Newsletter for Statistical Programmers and Biostatisticians #57

Programming

Roman Gryzodub - In my quest to streamline and accelerate the workflow for Statistical Programmers, I've developed SAS Copilot / Clinical Trials (available with a premium subscription). This innovative tool, specializing in SAS programming and aligned with CDISC standards, offers practical, code-centric guidance for statistical analysis in clinical trials.Your feedback on this tool is immensely valued as I continue to refine and improve its capabilities. Link to the original post.

Colm Smyth - Introducing the Cliniva P21 Checker Tool, specially designed for Pinnacle 21 Community reports. We're thrilled to share this tool with the public forum, as it significantly reduces the review time between older P21 runs and the latest executions, offering essential functionalities along the way. Link

From Hengwei Liu

• In SAS programming we need to set up SAS macros to facilitate the creation of tables for clinical studies.In R programming we can create R functions for that purpose. This short paper shows how to setup a function that can calculate the count and percentage for some categorical variables in the tables forclinical trials. Link
• In solid tumor studies, programmers often need to create waterfall plot for the best percent change from baseline in sum of diameters. Occasionally they may get a request to display multiple indicator variables below the waterfall plot. Check this short paper to understand how to display these indicator variables in a scatter plot below the waterfall plot. Link

A big win for supporters of R in the pharmaceutical industry. Study clinical reporting and upcoming submission used end-to-end R with pharmaverse packages!

On December 4th, 2023, Genentech Announced Positive Phase III Results for Inavolisib Combination in People With Advanced Hormone Receptor-Positive, HER2-Negative Breast Cancer With a PIK3CA Mutation. As Levi Garraway writes, Today we (Roche and Genentech) shared pivotal trial results for one of our investigational medicines, which demonstrated a robust and clinically meaningful improvement in progression free survival for people with PIK3CA-mutated, hormone receptor (HR)-positive advanced #breastcancer. It has been known for many years that activating PIK3CA mutations drive tumor growth, disease progression and treatment resistance. The results of this study may bring a new milestone for people with this type of breast cancer and for #cancer #precisionmedicine as a whole. We're deeply grateful for the patients and investigators who participated in this study, and we hope to bring this potential new medicine to patients as quickly as possible. Link

PHUSE just released a white paper on ‘Implementation of estimands using Data Standards’ and is looking for industry feedback. ??Public review closes on 17 January 2023.The white paper consists of best practices for using existing standards for estimand implementation and extensions to data standards where applicable.The Estimand Implementation Working Group (#EIWG) is consolidating EFSPI (European Federation of Statisticians in the Pharmaceutical Industry) feedback, so please share any feedback with your companies’ EIWG representative or directly to PHUSE via [email protected]. Link

Manimaran Pandiyan - CDISC complaint mock ADLB dataset along with log file using R programming .lubridate and hms packages which helps to derive datetime and timing variables and admiral package which helps to derive baseline flag, analysis flag variables and base values .#ADaM & #R. - Link

Biostatistics

Tim Morris has published a post on Intuitive understanding of the re-randomisation design. The re-randomisation design allows individuals to participate in a study more than once and be randomised each time they participate. They must have completed follow-up for any previous periods before being re-randomised. The individuals’ needs, rather than the design, dictates how many periods of participation each individual contributes, so individuals will have different numbers of randomised periods. It’s crucial that they are not forced to switch from or to stick with their previously randomised treatment. The analysis does not necessarily adjust for participant; in fact, using mixed models for this purpose can induce serious bias. Link

Haitao Chu and his team published recently a paper Network meta analysis to predict the efficacy of an approved treatment in a new indication. Drug repurposing refers to the process of discovering new therapeutic uses for existing medicines. Compared to traditional drug discovery, drug repurposing is attractive for its speed, cost, and reduced risk of failure. However, existing approaches for drug repurposing involve complex, computationally-intensive analytical methods that are not widely used in practice. Instead, repurposing decisions are often based on subjective judgments from limited empirical evidence. In this article, we develop a novel Bayesian network meta-analysis (NMA) framework that can predict the efficacy of an approved treatment in a new indication and thereby identify candidate treatments for repurposing. We conclude by discussing an illustrative example in psoriasis and psoriatic arthritis and find that the candidate treatment has a high probability of success in a future trial. Link

Alex Ocampo - “Correlation is not causation.” This mantra has reigned supreme in the scientific community for the last century. Yet the next generation of statisticians, including myself, is pushing boundaries; we’re delving deeper than mere correlations, pioneering the way in making causal claims from our analyses.?Yesterday, I was given the opportunity to present on causality to a broad audience at Novartis. I shared our recent work (Jemar Bather) on leveraging causal graphs to identify casual treatment effects in clinical trials. This shift towards causality is crucial to pharmaceutical drug development, as we should aim to present concrete evidence that treatments are the #cause of better outcomes.?Curious about our work? Dive into our latest publication in Statistics in Medicine: https://lnkd.in/esRGYCdj?And join us in championing the causal revolution! Check the post here - Link

Two very interesting posts from Ryan Batten, PhD(c)

• Reflection on modern methods: demystifying robust standard errors for epidemiologists. What are they? When should I use them? How do I know I need them? What exactly are they robust to? For me, this article was a great introduction and answered those questions. Link
• Study Designs for Extending Causal Inferences From a Randomized Trial to a Target Population. Extending results from a randomized trial to a target population can be challenging. Dahabreh et al. (2020) helped me think about aspects of this. Why? Link

Statistical recommendations for count, binary, and ordinal data in rare disease cross-over trials. Recommendations for statistical methods in rare disease trials are scarce, especially for cross-over designs. As a result various state-of-the-art methodologies were compared as neutrally as possible using an illustrative data set from epidermolysis bullosa research to build recommendations for count, binary, and ordinal outcome variables. How to make use of the quite limited data in rare disease crossover trials in an efficient way? - here are some answers and much food for thought, resulting from the EBStatMax project funded by the European Joint Programme on Rare Diseases (EJP RD). Thanks to Georg Zimmermann - Link

Gary Collins - Evaluation of clinical prediction models (part 1): from development to external validation. Evaluating the performance of a clinical prediction model is crucial to establish its predictive accuracy in the populations and settings intended for use. In this article, the first in a three part series, Collins and colleagues describe the importance of a meaningful evaluation using internal, internal-external, and external validation, as well as exploring heterogeneity, fairness, and generalisability in model performance. Link

Real World Evidence

Bengt Hedin -A on-demand webinar from Cytel senior consultant. The Road to First-in-Human Trials: Insights from a Real-World Example. The journey to the first-in-human clinical trial for a drug project has many potential bumps and detours. Together with our friends at Dicot AB, we’ll discuss and give their real-world examples of the success factors of the journey towards a first-in-human study on their drug candidate, LIB-01. We’ll discuss the prerequisites for keeping project plans, delivery of results and dealing with the consequences of decisions made during the drug development.

By attending this webinar, you'll be better equipped to make your journey to clinical trials smoother – this is a must for C-level managers, project managers, scientists and regulatory experts. Link

Clinical artificial intelligence quality improvement: towards continual monitoring and updating of AI algorithms in healthcare. From Jean Feng, Rachael V. Phillips, Ivana Malenica , @Andrew Bishara, Alan E. Hubbard, Leo A. Celi, Romain Pirracchio, MD MPH PhD . Machine learning (ML) and artificial intelligence (AI) algorithms have the potential to derive insights from clinical data and improve patient outcomes. However, these highly complex systems are sensitive to changes in the environment and liable to performance decay. Even after their successful integration into clinical practice, ML/AI algorithms should be continuously monitored and updated to ensure their long-term safety and effectiveness. To bring AI into maturity in clinical care, we advocate for the creation of hospital units responsible for quality assurance and improvement of these algorithms, which we refer to as “AI-QI” units. We discuss how tools that have long been used in hospital quality assurance and quality improvement can be adapted to monitor static ML algorithms. On the other hand, procedures for continual model updating are still nascent. We highlight key considerations when choosing between existing methods and opportunities for methodological innovation. Link

Trainings, Webinars & Events

Advanced Exploratory Visualization Techniques - ggplot2, plotly & purrr. This training is part of a workshop at Open Source in Pharma by Omar ElAshkar (University of Florida) that covered hands-on advanced visualization concepts in R. The session introduced ggplot2 package which provides modelers and analysts with a flexible, modular way to generate almost any customized plots with ease. A review of coupling ggplot2 & plotly with map functions from purrr, R package specialized for functional programming, which provides a superb experience for users, was provided. Issued by Open Source in Pharma. Link

Revolutionize Clinical Trial Data Exploration - teal. This badge is part of a workshop at Open Source in Pharma by Dony Unardi (Genentech) that covered a {teal} workflow in a clinical trial setting and then took participants through a series of exercises to create teal apps with varying degrees of complexity. The workshop illustrated a sample of module templates for baseline, safety, & efficacy analyses, highlighted new features, & demonstrated how to implement them & also customized existing modules, & built a bespoke analysis module from scratch. Issued by Open Source in Pharma. Link

Biostatistics Network meeting. Karolinska Institutet. The next Biostatistics Network meeting will be on Thursday 4 April, 2024. This follows the tradition of events to allow biostatisticians and others working in associated fields to join together, listen to prominent thinkers, and form collaborative relationships. One of the keynote speakers will be Professor Chris Holmes, a?professor of Biostatistics in Genomics at the Big Data Institute of the University of Oxford. Chris holds a joint position at both the Department of Statistics and the Nuffield Department of Medicine. His most recent paper is titled “Where Medical Statistics meets Artificial Intelligence” (NEJM, 2023). Link

- Training Course: R for Clinical Trial Statisticians. Who is this event intended for??This course is aimed at clinical trial statisticians who are experienced with other software languages (such as SAS) and would like to increase their proficiency in R. The content can also be applied more widely in a non-clinical trial context.What is the benefit of attending??Attendees will learn how to: understand data manipulation in R via the “Tidyverse”; produce R graphics with ggplot2; and perform & interpret basic statistical analyses in R. Link