Art.

This does not scare me, for I laugh in the face of arbitrary thresholds.

We wrote an article explaining why you shouldn't put several variables into a regression model and report which are statistically significant - even as exploratory research. bmjmedicine.bmj.com/content/4/1/.... How did we do?

Happily, we've had faster progress with another recent paper, just published! We use #CUREd data to show quite dramatic differences in how clinical & non-clinical NHS 111 staff handle calls about children: adc.bmj.com/content/earl... #NHS111 #BMJ #NHS @scharr-msg.bsky.social @arc-yharc.bsky.social

I've had a revised article sitting on the editor's desk at BMC Health Services Research, following really positive first reviews, for over 4 months. I can't get more than a stock response from editors. Anyone else in same position? Help, please?! @springernature.com

The first 12 months of our project evaluating same-day emergency care (SDEC) is nearly over. You can read about how the project is progressing on our website: sites.google.com/sheffield.ac... @fohsheffield.bsky.social @scharr-msg.bsky.social @arc-yharc.bsky.social #SDEC #NHS #NIHR

12 hour waits for a bed in the ED used to be a red line, never to be crossed. Now it is a just a part of life in ED. For every 72 patients waiting 8-12 hours in ED there will be one excess death. This is still unnacceptable. @rcem.bsky.social #resuscitateemergencycare rcem.ac.uk/almost-300-d...

New paper published in Statistical Methods in Medical Research: Comparison of statistical methods for the analysis of patient-reported outcomes in randomised controlled trials: A simulation study journals.sagepub.com/doi/10.1177/...

My colleagues @beckiesimpson.bsky.social, @richardjacques.bsky.social and I had fun at #EUSEM, though we are statisticians out of water among ED clinicians! We presented our work with NHS routine data: understanding avoidable admissions, winter pressures, and advisor skill-set & NHS111 performance.

A question for the #CausalInference community: best ways to include fairly generic measures of frailty and comorbidities (probably as a count) in a causal #DAG? #statsky #episky

Working with Taylor & Francis is simply unethical by this point. It’s unfortunately difficult to have AI use revoked from existing contracts but at least try, complain, if you have a book with them. And otherwise do not even consider.

My first... er, post(?) on here is a delightful one! Our new paper, a review of adaptive clinical trials and how their sample sizes are calculated and reported is live at BMC Medical Research Methodology. Unfortunately my fellow authors are not here yet.... #statsky #adaptivedesigns #clinicaltrials