PHUSE EU 2022 Papers - Index

167 papers. Source: PHUSE (Pharmaceutical Users Software Exchange) EU Connect conference proceedings.

#	Title ×	Authors ×	Abstract ×	Keywords ×	Extras	Pages
1	"Co-opetition": The New Leadership Buzzword for Driving Innovation!	Bhavesh Patkar	When the author talks to industry leaders today, he has observed that all companies today are looking to innovate and optimize, but are they doing it efficiently by always competing to be the best? Although competition is a traditional force for improvement and innovation, a new modern phenomenon of "Co-opetition" is emerging as a way to use the collective wisdom by "working together" to find more sustainable long-term solutions effectively building a sense of one community with shared challenges, purpose and development to drive open innovation. As leaders we tend to think in either/or terms, as in either compete or cooperate, rather than compete and cooperate. Doing both at once requires creative leadership, mental flexibility, and many other skills. In this abstract we look at the framework of building a successful co-opetition strategy, its advantages and how leaders today need to evolve to move from a zero-sum game to a plus-sum game		source presentation source	7
2	'Personal Balance Beam': Work-Life Balance in a Post-COVID World	Vahe Martirosyan, Arpine Melkonyan	With the pandemic slowing down and offices opening back up, we're spending drastically less time at home. Where are our hours going? Is working from home giving people adequate connections and the same opportunity to climb the career ladder? This talk will help you find that equilibrium		source	4
3	'Verify Outputs' Product Demonstration: Validation of Statistical Outputs Using Automation	Noa Sharon, Roi Kain	Presentation	data cleaning, audit trail	source	18
4	10 Things I Love About SAS Studio	Carrie Foreman	Poster	sas studio, sas enterprise guide	source	1
5	A Beginner's Guide to a Multilingual Shiny App	Karl Kennedy	Within the Shiny Environment you can leverage the power of multiple languages/technologies e.g., R (obviously!), JavaScript, Python, CSS, and R Markdown. The primary goal of this paper is to demonstrate the usage of these relatively advanced topics in a simple Shiny app so Shiny beginners can utilise and incorporate these concepts into their first Shiny apps. While also reducing the programming learning curve in implementing these new concepts. It will use the power of Python's string operations, CSS and JavaScript for front-end functionality, and R Markdown for reporting results. It will also cover some basic Shiny fundamentals along with good programming methodologies e.g., modularisation	txt, dataset	source presentation source	7
6	A Hitchhiker's Guide to the Corporate World for Newbies	Ritika Aggarwal	Stepping into the corporate world for the first time can be intimidating and exciting at the same time. It is extremely important to hit the ground running in a dynamic environment. A new phase comes with all the stressors of having new responsibilities, an unfamiliar culture, information overload, uncertainty in what is expected, and the list goes on. The key is to have a structured yet agile approach towards growth and learning at the inception of the career, to tip the balance. This paper brings out the essence of four cornerstones that can help build a strong foundation for a newbie in the corporate world - Communication, Collaboration, Self-leadership, and Feedback. My personal lessons learned will be shared at the end to provide the readers with the key steps and tips for sailing through the first of the career stages	clinical programming, clinical trial	source presentation source	5
7	A Native SAS Excel Interface - Replacing the SAS Engines	Ivo Marek	For years the `IMPORT` procedure and the `LIBNAME` statement have been able to access `EXCEL` spreadsheets, but they have some drawbacks. Why do they not recognise `EXCEL` cell formats? For example, if a numeric `EXCEL` column is formatted with one decimal place, why is the SAS format BEST12. automatically assigned to the imported column, even though the SAS format 8.1 is more convenient? Moreover, why use options like DBMAX_TEXT when the maximum string length can be determined? With the release of `EXCEL` 2007, the Office Open XML (OOXML) file format was introduced, with content and cell formats stored in XML files. By processing the XML files directly, the target SAS dataset structure can be derived more precisely from the source file. This calls for an easy and native SAS macro!	hash table, excel, format statement, call execute, dataset, proc import	source presentation source	9
8	Adaptive Design - Better Bend Than Break	Albin John, Athira Sudhakar	Presentation	sample, sample size, sample size re-estimation, adaptive design, effect size, randomization, interim analysis	source	15
9	Advanced SAS techniques to Generate Automated Topline RTF Reports - Used Daily by PMs	Mizan Alam	Before moving to final CSR, it is very important for the project managers (PMs) to review the Topline results on a regular basis for every single Dry -Run. The process of populating TLFs, updating TLF numbers, cross-references and incorporating results in Topline reports is laborious, time consuming and potentially error prone. An automated technique is presented in this paper which uses SAS and literate programming concept to generate study topline reports with specified text, TLFs, and cross-references with Table of Contents (ToC). Literate programming efficiently combines the text and statistical analysis in one single SAS source file, which can be reused and updated multiple times without applying any changes. This process will significantly reduce the amount of time to produce the study topline reports in RTF format. The goal of this process is to complete the final CSR, which is the final milestone of any clinical trial	adverse events, clinical trial, confidence interval, sensitivity analysis	source presentation source	15
10	Agile RnD - From Empiricism to Success	Marc Jantke	Pharma companies are trying to adopt agile methodologies and toolsets like Scrum and Kanban. Such frameworks are made and optimized for software development projects, and not easily applicable to the processes of clinical research and development. Yet the industry is trying to gain benefits by cherry-picking pieces of such agile frameworks, with limited success. This presentation takes a step back to look at the basis of agile methodologies: empiricism. The very foundation of Scrum is to drive decisions based on empirical knowledge in order to solve complex problems. All further aspects (transparency, inspection, adaptation) of agile are based on this fundamental concept; all organizational means (teams and roles), artifacts (backlog), processes (iterations and feedback cycles) and tools (software) are designed to support the empirical approach	clinical research, clinical trial	source presentation source	8
11	AI_PREMie: Saving Lives of Mothers and Babies Using AI	Patricia Maguire, John Curran	Presentation	sas viya, sample, data science, standard deviation, biomarker, champion model, data cleaning	source	26
12	Aligning Statistical Techniques Between R and SAS - A Practical Example Using Cluster Analysis	Stephen Prust, Mike Cooper	Presentation	cluster analysis, proc distance, proc step, proc transpose, dataset, data cleaning, macro variable	source	13
13	An Approach to Effectively Reproduce TFLs When Performing ADaM Conversion	Amine Khemiri, Caroline Pelckmans	Presentation	adam dataset, excel, artificial intelligence, clinical research, clinical trial, sdtm domain, raw data, cdisc	source video	19
14	An Automated Approach to Generating Call Programs from an Analysis Request	David Maher-McWilliams, Michael Allie	In clinical research statistical programming, it is common for an organisation to have a library of standard macros. This is an efficient way to re-produce analyses across a company's projects, with each project having call programs to the standard macros. Creating these call programs can be a manually intensive and time-consuming task, requiring a programmer to interpret the request, manually create the program and populate macro parameters. This is also a possible source of error from, but not limited to, misinterpretation and typos when populating parameters. Creating a structured analysis job request form can help to alleviate misinterpretation of requests, and this machine- readable document can be leveraged to programmatically generate and populate call programs	macro call, macro parameter, hash table, macro variable, dataset, clinical research, statistical programmer, data science	source presentation source	9
15	Automate to Accelerate: Saama's Smart Auto Mapper	Mike Collinson		machine learning, data transformation, clinical data, clinical trial	source presentation source	5
16	Automation in Biostatistics at Novo Nordisk	Matthew Phelps	Presentation	mock shell, data science	source video	37
17	Automation of Standard ADaM Variables Through an R Package - GenAdam	Mahesh Divakaran	For the last couple of years, R software applications in clinical research have been a hot topic in the pharma industry. It offers a variety of user-developed packages containing functions which efficiently manipulate complex data sets and create tables, figu res, and listings, perfect for biostatisticians and programmers in the pharmaceutical and biotech industry. But the development of CDISC compliant datasets using specifications or define as the base is a time - consuming manual effort be it through SAS or R. In this paper, we introduce an R package that can be an effective alternative in creating clinical trial ADaM datasets with a single function call followed by visualizing the created dataset. The entire concept spans across two phases. The first phase, where the focus of this paper lies, creates ADaM datasets with standard variables & predecessors presented. The second phase takes as input the study specification and maps variables to corresponding datasets	adam dataset, clinical trial, in=, clinical research, clinical data, sdtm domain, data science, data cleaning, r package	source presentation source	8
18	Automation of the CDISC Pilot study - a step-by-step guide	Stuart Malcolm		data pipeline, rest api, clinical data, clinical trial, excel, data cleaning, cdisc	source presentation source	9
19	Best Practices for Automated Delivery Set-up, from Raw Data to CRT, with Methods to Standardise and Streamline Programming and Delivery Processes	Daniel Carden	Presentation	raw data, adverse events, create table, macro call, safety population, system organ class, data lineage	source video	21
20	Bioresearch Monitoring Data Reviewer's Guide (BDRG)	Jack Field, Steve Fitzpatrick	Presentation	clinical trial, crf, clinical data, pharmaceutical industry, data validation rules, data cleaning	source video	20
21	Breaking Down the Data Standards Silos - A New Metadata Ontology for Data Analysis Derivations	Chris Price, Nelia Lasierra	In the last years, the FAIR principles have re-focused attention on the importance of data standards and formal vocabularies. Traditionally, these have existed in individual silos which has been a significant roadblock to end-to end automation efforts. Developing connectivity between these is essential to facilitate automation in data transformation and pooling activities. At Roche, data standards and vocabularies are described using ontologies and knowledge graphs and stored in an RDF-based metadata repository that enables their extraction in a machine-readable format. We conducted an experiment to evaluate the feasibility of extending our current ontologies describing SDTM and ADaM and to connect them via derivations to answer key questions related to the optimization of analysis, pooling, and curation processes	adam dataset, data science, data standard, raw data, cdisc	source presentation source	7
22	Building a Trial Pipeline with R Markdown and RStudio Connect	Ari Siggaard Knoph, Kristian Trøjelsgaard	Presentation	clinical programming	source video	18
23	Building an Efficient R Framework from Scratch	Pierre Wallet	Poster	data cleaning	source	1
24	CALL EXECUTE will figure it out: Data-driven Figure Programming	Mitchikou Pearl Tseng	In studies with multiple parameters of interest taken at varying timepoints under different treatment groups, parts, or periods, BY statement and/or employing SAS MACRO in SAS GRAPH are sometimes not convenient in the creation of repeated figure outputs. But, by incorporating CALL EXECUTE in our figure programming, we can successfully create repeating graphs for which input values are data driven and add specific conditional attributes that are flexible to the user. This paper will discuss a data-driven figure programming approach in the following steps: 1) Creation of a macro for your figure, 2) Addition of new variable(s) to your input dataset for possible specific tweaks to your figure and 3) Making use of `DATA STEP` + CALL EXECUTE to robustly call the figure macro and apply conditional tweaks to your output	call execute, macro parameter, in=, macro call	source presentation source	6
25	Capability Building is the New Normal in the Pharmaceutical Industry	Jaganadha Rao Gundu, Sreekar Mateti	The next few years will bring fundamental changes to pharmaceutical industry, and to adapt, employees in almost every role and industry will need to acquire new skills. The COVID-19 pandemic forced companies to operate differently, and it affected in many ways. Every company I know in pharma is looking for resources and many aspirants are looking for jobs in Pharma industry. There is a gap between companies and resources. This paper talks about how we can create the resources and how we can develop the people instead of searching for the resources. It talks about how we can adapt to the rapid swifts. This paper will also discuss how design thinking helps manager from new or experienced managers to solve the problems effectively. Design thinking has been successfully applied to developing new products and services already. This paper talks about how design thinking helps capability building in these unprecedented days	pharmaceutical industry, clinical research	source	8
26	CDISC Analysis Results Standard - Update and Progress	Karl Wallendszus, Hansjörg Frenzel	The CDISC Analysis Results Standard (ARS) Team is working towards creating standards to improve and facilitate the automation, reproducibility, reusability, and traceability of analysis results. The team is working towards this objective by creating an analysis results data model (ARDM) to support reuse of analysis results as well creating an analysis results metadata technical specification (ARM-TS) to support traceability, creation of analysis results, and creation of data displays. This paper provides an update of the team's progress and visualizes the development goals of the ARS team	adam dataset, data science, cdisc	source presentation source	7
27	CDISC Open Source Alliance (COSA)	Gerry Campion	Poster		source	1
28	Clinical Data Standards and Production of SDTMs. Are We Missing Important Points?	Eva Stigson, Paulina Kuczkowska	Presentation	sdtm programming, data transformation, adverse events, clinical data, clinical database, clinical trial, data standard, cdisc	source	15
29	Cloud Automation and Why I Should Use It	Roger Chivers	Presentation	financial services, life sciences	source	14
30	Collaboration Between Analytical Health Care Technology Leaders will Positively Impact Clinical Trial Research Innovation & Improve Health Care System	Mark Lambrecht, Andrew Warrington	Presentation	clinical trial, 21 cfr part 11, life sciences, clinical research, output delivery system, data science, adverse events	source video	19
31	Cross-functional Standardisation of Controlled Terminologies Through Knowledge Graphs	Daniela Bergann, Ena Kolundzic	Presentation	cdisc controlled terminology, controlled terminology, data integration, cdisc	source video	15
32	Culture Club: Our Experience of Creating an Environment where Individuals Feel Psychologically Safe, Fulfilled, Valued, Valuable and Impactful	Ivan Robinson, Huw Mason	Presentation		source video	14
33	Data Catalogues - The new Showman to organize your data, A CRO perspective	Limna Salim	In an era where data driven approach is prioritized by CROs, Data catalogues are an important functionality. Every team after adopting data catalogues find it productive to organize and review their data. Here we focus on creating secure Web based interactive catalogues across use cases mentioned from a CRO perspective. Also using a standard modular template optimizes structuring and retrieving information across the organization. Use Cases detailed in paper are as below- Macro Library Standard Study Repository Metadata Extractions for CDISC (SDTM/ADaM) programming Statistical Programming library A good catalogue contains elements such as Data discovery tools and keyword search engines that would allow users to discover metadata. It also provides the ability to download contents enabling reusing scripts in a downloaded format. This paper also reviews Different elements that catalogue contain. Development using open-source platforms like R and Python with Markdown files	clinical programming, data science, cdisc	source presentation source	8
34	Data From the Clouds ... and Back Again	John McDade	Cloud data storage is ever increasing, allowing users to access files from different devices and work collaboratively when required. This paper looks at different ways of accessing that data from SAS, with authentication when needed. Cloud data can be downloaded, modified at source or new data uploaded. Often programming with cloud-based data is reserved for application developers but we will look at some simple methods of connecting to the cloud within a clinical trial setting. The paper will look at downloading controlled terminology versions from NCI website and the CDISC Library and also reading and writing data to Microsoft SharePoint	proc http, proc step, excel, rest api, macro variable, sample, controlled terminology, xls, cdisc	source presentation source	12
35	DataSet-JSON - Smaller than XML, But Just as Powerful	Fabian Noelle	Presentation	safety population	source	10
36	Decoding in R	Himani Narang	Creation of ADaM datasets often requires generating variables by decoding values from codelists metadata as per CDISC submissions values. The newly created variables can be either in numeric or character formats, and creating them individually by merging with codelists dataset can be a cumbersome task. Many programming languages contain data structures that allow storing key-value pairs that can make this task easier. On the other hand, R does not have a direct built -in data structure for storing key -value pairs. However, other approaches can be used to mimic the functionality, such as named vectors and named lists. We have created a function in R named decode() which creates named vectors from codelists dataset. By simply accessing the vectors in a repetitive manner using mapply() function, the input values, or the keys, are decoded to the desired values and multiple variables in different formats can be generated at the same time	adam dataset, dataset, cdisc	source presentation source	5
37	Deep Learning to Classify Adverse Events from Patient Narratives	Qais Hatim, Tom Sabo	Presentation	adverse events, deep learning, machine learning, csr, data cleaning, clinical trial, data science	source video	12
38	Devices and Decentralised Clinical Trials Using a Storage and Advanced Analytics Platform for the Clinical Data Environment (ALYCE)	Stephan Cichos, Anja Thiede	Presentation	data standard, data transformation, feature selection, adverse events, clinical trial, data protection, data cleaning	source video	23
39	Diving into Machine Learning in a Data Lake	Cathal Gallagher	Poster		source	1
40	Do It Yourself: Case Report Form (CRF) Annotation	Caro Sluijter	There is more than one way to skin a cat; and there is more than one way to annotate Case Report Forms (CRF). As regulatory authorities and CDISC have guidelines to deal with metadata submission, we have a road map to create annotated CRFs. However, some guidelines contradict each other, and what do you do when your aCRF cannot adhere to all guidelines? As a CRO we annotate CRFs, but we also work with CRFs annotated by clients and we see differences regarding annotations and bookmarking. These different approaches are not wrong, but to be consistent we try to annotate CRFs following the FDA Technical Specifications Document and CDISC Metadata Submission Guidelines. CDISC guidelines are updated in March 2021, therefore we took a closer look at how we think we can best annotate CRFs. Furthermore, by automating annotation and bookmarking of CRFs we can speed up the process considerably	annotated crf, sdtm domain, cdisc	source presentation source	5
41	Don't Be a Mutt - Use SAS MUTT (Macro Unit TesT)	Michael Thomsen	Presentation		source video	16
42	DOT Language for Consort Flow Diagrams	Nicolas Guerro	SAS makes it possible to manage a wide variety of graphics using SG procedures, annotations and templates... Nevertheless, it is often seen that some graphics require many tweaks and pre-processing for the program to remain readable, data independent and reproducible. This is the case with flow diagrams which have been popularized by the CONSORT (CONsolidated Standards Of Reporting Trials) group to represent the interactions between populations in a structured way. Several solutions are available in existing papers, however the one that will be discussed here relies on the graph description language, DOT, to visually describe the relationship of objects before demonstrating how R is used to execute it and SAS to manipulate data, retrieve and refine the output	macro definition, proc step	source presentation source	13
43	DRM Version 2: The Standards Strike Back	Marc-Antoine Prodhomme	Poster: Data Review Model Version 2 (DRM V2) is an upgraded version of a Janssen proprietary data model that is integrated into an automated data flow designed to enhance review and visualizationsof clinical trial data. DRM V2 provides full traceability and describes both clinical and operational data consistently across studies and is considered as the Single Source of Truth, that eliminates multiple and inconsistent views of the same data. It enables all stakeholders to access the same data at the same time. DRM V2 incorporates CDISC SDTM data structures and follows proper adherence to the CDISC SDTMIG to reduce the effort necessary for data mapping to SDTM. This new version of DRM has now been implemented in numerous trials from different therapeutic area and is showing great accomplishment for the easiness of data consumption and early access to critical data	clinical data, data review, cdisc	source	1
44	Effective NLP API Access to Unstructured Text for Downstream Analysis and Models in Pharma: Bench to Bedside	Jane Reed	Presentation	machine learning, adverse events, clinical trial, data science, clinical research	source video	26
45	Enabling the End-to-end Data Pipeline with the elluminate Clinical Data Cloud	Achilleas Zaras	Presentation	statistical programmer, data cleaning, access control, audit trail, data migration, data wrangling, sas log	source	12
46	Entering the Pharmaceutical Industry - A New Graduate Perspective	Piotr Karasiewicz	Poster		source	1
47	Evolving into Data Science. What Has Changed?	Sarah Kirk, Ivan Robinson	Presentation	data science	source video	19
48	Exploratory Analysis - People-powered Pathways to Success within a Critical Risk-based Approach	Simon Jackson	Presentation	scatter plot, outlier, data cleaning, adverse events, bar chart, box plot, data science	source video	21
49	F2 (FAIR and Filing) Dashboard	Nigel Montgomery, Bhupendra Mistry	Poster	quality assurance, database lock	source	1
50	FACT - Fast Access to Clinical Trial Insights	Julia Gnatek, Steffen Vangsgaard	Presentation	standard deviation, clinical trial, proc freq, amazon s3, cloud storage, statistical programmer	source video	27
51	Finding Anomalies and Outliers in Clinical Trial Time Series Data	Pekka Tiikkainen	Correcting errors in clinical trial data as the trial is on-going is vital for an accurate safety and efficacy read-out. In this talk, we will show how time series can be used to identify trial sites and individual subjects with potential errors in their clinical parameters (e.g., vital signs or laboratory measurements). Each subject-level time series is summarized as scalars such as mean and entropy. These features alone are useful in identifying individual subjects with suspicious data. Furthermore, trial sites with systematic bias in their data are flagged (for example, all albumin measurements at a site might have a growing trend while there is no trend at the study level). Results of the algorithm are available to anyone in our development organization who are involved in ensuring the quality of our trial data. Actions resulting from the results can range from sending queries to planning complete study audits	time series analysis, outlier, clinical trial, vital signs, sample	source presentation source	6
52	From 'Being Alive' to 'Living': Why Does PRO Data Matter?	Nimisha Joshi, Rachel Thackray	Poster	clinical trial, clinical endpoint	source	1
53	From Artifact to Metadata: A Divergent Approach to Automation Through Standardisation	Nicolas Rouillé, Camie Britton	Companies of all sizes are turning to metadata repositories for standards management, from protocol development, to eCRF, SDTM and ADaM to Analysis Results. In comparison to sponsors, CROs face specific challenges when attempting to utilize approaches to standards management/metadata use given that not all aspects of clinical trial are always awarded to a single CRO. Against this backdrop, Cytel Inc. employed an approach to metadata and standards use as an accelerator of automation across its delivery processes focusing on the metadata extraction inherent in artifacts such as mock shells. The initial capture of metadata from these artifacts was then cycled through Cytel's delivery chain via a set of tools. As such, Cytel's approach to metadata began with artifacts, with the databasing/search/reuse of historical artifacts from prior projects	mock shell, clinical trial, data governance, data science, cdisc	source presentation source	4
54	From Silos to Knowledge	David Iberson-Hurst	Presentation	cdisc controlled terminology, proof of concept, cdisc	source	16
55	From Stored Process to Job Execution - Build Your SAS (Lego) App	Olivier Bouchard	Presentation	machine learning, data science, proc python, sas programmer, sas viya, proc step	source video	14
56	From Zero to Full-Fledged SCE (Or Other Business Relevant System) in One Hour	Hans-Peter Schreiter	Presentation		source video	14
57	Functional Programming: What, Why and How?	Tom Ratford		adam dataset, missing value	source presentation source	8
58	Generating Schedule of Activities as FHIR Resources	Andy Richardson, Derek Hummerston	To utilise data originating in EHRs, methods are required to link subjects' data to the study SoA. FHIR has a set of resources to plan, schedule, follow and collect study data from patients who are Research Subjects. Central to this is the ability to accurately define the SoA using FHIR resources. The HL7 Vulcan SoA project is developing a SoA FHIR Implementation Guide to support clinical research data collection for implementation by EHRs. This paper and presentation discusses work aimed at further understanding the requirements for defining and manipulating study SoAs; a necessary pre-requisite to generating the study requirements using FHIR definitional resources	clinical trial, visit	source	5
59	Generating Statistical TFLs Using ML, Deep Learning and NLP Techniques that Provide More Efficiency to the Statistical Programming Process	Jeyantharkumar Arunachalam	Presentation	mock shell	source	7
60	Going through an AdCom with the FDA: The Ultimate Programming Experience	Frederikke Høyer	An Advisory Committee meeting (AdCom ) where clinical data is being reviewed and debated with an Expert Panel is in some case needed before the FDA makes a regulatory decision regarding the approval of a drug. The preparation process requires close cross -functional collaboration beyond Biometrics where the programming team has a crucial role. The programming team must prepare for all known possible questions in advance which could result in thousands of analyses. On the day of the AdCom, the programmers play a critical role in the success as they are expected to be able to produce validated analyses in a matter of minutes for any unexpected questions from the Expert Panel. There is no possibility to come back after the meeting with any clarification or additional data as the Expert Panel will proceed with voting right after the Q&A. This paper will share our experience and lessons learned from the programmers' perspective	statistical programmer, clinical programming	source presentation source	5
61	Governance and Cataloguing of Analytics Assets	Paul McCaffery, Clement Benemara	Presentation	machine learning, data lineage, data dictionary, data cleaning, data science, data standard	source	15
62	GSK Diversity Paper: The Challenges We Faced with Real-world Data	Harrison James, Chris Leach	In early 2021, Pfizer published a paper titled "Demographic Diversity of Participants in Pfizer Sponsored Clinical Trials in the United States", which issued a challenge to other Pharma companies to analyse and publish their data. This paper will document how GSK's Stats and Programming considered their approach and implemented the integration of 930 trials, conducted between 2002-2019. This was extended to include census and epidemiology data. The paper will cover how the data was combined, and the complex tables and figures used to summarize the data. In summary, it will also cover the challenges faced and lessons learned	bar chart, dataset, clinical trial, histogram, reference data, adam dataset, cumulative frequency, final analysis, data cleaning, proc freq, proc step, cdisc	source presentation source	43
63	Handling Missing Data in Statistical Programming Using Multiple Imputations in SAS	Marta Tomczyk, Martyna Gawlowska	Presentation	proc mi, imputation, missing value, proc step, proc stdize, multiple imputation, histogram	source video	15
64	Harnessing AIML Technologies to Expedite Anonymisation of Unstructured Data: Implications for the EU CTR	Niamh McGuinness, Muhammad Oneeb Rehman Mian	The EU Clinical Trials Regulation (CTR) requires the publication of many documents submitted during the clinical trial lifecycle. Personal and commercially confidential information must therefore be protected through redaction and/or anonymization. In an effort to align existing anonymization solutions for clinical study report (CSR) data to the novel demands of EU-CTR, practical challenges emerge such as complex document formatting, high throughput, and varying regulatory requirements. In this regard, Artificial Intelligence and Machine Learning (AI /ML) technologies can be used to successfully build and deploy at scale a system for anonymizing CSR data. Clinical Natural Language Processing (c-NLP) aids in handling unstructured data, in particular c linical documents, to accelerate detection and classification of personal information for anonymization	clinical trial, csr, machine learning	source presentation source	4
65	Harnessing Routinely Generated git, github and Package-specific Metadata to Build a Portal into Open-source Projects	Mathieu Cayssol, James Black	Presentation	regression analysis, pharmaceutical industry, logistic regression, clinical trial, proc esm, missing value, proc mixed	source video	18
66	Health Authorities Inspections: What to Expect? A Statistical Programmer Experience	Alexander Carrington	A crucial aspect of clinical trials is the submission to the regulatory authorities, not only for the pharmaceutical companies manufacturing the compounds, but also for the Clinical Research Organizations involved in the projects. It's imperative that when submitting data analyzed in clinical trials we follow all necessary regulations, such as GCP or guidance developed by the regulatory body themselves, as well as our company processes. It's likely once a project is submitted to the appropriate regulatory authority, as a programmer, you will be expected to respond to queries/feedback and potentially take part in an inspection. The purpose of this presentation is to share Cytel's unique experience of being involved in a simultaneous EMA/FDA inspection and highlight the different approaches and expectations of the regulatory bodies	statistical programmer, clinical trial	source presentation source	5
67	How a Clinical Metadata Repository Facilitates Compliance with Industry Standards and Expedites Metadata Storage, Sharing, Analysis and Submission Times	Gilbert Hunter	Compliance with CDISC and other industry standards is mandatory for clinical trial submissions to the FDA, PMDA and MHRA. However, standards compliance is often still viewed as a road block, or considered an after-thought in the clinical trials process. This presentation explores the benefits of implementing industry standards from the start of a trial. And how designing a compliant clinical trial upfront leads to faster delivery of submission deliverables, ultimately avoiding submission delays. We explore how a Clinical Metadata Repository (CMDR) facilitates the implementation of industry standards, including how it handles new versions and changes to standards. This includes how a CMDR can be used to design studies in previous and current versions of standards, and how such a platform stays aligned with evolving standards	clinical trial, sdtm domain, cdisc	source presentation source	2
68	How Can We Integrate New Technology and Maintain Data Integrity?	Shafi Chowdhury	When a 14 year old teenager produced for me in an hour a complex figure that I think would take most SAS programmers at least a day, should we be worried? Or should we be excited that there are new things we can learn? Should we be happy about the new opportunities and the new paths they may lead to? Uninhibited and unburdened by our ingrained ways, left to their own devices, the young generation are able to solve problems quickly, efficiently and correctly. Although we are embracing software like R, too often it still feels like the hobby of an enthusiast within our team, who uses it in presentations here and there. Is it time to bring everything into the fold and find them a place within our process - be it R, Python or any of the many other tools around? This paper will explore how programmers can embrace new technology without affecting delivery integrity	data integrity, pharmaceutical industry	source presentation source	4
69	How to Avoid Conformance Issues When Metadata Varies Across Parameters	Caroline Pelckmans, Oussama Zineddine	Presentation	bds, artificial intelligence, clinical research, statistical programmer, mock shell	source video	24
70	How to Evolve as an Experienced Programmer, Present and Future, Leaving No One Behind	Bal Toora	Poster	clinical programming, pharmaceutical industry	source	1
71	How Will Knowledge Graphs Improve Clinical Reporting Workflows?	Alexey Kuznetsov, Jorine Putter	Presentation	proc means, adverse events, privacy, laboratory data, data cleaning	source video	15
72	iCDM: Interactive Dashboards that Innovate Centralised Clinical Data Monitoring at Danone Nutricia Research	Paul Vervuren, Dennis Menkveld	With an integrated approach and the use of automation and digital technologies, central data monitoring can bring substantial benefits for quality, patient safety, operational efficiency and costs. Danone Nutricia Research (DNR) initiated a project named 'interactive Central Data Monitoring' (iCDM). This project encompasses the development and roll-out of dashboards, development of methodology and the implementation of new process and data flow. iCDM fits into DNRs strategy on data and digitalization. In this paper we will discuss the technical approach, present the iCDM dashboard design and functionalities, share the project challenges and learnings, and reflect on the organizational journey	sdtm domain, clinical trial, edc, cdisc	source video presentation source	5
73	If You Build it, Will They Come? The Pros and Cons of a Build Versus Buy Strategy for Statistical Computing Environments (SCEs)	Venu Mallarapu	Presentation	access control, clinical research, statistical software	source	13
74	If you want to deliver standardised clinical data sets quickly, talk to the machine	Chris Austin, Michael Phillips	We trained a model to automatically map source electronic case report form variables to Study Data Tabulation Model standard variables. The application of data collection and data delivery standards to clinical trial data are not always perfectly aligned. We used a combination of machine learning and natural language processing techniques to train a model against four of the core safety domains. Overall, the results were good, with the model accurately mapping 80, 80, 65 and 75 percent of variables for adverse events, concomitant medications, demographics and vital signs, respectively	sdtm programming, training set, clinical data, response variable, adverse events, cdisc	source presentation source	6
75	Implementation of a Bespoke SDTM to Facilitate Analysis of Pooled Standardised Datasets	Sadie Kelly	Poster		source	1
76	Inference from weighted subjects	Andrea Parodi	LivaNova is the manufacturer of Vagus Nerve Stimulation (VNS) Therapy which is an approved adjunctive treatment delivered by an implantable device to treat epilepsy by reducing the frequency of seizures. We explored a potential clinical trial design to evaluate the effectiveness, safety, and healthcare utilization outcomes of the VNS Therapy for the treatment of drug-resistant epilepsy against the current pharmaceutical Standard of Care (SoC). The design described is a long-term prospective, multicenter, parallel-group, observational study. The primary endpoint is the maximum duration of continued, monthly response to treatment. The treatment choice is decided by the physician and patient, prior to receive the new treatment. From a statistical point of view, the issues are: Duration of response rarely follows a common distribution; Accounting for baseline differences in the non-randomized treatment groups	confidence interval, type i error, sample size, proc power and sample size, sample	source poster source	4
77	Integrating SQL Queries: From PROC SQL to SQLDF	Julia Psenner	Poster	proc sql, proc step	source	1
78	Interactive Data Visualisation for Faster and Improved Data Insights	Markus Niederstrasser	Presentation	outlier, data visualization, data review	source video	38
79	Introducing Raccess, an R Package that Provides a Generic Framework for Implementing Sophisticated Access Controls in Shiny Applications	Peyman Eshghi	Presentation	access control, r package	source video	15
80	Is Your Analytics Platform Ready for Handling Real-world Data for Regulatory-grade Real-world Evidence?	William Kuan, Olivier Bouchard	Real-world data (RWD) and real-world evidence (RWE) receive increasing attention in the life science industry. RWE has the potential to save on the cost of research and development, and shorten the time for patients to access innovative treatments. In order to use RWE for regulatory decision making, FDA has issued multiple guidance s documents to address the usage, benefit s, risks, and consideration of using RWE for both safety and effectiveness determinations. An open and trusted research environment ca n help an organization leverage the value of RWD through talents with different skill sets, ensuring confidence in the quality, reliability, relevance, and transparency of the RWE generation process. We discuss here the key considerations of building a RWE analytics platform and long-term strategies in using RWD and generating RWE with analytical capabilities for regulatory submission purposes	analytics platform, regulatory submission, clinical trial, big data, data cleaning, life sciences, insurance claims	source presentation source	7
81	It Takes Two to Tango - Pre-checker and Create Mapper	Carla Santillán, Marc-Antoine Prodhomme	Presentation		source video	25
82	Jumping Over the Compliance Barrier. How to Drive Innovation and Thrive as a Leader in a Regulated Industry	Erwin De beuckelaer	Presentation	sample	source video	32
83	Learning to 'Let Go'	Emma Moore			source presentation source	4
84	Let's Get Groovy - Word and PDF Processing in SAS	Katja Glass	Sometimes functionality from SAS is not sufficient to work with Word or PDF documents, whereas there are many solutions for problems we face available in JavaTM. There are various ways how other programming languages can be utilized in SAS. This presentation will show how `PROC GROOVY` can be used to apply available solutions for specific problems we face when working with PDF and word documents	proc groovy, proc step, macro variable, txt	source presentation source	8
85	Leveraging Information from Healthy Volunteers' Placebo Data	Koen Ven	Presentation	proc means, adam dataset, clinical trial, laboratory data, macro definition, cdisc	source video	17
86	Leveraging Manifest Files for Impact Analysis: Preparing Studies for Critical Deliverables (Interim/Final Run/Submissions)	Indresh Kumar Singh	Having inconsistent outputs, chronology of the execution timestamps not matching with the quality checklists are common issues encountered by programming teams during the preparation of the outputs needed for a critical milestone. The challenges with clinical trials is multiple sources of inputs, macros, multiple teams involved in changes, lack of traceability between inputs and outputs and impact analysis of the changes. In these dynamic situation teams face issues like- Communicating changes in outputs of one team that impact inputs to another team, Align and communicate modifications in macros or programs used by multiple programmers/teams, Lack of alerts for changes in input data sources. The presented approach uses manifest files in SAS Life Sciences Analytics Framework to automate the impact analysis across all teams/programs and share alerts to the impacted teams. The approach also provides a mechanism to visualize the dependencies for better clarity and communication	clinical trial, dataset	source poster source	4
87	Managing Never-ending Extension Trials	Frederik Vandvig Heinen, Lykke Pedersen	Presentation	statistical programmer, clinical programming	source video	19
88	Managing the Transition from Programmer to Manager	Mark Crangle	If and when the time comes for a programmer to consider moving into management, the decision can be a daunting one. On the face of it, line management responsibilities could seem alien to someone with a scientific background who is well used to applying logic and algorithms to solve problems. While some of the skills developed as a programmer will still be useful, the new manager will need to be able to step back from the day-to-day decisions on a project, and an array of new soft skills will be needed in order to successfully manage, motivate and develop others. This paper will draw on personal experience and examples from other industries to explore ideas and tips to prepare for a move into management, to handle the period where old and new responsibilities may overlap, and to continue to develop into a successful manager	sas programmer, sas programming	source presentation source	6
89	Maturing standardisation; the switch from spreadsheet-based standards to an MDR	Louella Schoemacher, Jasmine Kestemont	In order to work consistently, structured, and time efficient throughout multiple studies, standardisation is key. Standardisation can be relatively simple with having the company standards in a spreadsheet and using this spreadsheet to share study-specific metadata requirements. However, with a growing number of compounds, indications, studies, and number of CRO's involved in creation of the SDTM package (aCRF, SDTM datasets, Define -XML and SDRG), it gets harder and more important to have and maintain the standards. Furthermore, individual studies can have specific requirements which may result in deviations even within a compound or indication. When the number of studies grows, a simple spreadsheet may no longer be sufficient to capture all subsets of metadata requirements. When the standards cannot keep up with the growing number of studies, this can result in an overgrowth of different metadata definitions, leading to undesirable inconsistencies	adverse events, sdtmig, cdisc	source presentation source	5
90	Migrating a Large User Base from Multiple Legacy Systems to entimICE-AZ: Issues and Lessons Learned	Panteleo Nacci	This paper discusses some of the user-related issues encountered when AstraZeneca (AZ) migrated GCP-related analysis and reporting activities from a number of legacy systems to a modern Statistical Computing Environment (SCE). At the same time historical and closed studies were moved from several file storage locations to a read-only one (Clinical Trial Data Storage, CTDS), from where they could be imported into the new environment if and when additional analyses were needed. While data, programs, etc. for studies migrated to the new SCE were remapped, those moved to CTDS were copied 'as is', thus maintaining their original folder structures	output delivery system, data integrity	source presentation source	5
91	Minimising 'Data Loss' During Mapping of Raw Data to SDTM	John Monaghan	Presentation	macro variable, sdtm programming, raw data, order by, data cleaning, proc iml, proc step, cdisc	source video	35
92	Modelling Electrocardiogram Waveform Interpretations for SEND Submissions	Jack Baker	Poster: Modelling data for electrocardiogram (ECG) waveform interpretations can be challenging in SEND. Currently, within the latest confirmed data endpoints for exchange (CoDEx), counts of arrhythmia types in continuous ECG measurements are not in scope for SEND. This poster explores methods to present counts of arrhythmia types in continuous ECG measurements within the ECG domain and provides best practice recommendations for future SENDIG guidance. Acknowledgements I would like to acknowledge and thank the Labcorp Safety Pharmacology team who contributed toward the great discussions and exploration of this data, as well as the CDISC SEND Safety Pharmacology SubTeam for their progress and work on the SEND model. Modelling in SEND Format Let's look at how this data has been modelled in the past. Previously this data was modelled by essentially 'forcing' the data into the current domain model. Figure 1 shows an example of how this data can be modelled in the current SENDIG 3.1	heatmap, controlled terminology, dose, cdisc	source	1
93	My Journey from SAS to R	Lucy Watson	Presentation	data science, missing value, proc freq, safety population, string functions, sample, proc step	source	15
94	No Compromise on QUALITY: Coding tips and Macros to improve Efficiency	Akhil Vijayan	When it comes to optimizing programming time, especially in ISE and ISS studies and in circumstances when there is a new data cut, programming reusability is favored by many organizations. But programmers still face issues while reusing the codes like truncations, handling variables with length more than 200, handling of QNAMs in SUPP domains, etc. In this paper, we will discuss techniques and best practices for reducing the likelihood of subtle errors in software reuse scenarios like: Coding tips to handle supplementary variables efficiently. Methods to avoid truncation in data Importance and examples of custom checks that programmers can implement to improve efficiency. The paper also discusses a few standard macros which support custom checks like Macro to ensure all variables with length more than 200 are successfully mapped in SDTM * Macro for automating SAS formats from P21 specification template	proc step, proc sql, cdisc	source presentation source	10
95	Non-inferiority Trial - Growing Significance	Anubrata Kundu	Presentation		source	23
96	One MDR, Multiple EDCs	Kairav Tarmaster	Presentation	life sciences	source	16
97	Overcoming Challenges of Your Integrated Summary ISS/ISE Submissions	Jenet Cp, Akhil Vijayan	The U.S. Food and Drug Administration (FDA) has advised that Integrated Summaries of Effectiveness and Safety (ISE, ISS) should be included in the Common Technical Document (CTD) while filing a New Drug Application (NDA). Both ISS and ISE are intended to provide a pool of databases that contain the results of all clinical trials. Integrating multiple studies from different phases of clinical trials can be challenging and it demands precise observation and planning. The most difficult and time -consuming part of the I SS/ISE process is creating the integrated datasets. Incorrect data integration leads to inaccurate analysis results that do not reflect the study drug's actual behavior	adam dataset, sdtm domain, controlled terminology, data cleaning, clinical trial, data science, adverse events, cdisc	source presentation source	7
98	Patient Narratives - Speeding Up Narrative Generation Using ODS Document Stores	Neville Cope	Patient narratives are a critical part of many clinical trials, and generating them programmatically with SAS for hundreds of trial subjects can take a long time. What if, instead of generating a narrative that processes many sentences and tables repeatedly for each subject, we could generate the sentence/table for all subjects just once and then manipulate all the outputs using `ODS` Document Stores? Well - you can! This paper will demonstrate techniques using `PROC DOCUMENT` to manipulate a collection of SAS outputs from one structure into another, speeding up the generation of a large numbers of narrative reports	proc step, proc report, adverse events, medical history, by group processing, sas output, clinical trial	source presentation source	14
99	Performance Traps in Multiple Imputation Analysis Using Mixed Model Repeated Measures	Piotr Karasiewicz	Presentation	covariance, proc step, proc mixed, fixed effects, change from baseline, degrees of freedom, unstructured covariance, imputation, secondary endpoint, correlation, sample, covariance structure, restricted maximum likelihood, data science	source video	17
100	Pharm R'Clock News: A Contributor's Guide to the {admiral} Package	Sophie Shapcott, Thomas Neitmann	Presentation		source video	7
101	PHUSE Education - Connecting the Dots	Sascha Ahrweiler	Poster		source	1
102	Practical Implementation of Good Programming Practice and Other Theories in Macro Development	Fred Ross	When putting theory into practice during the development of a macro, concepts and ideas might not always hold up as intended. This could be due to various reasons such as but not limited to: the nature of the project the macro is intended for and/or limited time to spend on the development of the macro. In a previous PHUSE presentation I have outlined concepts for Good Programming Practice and other ideas to follow or consider when developing macros. In this presentation I will explore these previously established ideas and the practicality of what worked and what didn't in a real world example	macro call, raw data	source presentation source	4
103	Predictive Analytics Framework for Diversity-focused Trial Enrolment: Synergistic role of Registry Data, Lived Experiences, and Programming Power	Kapil Khambholja, Vatsal Chhaya	Given the advocacy of clinical trial diversity in the recent past following the FDA guidance, stakeholder involvement has been the buzzword in regulatory decision-making. one of the several effective ways to ensure this includes accelerated use of real-world data, especially in complex disease contexts like rare diseases in oncology. This contains differential effects related to safety and effectiveness in different patient cohorts. Thus, predictive analytics of outcome registry data with stratification of the target population based on age, gender, and ethnicity may prove great trial enrolment insights. The present paper proposes the predictive analytics framework including synergistic use of standardized terms in disease ontologies, AI/ML capabilities, and insights from real-world experiential systems	clinical trial, predictive analytics, insurance claims	source presentation source	5
104	Preparing for the "Q" in RBQM: The Framework for Risk-Based Quality Management	Jason Konn, Alison Banda	Presentation	clinical trial, likelihood function	source	9
105	Processing Data and Documents with Natural Language Processing and Quantitative Anonymisation Strategies in Support of CTIS Disclosure Requirements	Elliot Zimmerman, John Lyons	Presentation	adverse events, clinical trial, phase ii, outlier	source video	25
106	Protecting the Intangible: The Impact of Culture on Business Outcomes	Diana Stuart	Presentation		source	9
107	Protocol Deviation Programming	Oussama Zineddine, Caroline Pelckmans	Presentation	artificial intelligence, clinical research, protocol deviations, sdtm domain, macro function, per protocol, statistical programmer, cdisc	source video	25
108	Python to Automate an AR Project Identifying Discrepancies across Fields from a Government Open-source Dataset of COVID-19 Vaccine Registrations	Aldir Medeiros Filho	This paper presents a Clinical Research Operations' case study & proof of concepti to assess if open-source software programming language and open-source datasets could be used as clinical operations managerial tools to deal with longitudinal 'big data' during clinical trials life cycles. Shares learning process and challenges of using open-source triad Google Collab's IDE & Pandas & Python, to conceive a Clinical Risk Based Monitoring project to identify, track AND resolve intra & inter discrepancies across government opensource datasets documenting distribution & application of vaccines against covid-19 covering a population of 7 million across 295 counties	clinical trial, outlier, excel, data science, missing value, big data, clinical research	source video presentation source	14
109	Quality Tolerance Limits: Statistical and Programming Consideration	Dhruba Sikdar, Nathalie van Borrendam	Quality Tolerance Limits (QTLs) as part of overall ARBM strategy is gaining traction in the industry and Health Authority (HA) alike. Predefined QTLs, taking into consideration medical and statistical characteristics of the variables as well as trial design are the cornerstone of this process. In the current QTL strategy at Janssen R&D, the QTL core team (Physician, Statistician, Clinical & Statistical Programmer, Trial Manager, Central Monitoring Manager) has the overall responsibility to define, monitor and report on the QTLs. The Statistical Programming is involved at the Define stage by providing specific programming inputs. During the Monitoring stage they are providing additional data for the study teams when a breach occurs. They are heavily involved at the Reporting stage to create CSR outputs. This paper discusses programming and statistical considerations to make while defining QTL parameters and corresponding thresholds	proc freq, clinical trial	source video presentation source	7
110	R for Submission - Experiences from the R Consortium Working Group	Ari Siggaard Knoph	Presentation	data cleaning, clinical programming	source video	16
111	R Package Development at Novo Nordisk	Anders Bilgrau	Presentation	data cleaning, macro definition	source video	40
112	R Shiny Application for Exploring Deep Sequencing Data	Annemie Van Dijck, Ilse Augustyns	Presentation	clinical programming	source video	32
113	R Shiny Apps: To QC or Not to QC?	Katie Withycombe	Presentation	adam dataset, sas output, data science, do over, independent qc, phase ii, proc freq, cdisc	source video	19
114	R Shiny to Drive Automation in Clinical Analysis and Reporting	Jocelyn Wang, Goutham Ghanta	Poster: R is becoming increasingly popular in the pharma industry, with collaboration on "PharmaVerse" developments where R Shiny is central. R shiny is used to share content with an interface enhancing the end-user experience. The intent with this poster is to show one application of R to inspire automation in the regulatory environment. All clinical drug projects are required to create periodic safety update according to a specific standard, which is a repetitive activity that can be highly automated and simplified. Below is an example of how R Shiny can be used to automate and simplify generation of Development Safety Update Report (DSUR). This implementation supports data import, metadata review and generation of the DSUR output automatically in a standardized way. It is assumed that data are highly standardized using the CDISC SDTM format	pharmaceutical industry, raw data, cdisc	source	1
115	RBM Beyond the Mandate: The Importance and Impacts of Risk-based Monitoring Today	Gina Wood, Alison Banda	Presentation	clinical trial, data cleaning, edc, proc freq, outlier	source	14
116	Reflections on the First Year with Git	Christian Olsen	Presentation	clinical programming	source	39
117	Release Your Creativity on Your Company Laptop	Johannes Ulander	Trying to be a creative programmer on a corporate laptop has its limitations. You have this great idea where you believe that you can automate a tedious task so that your team can free up resources for more important tasks, just to realize that the corporate software catalogue lacks the tools you need and you cannot download tools from internet. And if you cannot show that this actually works, no one will believe you. In fact, in most cases you just need a web browser and your company's cloud-drive (and a tiny bit of knowledge of JavaScript.) This presentation will show how I created a data template engine that displays forms and edit them, can collect data from subjects and generate SDTM domains. Only using what you most likely already have on your laptop today. A text editor, a web browser, and a cloud-drive	sdtm domain, bootstrap, cdisc	source presentation source	10
118	RESOLVE - A Lesser-known Powerful SAS Data Step Function for Retrieving Macro Variable Values	Navneet Agnihotri	The SAS macro facility is an extremely powerful tool of the SAS System. The SAS user can create macro variables from and within a `DATA STEP` by CALL SYMPUT routine and the values of the macro variable created in a `DATA STEP` can be retrieved by SAS `DATA STEP` functions SYMGET and RESOLVE. One would have very often encountered SYMGET function employed in macro program but very rarely could find reference of RESOLVE function in any macro program. Though RESOLVE function performs similar tasks as SYMGET function, it is however much powerful than SYMGET. It accepts a wider variety of arguments than SYMGET and resolves any macro expression unlike SYMGET which resolves only a single macro variable. This paper will compare and demonstrate the usefulness and relative ease of the lesser-known RESOLVE function compared to SYMGET function in retrieving values of macro variable(s)	macro variable, data step, &variable, macro call, call symput, macro facility, sas log	source presentation source	5
119	Risk-based Monitoring: Quality Tolerance Limits - Threshold Setting Methodologies	Nathalie van Borrendam, Annett Keller	Presentation		source	27
120	SanityCheckR: data quality control with R Shiny	Diana L. Santos Ferreira	Data exploration and validation is a critical step in any life sciences project workflow and performing it usually implies programming skills and writing many lines of code which can be time consuming. With the aim of empowering all team members, with and without programming skills, to interact in a quick and easy fash ion with data sets, we developed SanityCheckR an interactive application to explore, visualize and summarize SAS dataset files, using R language and the Shiny framework . The dashboard was customized to allow the user to interact with the dataset (for example, dynamically filtering variables) and provide easy access to the most common quality control tools, such as descriptive and pivot tables, plots to visualize variables and reveal missing data patterns , accelerating and simplifying the exploration/validation processes. In addition , this tool can be used in a live and interactive manner enabling more informative discussions in, for instances, data review meetings	missing value, proc means, quality control, data cleaning	source poster source	5
121	SAS and Python Working Together	Jim Box	Presentation	data science	source	18
122	SAS Macro Library Management and %library() Macro	Magnus Mengelbier	Presentation	macro definition, sas library, data cleaning	source	14
123	SAS Unit Validation of Your Macros to Upgrade Calibre!	Roshan Stanly, Siji Paul	SDTM, ADaM, and TFLs submissions to the FDA require standardized programming efforts. Few of these aspects on each of these tasks remain almost similar across studies. Macros can help in avoiding repetition of the same efforts and improving efficiency, but validating these macros before using them in live studies is an ordeal. SAS unit testing has been found to be the most effective method for validating macros. SAS Unit is a unit testing framework for SAS macros that controls the execution of test scenarios and generates documentation in HTML format. In this paper, we will be discussing the end-to-end process for the validation of SAS macros, starting from the installation of the SAS unit to the final documentation	data cleaning, numeric functions, macro definition, macro call, macro parameter, macro variable, cdisc	source presentation source	11
124	Scaling within a Fast-moving Biotech Company. Key Learnings for Leaders and Innovators in Our Life Sciences Industry	Stijn Rogiers, Jasmine Kestemont	"Build the plane while flying it" is probably a good quote for the journey ahead of us. Argenx is a fast-moving and growing Biotech company. This means we need to rapidly scale for future success while delivering on the ongoing clinical trials and submissi on milestones. In this paper, we focus on the journey so far as well as the future ahead of us. This includes an analysis of the argenx processes, talents, system landscape, and adoption of industry standards within Data Management. Next, mapping these to the future needs including integrations and automations, identifying potential gaps, and most importantly creating a business plan with realistic timelines for management approval. Finally, moving into the implementation phase, following up on the plan as agreed with leadership. 2	clinical trial, clinical research, data cleaning	source presentation source	11
125	SDTM Hackathon - To Find a Metadata-driven Software System that will Allow BMS to Automatically Generate Mapped SDTM Datasets	Rashad Rasul, Nicole Thorne	Poster	annotated crf, biostatistics, cdisc	source	1
126	Seeing the Full Picture with the R Shiny Data Visualisation Toolbox	Anastasiia Konvisarova		clinical trial, dataset	source presentation source	20
127	Set Up and Drive Your Routine Business Task with Power Automate	Satish Patil, Bhavin Solanki	Presentation	clinical trial, open label, phase i, proc freq	source video	33
128	Shiny-tise Your PROC COMPAREs: An Interactive Shiny Application for Data Review and Comparisons	Paul Jenkins	Presentation	unique key, dataset, numeric variable, proc compare, data cleaning, clinical programming, in=	source video	34
129	Shrinking the Atlantic - Leading in a Global Environment	Marc Drucker	Presentation		source video	12
130	Standardised SAS Macro Considerations for Submission-ready Code	Will Greenway	Standardised macros can be great but can often overload a user's log. Not all macro comments and processing need to be shown to a user where only the end result really matters. Giving users what they need, when they need it, can result in better quality macros, downstream processes and clearer logging. This paper is based on learnings from a "submission-ready code" process which outputs resolved code (without macro calls). The original suite of standardised macros generated resolved code containing meaningless processing steps and orphaned comments that were at best irrelevant, and at worst confusing or misleading. I intend to cover my experience with various SAS options and methods for controlling log and MFILE output, debugging considerations, and the impact of different SAS comment types, all in the context of "submission-ready" code. And, as always, there are hidden obstacles along the way	macro variable, &variable, macro call, options statement, macro definition, proc step, sas log, macro facility, macro loop	source presentation source	11
131	Standards Subject to Change: Data Standards Flexibility in an Evolving Landscape	Gina Wood, Alison Banda	Presentation	trial design model, excel, clinical data, data mining, label statement, clinical research, clinical trial, cdisc	source	28
132	Statistical Methods for Finding Human Bias in Machine Learning Models	Jim Box	Presentation	machine learning, data science	source	19
133	Streamlining the Usage of SAS, R and Python in our Process Workflow	Sayem Rahman, Shafi Chowdhury	Presentation	dataset, adam dataset, clinical trial, data cleaning, cdisc	source	8
134	Submission Readiness	Ali Dootson	Presentation	clinical data, data transformation, data lineage, data cleaning, analytics platform, clinical trial, data standardization	source	16
135	Submission Standards for RWD: The Good, the Bad and the Ugly	Jeffrey M. Abolafia, Sarah Ferko, Ingeborg Holt	Presentation		source video	33
136	Table of Contents (TOC) - An Application of PowerShell	Sumit Pratap Pradhan, Navneet Agnihotri	Trial Master File (TMF) is set up at the beginning of clinical trial. It is a collection of all essential documents pertaining to the trial. After each delivery (e.g. dry run), we are supposed to file documents like specification, qc plan, datasets, table, listing and figures on TMF. Uploading specification is easy as this is single file but uploading documents like datasets/tables requires to create a zip file and table of contents (TOC) because it is collection of multiple files. TOC is rtf file having listing of files included under Zip file. Manually creating TOC will require significant time. Since this is repetitive task, it can be automated to save the effort. PowerShell can be used here. PowerShell is a cross-platform task automation solution made up of a command-line shell, a scripting language. With PowerShell, an GUI will be developed which will ask key information from user and based on input provided, shell scripting and SAS will be used for generating TOC	adam dataset, sdtm domain, cdisc	source video presentation source	6
137	Technology Disruption in the Last Decade Can Inform Our Future Innovation	Chris Decker	Presentation	phase iii, clinical data, output delivery system	source video	19
138	Testing Times: Implementing CI/CD Test Automation Methodology for SAS	Paul Shannon	The benefits of DevOps practices including Continuous Integration and Continuous Delivery (CI/CD) are widely documented throughout the software sector. Significant savings in resource requirements can be achieved while simultaneously improving the quality of the end product. Cloud deployments with SAS Viya means CI/CD can be implemented for SAS Program code and SAS Platforms including such as a validated Statistical Computing Environment (SCE). Test automation can work to support validated SCEs, particularly with applying updates and subsequently the re-validation process. Using real-world examples, this paper demonstrates how test automation can be applied in a SAS programming environment. Consideration is given to differences for SAS 9.4 and Viya, mock data sources, third-party testing utilities and ultimately how much time can be saved using CI/CD	sas viya, data cleaning	source presentation source	7
139	The Development of CDISC Biomedical Concepts: Update and Next Steps	Bess LeRoy	Presentation	controlled terminology, data governance, vital signs, cdisc	source video	35
140	The Friendly Mapper - To Map Better Together!	Rajesh Saha	Presentation	data cleaning, clinical programming	source video	9
141	The Future-proof Data Fabric for Digital and Decentralised Trials	Venu Mallarapu	Presentation	data lake, data mart, data pipeline, machine learning, artificial intelligence, clinical data, clinical trial	source	12
142	The impact of migrating to SAS Viya 202X (aka SAS Viya 4)	Juan Sanchez	Currently, most of the SAS-based Statistical Computing Environments (SCEs) are built on the SAS 9 version; some on SAS Viya 3. Recently, SAS Institute released its modern cloud -native, unified version with multi -lingual programming capabilities, SAS Viya 202X (Viya 4). This journey to SAS Viya 4 brings specific challenges for GxP environments from a technical and change management point of view. Due to the compliance aspects, this transition can take from a couple of weeks up to 2 to 3 years , depending on the complexity of your overall statistical compute platform	sas viya, sas 9	source presentation source	13
143	The Integration Dilemma	Angelo Tinazzi	As of today, our Industry has not defined any approach, nor does an official regulatory agency preference / recommendation exists on how to integrate data of different studies to support either ISS/ISE. In 2020 PHUSE released a white paper, "Integration St rategies in Support of ISS/ISE Submissions" [2], where three integration approaches are proposed: 1. SDTM(s) → iADaM 2. ADaM(s) → iADaM 3. SDTM(s) → iSDTM → iADaM Over the last five years at Cytel we did experience all three options for several submissions. All three options have pros/cons and their applicability strictly depend on whether and how standards were applied in individual studies. The purpose of this presentation is to share experience we gathered by applying these three options, such as challenges and some technical tips we did apply to streamline for example the integration of several legacy studies into an iSDTM	adam dataset, reviewer guide, medical coding, adverse events, data integration, liquidity gap, clinical trial, open label, data cleaning, concomitant medications, define-xml, sdtm domain, data science, excel, clinical research, cdisc	source video presentation source	15
144	The M Files: I Want to Believe	Gregory Nunn	Presentation	mprint	source	13
145	The Real Benefit to Using Real-world Evidence Studies" Using RWE to Revamp Clinical Trials	Artur Sargsyan	In past decade the amount of clinical data generated outside of clinical trials has hit an unseen level. Having that data on hands enabled clinical researchers to step into new era of clinical trials based on real world evidence that could potentially be a game changing in pharmaceutical industry. This paper is intended to provide a brief discussion about the benefits of Real-World Evidence Studies conducted to enhance traditional Clinical trials	clinical trial, clinical data	source presentation source	5
146	The Role of Data Analytics Functions in Studies Utilising RWD for Submissions	Mark Graham, Ravi Bahia	Presentation	clinical trial, data science, data ingestion, open label, phase ii, raw data, clinical programming	source video	15
147	The Scrum Mindset	Florence Ratcliffe	Poster		source	1
148	The Secret to Getting Your Dream Job Offer	Tom Oades	Presentation	statistical programmer, clinical data, adam dataset, adam programmer, clinical trial, data standard, pd, cdisc	source video	13
149	The Table Has Turned	Catharina Dahlbo, Anna Berg	Poster		source	1
150	Time Goes Fast. Are You Really Ready for Unblinding?	Stijn Vermeylen	Presentation	clinical research	source video	23
151	Title:Pixel Plots: Identify Missing Data, See High-level Patterns and Break Axis Constraints	Paul Grimsey	Before starting analysis, it is useful to know if all your data is present and if there is anything unexpected or needing further investigation. By representing each received result by a single coloured cell, pixel plots can help you to quickly identify missing results and unexpected patterns in the data. This visualisation gives the Data Analyst / Scientist a very good feel for how much data is present and for which patient visits. Limitations of plot axes are overcome by exporting to `EXCEL`, which removes issues found when plotting too much data into static plots. This paper will show how to create pixel plots in R and SAS, along with template code for both languages for the basic visualisation, using a single SDTM dataset. More complex real life examples of the visualisation will also be presented, highlighting where this visualisation can be a very useful addition into the Data Scientist's toolkit	excel, sample, proc step, sdtm domain, proc report, clinical trial, proc sgplot, data science, cdisc	source presentation source	16
152	Trimmed Mean Analysis Is Not Rocket Science	Kateryna Fedoryshyna	Presentation	trimming, sample, adam dataset, change from baseline, response variable, missing at random, data science	source video	18
153	Tumour Assessment Evaluation Using MS Azure and Power BI	Robert Adams, Clara Beck	Presentation	pd, standard deviation, surrogate endpoint, missing value, clinical trial	source video	22
154	Using Individual-level Real-world Data to Inform Disease Area Data Models Employed for Clinical Trial Simulation and Design	Sheila McCartan	Presentation	clinical trial, trial summary, change from baseline, correlation, clinical data, baseline characteristics, data standard	source video	41
155	Using Janssen's ADaM Submission RPA Bot to Successfully Prepare for Delivery of an ADaM Submission Package	Andrew Noller	Poster	cdisc	source	1
156	Using Power BI to Present the QC Status for Studies	Spencer Renyard	PHASTAR hosts the QC trackers for studies on a SharePoint site, which has multiple benefits including multi-user edit access and version history. Power BI can access the QC tracker files and use them as data sources, providing the ability to analyse and summarise this data to provide information on the QC status of the studies in real time. With a centrally located, easy to use dashboard created with Power BI, the study or studies of interest can be selected, and a standard, validated set of summary tables and plots present the overall QC status for the study. This report can be used by the lead programmer and project manager to track the QC status and provide feedback to interested parties such as clients/customers and portfolio leads	excel, dataset	source presentation source	8
157	Using Python to Compile CSR	Loskutov Volodymyr		csr, clinical trial, create table	source video presentation source	4
158	Using R to Write Patient Narratives	Narek Torosyan	The Patient Narrative is an impactful document that summarizes and describes a subject 's clinical experience resulting from participation in a clinical trial. There are many ways to create this. Let us explore how we can create this document using R. We will identify the packages that are essential to write this program. We will see examples of code and why we chose R instead of traditional methods	clinical trial, adverse events	source presentation source	3
159	Validation Tools Assemble - Proc Compare No Longer Enough to Save the Day On Its Own!	Eoin Byrne, Peter Crumey	Statistical Analysis of any form relies heavily on the quality of the data, and this is paramount in clinical trials to ensure patient safety. This then puts a greater emphasis on the validation, as to ensure that all interested parties can trust the analysis. When a complete match happens and we see "NOTE: No unequal values were found. All values compared are exactly equal" it is easy to say that that is enough, but this is not always the case when using `PROC COMPARE` in SAS as not all complete matches are created equal	proc compare, proc step, data cleaning	source poster source	7
160	Veranet: R Shiny Meets Network Meta-Analysis and Data Visualisation	Anton Schönstein	Presentation	confidence interval, fixed effects, proc mixed, forest plot	source video	21
161	Vision and Demonstration of the OpenStudyBuilder for Supporting Structured Protocol Content and Submission Deliverables Using Concept-based Standards	Martin Gram, Mikkel Traun	Poster	cdisc	source	1
162	Visualisation for Success: Driving KPIs and Organisational Process Improvements via Portfolio-level Analytics	Philip Johnston, Julie Ann Hood, Sarah Angelo	Poster		source	1
163	Visualise Me! Down the Rabbit Hole with Standardised Dashboards for Safety and Study Follow-up	Jeroen Gielen, Stefan Verbeeck	Presentation	clinical research	source video	22
164	Welcome to the Pharmaverse	Michael Stackhouse	Presentation	data cleaning, demographics table, txt	source video	12
165	What History Teaches Us in Designing a Modern Statistical Computing Environment	Caroline Phares	Presentation	clinical programming, clinical trial	source video	11
166	What's the Difference?! CRO vs Pharma	Tom Ward, Piotr Karasiewicz	Poster	data cleaning, clinical research	source	1
167	Who is Subject X?	Vicky Poulsen, Henning Pontoppidan Föh	Presentation	clinical programming, clinical research, pharmaceutical industry, sas consultant	source video	26

Back to Conf go 2023 go 2021

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.