Eslam Maher

Just learnt the difference between & and && in R

Oncology Drug Development: Challenges of Increasingly Available Therapies

Important reflections and considerations for oncology trials by the FDA (2026) ascopubs.org/doi/10.1200/...

Big Book of R

www.bigbookofr.com

The Hochberg Procedure for the Comparison of Multiple End Points This Guide to Statistics and Methods provides an overview of the Hochberg procedure, an effective tool for avoiding false-positive conclusions in clinical trials reporting multiple end points.

The Hochberg Procedure for the Comparison of Multiple End Points jamanetwork.com/journals/jam...

I love how easy and flexible implementing this is.

BOP2-FE: Bayesian optimal phase II design with futility and efficacy-stopping boundaries The primary purpose of an oncology single-arm trial is to evaluate the effectiveness of anticancer agents and make a go/no-go decision while maintaining patient safety. We propose a flexible Bayesi...

BOP2 with both futility and efficacy analyses www.tandfonline.com/doi/full/10....

Leveraging generative AI to transform statistical analysis plan authoring in clinical trials - Rogier Landman, Birol Emir, Richard Zhang, Giannis Manousaridis, Yi-Chien Lee, Lili Jiang, Gordon Siu, Fe... Background Statistical analysis plans are critical regulatory documents that define the statistical methodology, objectives, and data-handling procedures for cl...

Leveraging generative AI to transform statistical analysis plan authoring in clinical trials journals.sagepub.com/doi/10.1177/...

Understanding Maximum Likelihood Estimation A tool to understand maximum likelihood estimation

rpsychologist.com/likelihood/

@glenmartin.bsky.social is giving an online seminar:

"Missing Data: The Missing Component in Guidance for Developing, Validating and Implementing Prediction Models"

📅 Date: Tuesday 10th March 2026
🕐 Time: 1–2pm GMT
📍 Location: Online

🔗 Register here to receive the Teams link: lnkd.in/ejk7EFzC

By "no" I meant Wilson CI does work great without adding pseudocounts indeed

Brown et al. (2001) has good discussion here (see section 3.1.2) www.jstor.org/stable/2676784

So they find that 0.95 CI Wilson's interval using z = 1.96, approximates to the simpler rule of "add 2 & 2 then use Wald's". Additionally, the midpoint of this "adjusted Wald" interval would be the same as Wilson's.

@richarddmorey.bsky.social they also mention the Beta(2,2) analogy!

No, not at all. I re-checked Agresti and Coull (1998), and they seem to "strongly" favor Wilson's but were looking for a simpler formula, like Wald's, to be presented in classrooms, but with better coverage. www.tandfonline.com/doi/abs/10.1... 1/2

Isn't this simply the Agresti-Coull intervals that are sometimes commonly reported? They seem to improve the performance intervals of Wald's in small n and extreme proportions by simply adding those pseudocounts. Love the Beta(2,2) prior analogy!

new package alert! {tidychain} is a #rstats packaged inspired by the below authors experience in showing how an excel file was changed / manipulated by looking at the underlying xml files to prove fraud in research

datacolada.org/109package

usrbinr.codeberg.page/tidychain/

How is it that a paper that claims to show how to do target trial emulation does not address confounding by indication and its ramifications for data collection? www.jclinepi.com/article/S089... #EpiSky #StatsSky

The `bmabasket` package used for Bayesian model averaging in basket trials was apparently removed from CRAN on 2025-12-22 cran.r-project.org/web/packages...

The fundamental problem of risk prediction for individuals: health AI, uncertainty, and personalized medicine Background: Clinical prediction models for a health condition are commonly evaluated regarding performance for a population, although decisions are made for individuals. The classic view relates uncer...

Love this preprint arxiv.org/abs/2506.17141

Developing Practical Method Based Guidance for Trial Statisticians for the Design and Analysis of SMART trials at Imperial College London on FindAPhD.com PhD Project - Developing Practical Method Based Guidance for Trial Statisticians for the Design and Analysis of SMART trials at Imperial College London, listed on FindAPhD.com

A great opportunity for a PhD studentship in Sequential Multiple Assignment Randomised Trials (SMARTs) at Imperial's CTU with @vcornelius.bsky.social www.findaphd.com/phds/project...

Evidence of Unreliable Data and Poor Data Provenance in Clinical Prediction Model Research and Clinical Practice Clinical prediction models are often created using large routinely collected datasets. It is essential that prediction models are developed with appropriate data and methods and transparently reported...

Evidence of Unreliable Data and Poor Data Provenance in Clinical Prediction Model Research and Clinical Practice www.medrxiv.org/content/10.6...

Major Chinese funder to stop paying fees for 30 pricey open-access journals Move comes amid effort to grow the country’s own journals

Major Chinese funder to stop paying fees for 30 pricey open-access journals www.science.org/content/arti...

The scary part of this is that we have created a #Statistics educational system that is guaranteed to create a low bar by teaching formulas, recipes, isolated stat tests, no experimental design, no Bayes, no questioning of assumptions, no questioning outcome variables or study designs, ... #StatsSky

gsDesign2: Group Sequential Design with Non-Constant Effect The goal of 'gsDesign2' is to enable fixed or group sequential design under non-proportional hazards. To enable highly flexible enrollment, time-to-event and time-to-dropout assumptions, 'gsDesign2' o...

gsDesign2 allows for designing trials with non-proportional hazards 🤩
cran.r-project.org/web/packages...

Einstein Circle Lecture | Nan van Geloven - News - BIH at Charité How can prediction models support clinical decision-making when treatment effects are already embedded in the data used to develop them? In this lecture, Nan van Geloven, Professor of Biostatistics at...

An important topic for real world studies IMO: Einstein Circle Lecture - (The Lack of) Causal Inference in Prediction Models

Feb 19, available via Zoom: www.bihealth.org/en/notices/e...

Mats Stensrud | Why reporting hazard ratios in medical research is a bad idea YouTube video by the Causal Inference Interest Group (CIIG)

200+ attendees joined us for this seminar with Mats Stensrud today. If you missed it, you can watch the recording here:
youtu.be/KL4eFZcw-sg

1/2 Nail in the coffin of dichotomania: @erik-van-zwet.bsky.social 's paper with @stephensenn.bsky.social and myself just published: onlinelibrary.wiley.com/doi/10.1002/... with extended discussion at discourse.datamethods.org/t/dichotomiz... #Statistics #StatsSky #rct #clinicaltrial

Application of External Control Arms in Oncology Drug Development - Friends of Cancer Research Tuesday, April 7, 2026 10:00AM – 2:00PM ET The Ritz-Carlton 1150 22nd St NW, Washington, DC 20037 Conference will also be streamed virtually Register Today Friends of Cancer Research (Friends) is conv...

Application of External Control Arms in Oncology Drug Development, 7 April

Virtual or in-person in Washington
friendsofcancerresearch.org/event/applic...

Evaluating Safety Rules in Ongoing Dose‐Escalation Trials Using BOIN: A Comparative Analysis of Criteria and Their Impact on Dose‐Level Safety Assessments Ongoing dose-escalation trials present unique challenges in assessing safety, all with the goal to establish the maximum tolerated dose (MTD) and/or the recommended phase II dose (RP2D). Due to the e...

Evaluating Safety Rules in Ongoing Dose-Escalation Trials Using BOIN: A Comparative Analysis of Criteria and Their Impact on Dose-Level Safety Assessments onlinelibrary.wiley.com/doi/10.1002/...

Haven't read the statistics diaries in the post, but I love the idea of it.

I remember this day and my fascination with what physicists do. Exciting stuff!