show you how the put your output data elements into any row and column in the new worksheet. All of this can be done while you execute your SAS code only once |
excel, xls, ods excel, ods pdf, output delivery system | source | 11 |
| 214 | UTF What? A Guide for Handling SAS Transcoding Errors with UTF-8 Encoded Data | Michael Stackhouse, Lavanya Pogula |
The acronyms SBCS, DBCS, or MBCS (i.e. single, double, and multi-byte character sets) mean nothing to most statistical programmers. Many do not concern themselves with the encoding of their data, but what happens when data encoding causes SAS to error? The errors produced by SAS and some common workarounds for the issue may not answer important questions about what the issue was or how it was handled. Though encoding issues can apply to any set of data, this presentation will be geared towards an SDTM submission. Additionally, a common origin of the transcoding error is UTF-8 encoded source data, whose use is rising in popularity by database providers, making it likely that this error will continue to appear with greater frequency. Therefore, the ultimate goal of this paper will be to provide guidance on how to obtain fully compliant SDTM data with the intent of submission to the FDA from source datasets provided natively in UTF-8 encoding | proc step, proc print, character variable, length statement, raw data, proc contents, flat file, dataset, permanent dataset, proc options, cdisc | source | 15 |
| 215 | Utilize Dummy Variables/Datasets in Graph Generation | Amos Shu | The input SAS dataset does not always perfectly contain all necessary data that can be used for graph generation in the real world. Creating dummy variables or datasets is an effective way, sometimes the only way, to generate the desired graph. This paper discusses three different ways to utilize dummy variables/datasets in SAS visualization | binary variable, adverse events, in= | source | 4 |
| 216 | V is for Venn Diagrams | Kriss Harris | Would you like to produce Venn diagrams easily? This paper shows how you can produce stunning two -, three-, and four-way Venn diagrams by using the Graph Template Language, in particular the DRAWOVAL and DRAWTEXT statements. From my experience, Venn diagrams have typically been created in the pharmaceutical industry by using Microsoft EXCEL and Microsoft PowerPoint. EXCEL is used to first count the numbers in each group, and PowerPoint is used to generate the two- or three-way Venn diagrams. The four-way Venn diagram is largely unheard of. When someone is brave enough to tackle it manually, working out the numbers that should go in each of the 16 groups and entering the r ight number into the right group is usually done nervously! |
macro variable, macro call, graph template language, %if-%then-%else, excel, macro definition | source | 22 |
| 217 | Validating R - Part of the Uphill Battle in the Pharmaceutical Industry | Peter Schaefer, Debra Fontana |
R, the open source language for statistical computing, is widely accepted by data scientists across many application areas. However, several myths around R still seem to prevent its acceptance by the pharmaceutical industry. In this paper, we make another attempt to disprove the 'R can't be validated' thesis and show the path to a validated R system. As a bonus, we will describe a concept for a powerful validated system for data analysis and reporting. Born as beta version infant in 2000, R, the open source programming langua ge for statistical computing, has made an astonishing career: It has grown in popularity by academia and industrial users to surpass other languages. As much as there seems to be this paradigm shift across many industries, the pharmaceutical industry is behind the curve and a number of myths around R and its usability in regulated clinical trial environments are certainly hard to eradicate | data cleaning, validation documentation, pharmaceutical industry | source | 10 |
| 218 | Visualizing and Modeling Clinical Trial Data | Alex Ford, Andrea Coombs |
Across the globe, health and life science organizations need the ability to access, explore, and model data. Study start-up units view enrollment and screen fail rates across sites; monitors investigate serious adverse events; biostatisticians test interactions between predictive variables and health outcomes; business analysts forecast the budgetary spend necessary to complete a trial. In this presentation we will discuss how SAS Visual Analytics, Visual Statistics, and Visual Data Mining and Machine Learning can help answer these questions and many more facing the health and life science industry. This suite of visualization tools provides a modern, integrated environment for governed data discovery that can be leveraged by many users across the organization including those without advanced analytical or programming skills. We will demonstrate data exploration, descriptive and predictive modeling, and machine-learning techniques for analyzing structured and unstructured data | proc hpsplit, neural network, life sciences | source | 7 |
| 219 | Visually Exploring Odds Ratios between Risk Factors for Fetal and Infant Mortality in Kent County, MI | Ruth Kurtycz | Infant mortality is considered to be an important measure of the health and well-being of a nation, state, county, or municipality. The Kent County Fetal Infant Mortality Review (FIMR) program examines the case histories of all the deaths of infants one year of age or less whose mothers are residents of Kent County, Michigan. Each de-identified infant mortality case is reviewed by the FIMR Case Review Team (CRT) to see if any of more than 175 risk factors are present, and if they contributed to the infant's death. Each risk factor is grouped under a larger category such as family planning, medical issues, substance use, social support, or mental health. Understanding the associations between these categories of risk factors, including which are most likely to occur simultaneously, is an important part of developing resources to support mothers in these situations and reduce infant mortality | odds ratio, proc step, proc sql, sample, macro variable, sample size, macro loop, proc freq, macro definition, %if-%then-%else | source | 11 |
| 220 | Want Submission-Ready Datasets Package from the Get-Go? | Kishore Pothuri, Bhavin Busa |
As per the Study Data Technical Conformance Guide, each submitted SDTM/ADaM dataset should have its contents described with complete metadata in the define.xml file. It is expected that an aCRF be included along with the SDTM domains per the Metadata Submission Guidelines. In addition, it is strongly recommended to have a Study Data Reviewer's Guide (cSDRG) for SDTM and Analysis Data Reviewer's Guide (ADRG) for analysis data submission. And so, to get to a 'submission-ready' stage for study datasets, it is not just about the datasets itself, but also requires one to check for data validation rules and to prepare all other submission components that go along with it. In this paper, we will provide a detailed approach and our recommendations to build efficiencies in developing CDISC datasets and all required submission documents. And provide step-by-step process to create submission-ready dataset package towards NDA submission from the get-go | sdtm domain, sdtm programming, data cleaning, adam dataset, proc step, raw data, data validation rules, cdisc | source | 8 |
| 221 | Watch Out For Blind Spots While Keeping Up With the Speed of Evolving Standards and Regulations | Srinivas Veeragoni | The new world of evolving rules, standards and regulations of clinical data, and the automation of processes for implementing these new standards are similar to the evolving features in automotive industry ranging from blind spot warnings to completely self-driving vehicles. One thing that is common and important in both these cases is to apply common sense on top of everything. To keep up with the pace of standards evolution, sponsors and CROs are trying to automate many pro cesses in every stage from data collection to final submission. However it is important to ensure that data usability is not lost while maintaining the structural integrity of the data and compliance with standards. This paper will focus on some critical ground rules to be aware of while automating and implementing standards on clinical data | clinical data, data cleaning, clinical trial, raw data, controlled terminology, adam dataset, cdisc | source | 14 |
| 222 | Welcome to the Three Ring %CIRCOS: An Example of Creating a Circular Graph without a Polar Axis | Jeffrey P. Meyers | An internal graphics challenge between SAS and R programmers within Mayo Clinic led to the creation of a CIRCOS plot with the SAS 9.4M3 features. A CIRCOS plot is a circular visualization of the relationship between objects, positions and other time-point related data. The CIRCOS plot features a curved rectangle for each node or object around the perimeter of the circle, and a Bézier curve for each path between nodes that is pulled towards the center of the circle. Each rectangle and curve's size is determined by their proportion of the total population. The primary challenge of creating this plot with the current SAS SG procedures is that these procedures do not include polar axes to simplify plotting circular or curved objects. The macro %CIRCOS is an example of overcoming these limitations creatively using trigonometry to prove that these types of graphs are still possible without polar axes | macro variable, do loop, proc sql | source | 14 |
| 223 | What makes a good QC Programmer? | Timothy J. Harrington | This paper is a discussion of the methods employed by Quality Control (QC) Programmers, when performing QC by means of independent SAS programming, for the most effective identification and reporting of discrepancies between a SAS Programmer/Analyst (SPA) and QC program set of results, with a given program specification. The discussion concerns both comparing the output of a program, that is SAS data sets, tables, listings, or graphic output, and on reviewing log files for coding issues and input data problems. Specific skills and traits are identified which are compatible with Good Clinical Programming Practices (GCPP) and the recently implemented 42 CFR Part 11 during the QC process. The content of this paper is based on many years of experience and past formal training of the author and of both current and prior project team members | proc step, missing value, proc compare, character variable, proc sql, macro variable, likelihood function, proc freq | source | 8 |
| 224 | What Medical Reviewers Can Do With Standardized Data and Metadata Received in Module 5 | Eileen Navarro | source | 30 | ||
| 225 | When a Clinical Trial Lasts Forever: A Reporting Solution for Post-Market Surveillance of Cardiac Rhythm Devices | Becky DeBus | Post-market surveillance data is vital to the medical device industry for monitoring product performance, patient safety and patient outcomes. The collection and analysis of this data poses some special challenges. This paper discusses a unique post-market study of implantable cardiac rhythm and heart failure (CRHF) devices that began in 1983 and continues to this day. The study currently contains data collected for approximately 210,000 devices implanted in more than 75,000 patients at over 300 sites. These data have been collected in multiple geographies using numerous platforms under evolving data collection standards. In recent years the complexity of the data and expanding analysis needs have outpaced the scope of the existing reporting structure. A solution was needed to more effectively document reporting requirements, increase efficiency, and provide broader internal access to the data | regulatory reporting, sample, data cleaning, clinical trial | source | 8 |
| 226 | When Powerful SAS Meets PowerShell | Shunbing Zhao | PowerShell is an MS Windows-based command shell for direct interaction with an operating system and its task automation. When combining the powerful SAS programming language and PowerShell commands/scripts, we can greatly improve our efficiency and accuracy by removing many trivial manual steps in our daily routine work as SAS programmers. This paper presents five applications we developed for process automation. 1) Automatically convert RTF files in a folder into PDF files, all files or a selection of them. Installation of Adobe Acrobat printer is not a requirement. 2) Search the specific text of interest in all files in a folder, the file format could be RTF or SAS Source code. It is very handy in situations like meeting urgent FDA requests when there is a need to search required information quickly. 3) Systematically back up all existing files including the ones in subfolders | sas programmer, macro definition | source | 7 |
| 227 | Why choose between DATA Step or PROC SQL when you can have both? | Charu Shankar | As a SAS coder, you've often wondered what the SQL buzz is about. Or vice versa you breathe SQL & don't have time for SAS. Learn where the SAS DATA STEP has a distinct advantage over SQL. Learn where you just can't beat SQL |
proc sql, proc step, data step, cpu time, input statement, raw data | source | 5 |
| 228 | Why organizations need MDR systems to manage metadata? | Abhinav Jain | In the last decade, CDISC standards undoubtedly have transformed all data handling aspects in the sphere of clinical research. Pharmaceutical sponsors and regulatory agencies now have widely recognized standards for data collection, tabulation, analysis , and submission. Sponsors have made a significant investment of time and resources to develop skills in understanding and implementing the standards. As future advancement, clinical research industry is now heading towards utilizing technology to gain efficiency in standards implementation. MDR (MetaData Repository) system is one such tool that helps organizations to manage and centrally deploy the standards. In this paper, I intend to cover the features offered by an MDR System. How organization, therapeutic area and study-level data-collection metadata can be stored and efficiently managed with MDR system. How an MDR system is relevant to a statistical/clinical programmer | metadata management, data standard, clinical data, clinical research, data validation rules, excel, data cleaning, cdisc | source | 7 |
| 229 | Why SAS Programmers Should Learn Python Too | Michael Stackhouse | Day to day work can often require simple, yet repetitive tasks. All companies have tedious processes that may involve moving and renaming files, making redundant edits to code, and more. Often, one might also want to find or summarize information scattered amongst many different files as well. While SAS programmers are very familiar with the power of SAS, the power of some lesser known but readily available tools may go unnoticed. The programming language Python has many capabilities that can easily automa te and facilitate these tasks. Additionally, many companies have adopted Linux and UNIX as their OS of choice. With a small amount of background, and an understanding of Linux/UNIX command line syntax, powerful tools can be developed to eliminate these tedious activities. Better yet, these tools can be written to "talk back" to the user, making them more user friendly for those averse to venturing from their SAS editor | sas programmer, data cleaning | source | 7 |
| 230 | Working with the SAS ODS EXCEL Destination to Send Graphs, and Use Cascading Style Sheets When Writing to EXCEL Workbooks | William E. Benjamin Jr. | This Hands-On-Workshop will explore the new SAS ODS EXCEL destination and focus on how to write EXCEL Worksheets with output from SAS Graph procedures and spice it up using Cascading Style Sheet features available on modern computer systems. Note that the ODS EXCEL destination is a BASE SAS product, which makes it available on all platforms. The workshop will be limited to the Windows platform, but it should be simple to port the code to other operating systems. The code will be on the computers and you will get a chance to see how it handles |
excel, xls, output delivery system, sas base, proc template, proc step, cdisc | source | 19 |
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.