Network Update #64
Industry
Today, I would like to start with a very interesting post from FDA. Oncology Accelerated Approval Confirmatory Trials: When a Failed Trial Is Not a Failed Drug - via Journal of Clinical Oncology / American Society of Clinical Oncology (ASCO). FDA authors Gautam Mehta, MD, and Richard Pazdur, MD. Link
Another important step for open source in the pharmaceutical industry. FDA Approves Genentech’s Itovebi, a Targeted Treatment for Advanced Hormone Receptor-Positive (Link). And the submissions have been done in big part with open source pharmaverse packages. Thanks Ross Farrugia for your informative post! Link
R Consortium Exciting Development for R in Pharma! On September 20, 2024, the R Consortium’s R Submissions Working Group successfully submitted its latest test package—Pilot 4—to the FDA CDER via the Electronic Common Technical Document (eCTD) gateway. This package includes WebAssembly, which aims to simplify the review process for agency staff. Link
Programming
Jagadish K. - Principal Component Analysis (PCA) is a dimensionality reduction technique used to simplify large datasets while preserving the most important information. It’s particularly useful when the data has many variables, many of which may be correlated. PCA transforms the original variables into a set of new, uncorrelated variables known as principal components, which capture the largest amount of variance in the data. Read this post to understand the topic better. Link
Anna Wiksten and Tiina Kirsil? Same Same but Different: Leveraging R for Comparing Clinical Study TLF Versions. Large amounts of outputs are generated during a clinical trial. Whenever there are changes in the underlying data, expected or unexpected, the outputs need to be re-generated. In this paper we present a set of R functions that offers a powerful and flexible solution for comparing and reviewing clinical study output versions. Using these functions one can speed up the review of the outputs by quickly identifying the outputs that have and haven’t changed. Hence, no time is wasted for reviewing the outputs which have not been affected by data changes. Link
Maciej Nasiński A big shoutout to Maciej who decided to build a shiny app for the pacs package. The pacs package won the UseR poster session and got a lot of good feedback. One of the best R programmer Kirill Müller gave pacs package a github star, which was another motivation for me to work on the dashboard.
The R Dev Dashboard is a powerful tool designed to benefit thousands of R developers, package maintainers, and organizations. With features like package information, download trend analysis, version comparison, CRAN check results, and dependency analysis, this dashboard simplifies package management and provides essential insights. Link
Bartosz Jab?oński - The SAS Packages Framework, version 20241014, is ready. The framework is, as usual, available at GitHub. Release is "sponsored" by letters D and S, and the number 2, because the major enhancement added is support for PROC DS2 threads and packages. From now on two new types: DS2PCK and DS2THR, allow you to add PROC DS2 packages/threads code to your SAS package. Of course, documentation updated accordingly. Details about the release: Link
Sunil Gupta Which R function is the closest to data frame options? For filtering records and selecting variables, R programmers can apply the subset function or data frame options to get the same results! Link
I encourage you to see a post from Dr. Alexander Krannich on packages that help you with providing comprehensive summary of your data and why he uses gtsummary instead of tableone. Link
Andreas Ruf Missing Data with the powerful R Package mice. Dealing with missing data is one of the biggest challenges in data analysis in my opinion, but the R package "mice" (Multivariate Imputation by Chained Equations) has been a real game-changer in this field. Developed by Stef van Buuren, mice offers a robust framework to handle missing values, ensuring that the analysis remains accurate and reliable. Link
Hengwei Liu - How to create customized teal modules. Teal is developed by Roche and it has many modules for tables and figures. There are situations where you may want to create your own teal modules. I created a short video about two different ways of creating customized teal modules. if you are interested it is here. Link
Eric Nantz - Episode 182 of the R Weekly Highlights Podcast is out! Link
?? Nested tests in testthat (Roman Pahl
?? Post summaries with AI & Hugging Face Athanasia Monika Mowinckel
Biostatistics
Bayesian statistics for clinical research. Bayesian analysis combines previous information (represented by a mathematical probability distribution, the prior) with information from the study (the likelihood function) to generate an updated probability distribution (the posterior) representing the information available for clinical decision making. Owing to its fundamentally different conception of probability, Bayesian statistics offers an intuitive, flexible, and informative approach that facilitates the design, analysis, and interpretation of clinical trials. In this Review, we provide a brief account of the philosophical and methodological differences between Bayesian and frequentist approaches and survey the use of Bayesian methods for the design and analysis of clinical research. Ewan C Goligher MD, Anna Heath PhD, Michael Harhay Link
Yu Du Covariate Adjustment for Linear Models: Understanding FDA Advice on Standard Errors
Dr. Alexander Krannich 4 Regression Models You Should Know with R codes!
1?? Linear Regression
2?? Poisson regression
3?? Logistic regression
领英推荐
4?? Cox regression
See the full post with graphs and R code here - Link
Prem Kant Shekhar - Beginner Biostatistics Insight: Mixed Model for Repeated Measures (MMRM) in R using the mmrm package. In clinical trials, especially those with repeated measures over time, it’s crucial to handle both correlations between time points and missing data. This is where the Mixed Model for Repeated Measures (MMRM) comes in, making it an excellent tool for analyzing longitudinal data. Link
Real World Evidence
Fei Tang, PhD, MPH and Grace Hsu. Maximizing the Potential of Real-World Data with Bayesian Borrowing. Concerns about data quality in RWE generation, including bias and small sample sizes, call for advanced methodologies to enhance the robustness of RWE. Fei Tang, PhD, MPH and Grace Hsu discuss how Bayesian borrowing stands out as an approach that can significantly increase the scientific potential of Real World Data. Link
Nan Xiao - Group sequential trials in industry: a 30-year perspective. This post summarizes our recently published SBR paper, which examines real-world applications of group sequential designs in industry trials over the past 30 years - Some Group Sequential Trials From Industry Over the Last 30 Years (published in Statistics in Biopharmaceutical Research.)
In the paper, Keaven Anderson discussed key design considerations, challenges, and how open source R packages can support trial design. Link
Robert Szulkin Prediction of High Nodal Burden in Patients With Sentinel Node–Positive Luminal ERBB2-Negative Breast Cancer was recently published online in jama surgery. In this study we developed a prediction model to identify breast cancer patients with high nodal burden in the axilla. We developed a nomogram which could be used by clinicians and showed that our model is predictive of recurrence-free survival.
Why is this important?
In our previous publications we have shown that axillary lymph node dissection (ALND) does not improve recurrence-free survival (NEJM 2024, de Boniface et al), and we have argued that ALND should not be used to stage high nodal burden, pN2-3 (The Lancet Oncology 2024, deBoniface et al). However, it is desirable to identify patients with high nodal burden since they have a worse prognosis and could benefit from treatment CDK4/6 inhibitors. Our prediction model is based on clinical measures which do not require an ALND. Link
Events & Webinars
Join the #OpenStudyBuilder #community #meeting! ?? Date: 4.11. | ?? Time: 15:00-15:50 CEST | ?? Virtual Event Discover the power of OpenStudyBuilder through:
?? Q&A Session: Get expert insights and answers.
?? Lively Discussions: Engage on various subjects.
?? Interactive Demo:
IBIG Forum: Parma, 29-31 October 2024 (Chiesi Farmaceutici)
? Don’t miss out—there are only a few days left to secure your spot!
Join us from October 29-31 at Chiesi in Parma (?? Via Paradigna 131A, 43122 Parma) for three days packed with learning, networking, and the latest developments in biostatistics!
??This is the perfect chance to connect with peers, exchange ideas, and stay at the forefront of our field.
?? Pre-forum course registration: https://lnkd.in/dUkRryAu
?? Forum registration: https://lnkd.in/dFTbpVrF
?? Last call to register! Don’t wait—secure your place and be part of one of the most significant biostatistics events in Italy this year! Link
The online CAUSALab Methods Series at Karolinska Institutet is back!
Fall 2024 kicks off October 29 with a virtual talk from Jonathan Sterne (University of Bristol), Estimating COVID-19 vaccine effectiveness using linked English electronic health records in the OpenSAFELY secure analytics platform.
Register to attend the session ?? Link
Learn more about the virtual series: Link
Methods for Evaluating Models, Tests And Biomarkers. MEMTAB 2025: "Methodology That Stands the Test”
MEMTAB is the leading international conference about methods to evaluate models, tests & biomarkers for healthcare. It allows debate & dissemination of best methods for developing, evaluating & identifying reliable models, tests & biomarkers for use in clinical practice.?
In 2025, for our 7th International conference, we return to the University of Birmingham and raise the conference theme: "Methodology That Stands the Test”.
29th April - 1st May 2025 Link
Data scientist |Bio-Statistician | AI | Gen AI |LLM
4 个月Huge congratulations on your wedding and wishing you a lifetime of love and happiness!
Congratulations on getting married and for the work you do !
Father ?????? Founder of Better Biostatistics ??Life Sciences Educator ????
4 个月CONGRATULATIONS on getting married Krzysztof Orzechowski! Thank you for all that you do for the community and I wish you nothing but the absolute best during this time of celebration and as you being the next chapter of your life! My recommendation - be in the moment - your wedding will be the fastest day of your life - Try your best to take it all in and really have it in the front of your mind that the day is YOUR wedding day and enjoy it fully! All the best!