First SUGCDSC Conference 2018

Proceedings

First SUGCDSC Conference 2018

Date: Friday, August 31, 2018
Location: Silver Spring Civic Building at Veterans Plaza, 8525 Fenton Street, Silver Spring, MD 20910, Washington DC Area


Conference Paper
First SUGCDSC Conference, 2018

Neem Leaf Standards for Global Clinical Data: A White Paper on Standards that can Change the Landscape of Global Clinical Data Standards

Author & Speaker: Venu Perla, PhD, Founder & President, SAS User Group for CDSC


Conference Paper
First SUGCDSC Conference, 2018

Macro to Automate SDTM & ADaM datasets Creation from Specification Template

Speaker: Vidya Muthukumar, Advanced Clinical
Author: Nirosha Reddy, Advanced Clinical, IL

Abstract: As CDISC standards are required in clinical trials, many tools have been developed for automating SDTM and ADaM datasets creation. The main purpose of the macro is to increase the efficiency in building datasets, limiting errors in validation by creating a validated standard macro and also reducing the number of CDISC compliance findings. A macro program (called %make_empty_dataset) has been developed to create zero record SDTM & ADAM datasets by reading the information from its specification template. Macro first imports the specification template and will create a macro variable for number of datasets present in dataset tab (contains all the dataset names present in the spreadsheet) and will load the variable attributes into macro variables and finally creates zero observation datasets using the do loop, the datasets contains all the variables in the order of specification template and all variable attributes (i.e.name, type, label, length). In addition to creating all the datasets based on the order we have in template, this macro overcomes the challenge of inconsistencies between specification and the actual data sets.  A benefit is efficiency for creating the datasets.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

Automate the Process to Ensure the Compliance with FDA Business Rules in SDTM Programming for FDA Submission

Speaker: Xiangchen (Bob) Cui, PhD, Alkermes Inc., Waltham, MA
Authors: Xiangchen (Bob) Cui, Hao Guan, Min Chen, and Letan (Cleo) Lin, Alkermes Inc., Waltham, MA

Abstract: FDA has published “FDA Business Rules” [1], and expects sponsors to submit SDTM datasets which are compliant with the rules, as well as CDISC IG [2]. These rules assess if the data supports regulatory review and analysis. Some of them are specific to FDA internal processes, rather than to CDISC SDTM standards. Pinnacle 21 is the most commonly used tool by both the industry and FDA to check compliance with both FDA business rules and CDSIC rules. However, Pinnacle 21 is usually used at a late stage of SDTM programming development cycle, and it cannot help users to resolve its findings regarding “Error” and/or “Warming” messages, even if it is used at the very early stage.

This paper presents a systematic approach to automate SDTM programming process to ensure compliance with FDA Business Rules. This process contains study data collection design, data collection (edit-checking), standard SDTM programming process, and in-house macros for automatically reporting and/or fixing the issues to address non-compliance with “FDA Business Rules”. It avoids inefficient use of resources for repeated verification of the compliance and/or resolution of the findings from Pinnacle 21 for these rules. In fact, some of these non-compliant issues are often very “costly” and/or too late to be fixed at a late stage. The sharing of hands-on experiences in this paper is to assist readers to apply this methodology to prepare both FDA Business Rule and CDISC Standards compliant SDTM datasets for FDA submission in order to ensure the technical accuracy and submission quality, in addition to cost-effectiveness and efficiency.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

Thanks for Visiting: Tips and Tricks for Creating and Using the Subject Visits Domain

Author & Speaker: David Vanthof, QST Consultations, Ltd.

Abstract: The pooling of clinical trial data for the creation of the Subject Visits domain presents many challenges. Clinical trials often collect multiple assessment dates and/or datetimes at each visit which should be checked for accuracy as well as the need for the creation of date ranges for visits. In addition, EDC systems often restrict forms available at scheduled visits to planned assessments and require the creation of an unscheduled visit to submit unplanned assessments that occur at scheduled visits. This leads to assessments completed on the same date being tied to both scheduled and unscheduled visits. Another challenge with clinical trial data is the pooling of vendor data, such as laboratory and electrocardiogram transfers, with EDC data. Vendor data often uses differently terminology for visits and has limitations on the visit values that are exported.

This paper will detail how to pool the data together and account for the challenges described above. The paper will include methods to account for a singular date that is tied to multiple visits, visits with date ranges, unscheduled assessments at scheduled visits, and differences between vendor and EDC data. The paper will also describe validity checks on visit dates as well as how to use the Subject Visits domain when processing by visit domains to ensure more accurate data submitted to SDTM.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

Improving Traceability for Complex Algorithms in ADaM Datasets

Author & Speaker: Priya Saradha, Levstat® LLC

Abstract: Currently, ADaM datasets implement traceability using the variable triplet SRCSEQ, SRCDOM, and SRCVAR to the maximum extent possible. These variables enable the user to identify source records from SDTM or other ADaM datasets. The usage of these variables has been largely limited to identifying a single record from the source dataset. Traceability becomes a challenge when multiple records from a single dataset or multiple datasets contribute to the derivation of a single record in the output dataset. To address this challenge, we adapted the usage of SRCSEQ by creating it as a character variable, SRCSEQS, to include all the sequence numbers and their respective sources contributing to a specific record. This paper intends to discuss this simple, yet effective, approach in detail providing implementation examples.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

NONMEM- SAS Programmers Using SAS should Map to CDISC Standards PC and PK Domains for Non Clinical and Clinical Study Data

Author & Speaker: Meghnath Chavva, Booz Allen Hamilton, Silver Spring, MD

Abstract: Nonlinear Mixed Effects Modeling (NONMEM) is a type of population pharmacokinetics/pharmacodynamics (pop-PK/PD) analysis that is used very widely in clinical pharmacology modeling & simulation research. At different stages of drug development the population PK approach, coupled with PD modeling, allows integrated analysis, interpretation, and prediction about drug safety, potential drug interaction efficacy, dose-concentration
relationship, and dosing strategy. 

The analysis results help regulatory agencies evaluate new drug submission, review safety/effectiveness of a drug, and guide drug labeling. Pharmacometricians work with NONMEM® software which requires pop-PK/PD data in text (ASCII/CSV) format, whereas agencies request data be submitted in SAS transport files (xpt). SAS is used to create, maintain, update, and recreate NONMEM data sets required for modeling purposes, and facilitate the creation of regulatory format as well as the text file. 

Utilizing SAS also assists in maintaining data integrity, handling large data, tracking data manipulation, auditability and derivation through log. The use of SAS for the preparation of pop-PK/PD analysis data sets that are consumed by NONMEM® software has led to a greater demand for specialized ‘NONMEM’ scientific SAS programmers.

These professionals are tasked to pool the data from multiple clinical studies, manipulate data coming in diverse formats, combine and validate records in a single SAS output data set. This paper explains a NONMEM data set structure, some core variables, group interaction, and a programmer’s tasks and challenges involved.

CDISC Data Standards: – Domain Models Based on the General Observation Classes – Findings Most subject-level observations collected during the study for Non Clinical and Clinical should be represented according SDTM general observation class called findings.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

Text Mining Protocols with SAS Enterprise Guide to Create Therapeutic Library

Author & Speaker: Vidya Muthukumar, Advanced Clinical, IL

Abstract: In a CRO or Pharma setting, one may have received emails from management asking if anyone had experience working in indications such as Asthma and ABPA (Allergic bronchopulmonary aspergillosis) or Aesthetic or Cosmetic Dermatology products or worked with biomarkers to make a pitch for potential capabilities bidding projects.

While it may not be easy for programmers or biostatisticians to quickly recall protocol indications, it is likely that team members may overlook key experience, if they worked on different studies.

This abstract demonstrates a method to text mine keywords from protocols in PDF format to search for key indications and output the data to a library as an index of therapeutic areas. Since SAS cannot read PDF files directly, PDF files need to be converted to .txt or .xls files using Adobe Export PDF. Text mining a PDF protocol requires the first few pages to be converted to excel file which will be used to text mine and create a therapeutic library.

Using SAS EG and Proc SQL, one can build a text-mined SAS hash object table to match key indications from protocols and output it to create a therapeutic library of studies.

The object of this abstract is to provide an efficient solution using text mining and SAS EG to create a therapeutic library to not only help streamline a company’s capabilities pipeline but also potentially help with resource allocation to key therapeutic areas.

Full Paper: Click here


Conference Paper
First SUGCDSC Conference, 2018

Levista – An Integrated Programming Environment with Visualization

Author & Speaker: Priya Saradha, Levstat® LLC

Abstract: 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 LEVSTAT, 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. Visualization of various components of the specification document in one screen, increases the efficiency of the specification development process and decreases the error rate by eliminating the need to switch between multiple documents/tabs. The time spent on generating the specifications, value-level metadata and codelist is greatly reduced when using the graphical interface. This ability to connect the individual components on a single screen provides a better interface for the user to input information and helps gain control over the information being input for various components. Aside from spec development, the platform helps integrate and manage the statistical review of outputs generated. Clients, Statisticians and Managers will be able to enter their comments after review as part of the output document, which is integrated into a tracking database. The utility of tracker database is extended to store key milestone information for each study enabling the team to keep track of timelines, amount of work completed and outstanding, current status of each output. The platform provides the ability to graphically project all the information from the tracker database on a customizable dashboard. In summary, this platform integrates project management, review management, ADaM development, and metadata management under a single visual environment. This environment enables the study team to complete the programming tasks in the most efficient manner and establishes a unified platform to communicate accurate information throughout the team at any instant. Through this paper we intend to share the overall functionality of the interface, highlighting the most useful features, and present metrics of productivity which show increase in efficiency and reduction in error rate.

Full Paper: Click here


Click here for the conference program.