Just updated details about future courses & events on the prognosis website, including Stats Methods for Risk Prediction (15 - 17 Oct, 2024) Stats Method for IPD Meta-Analysis (2-4 Oct, 20244) MEMTAB 2025 (30 April - 2 May, 2025) www.prognosisresearch.com/courses-events
**MEMTAB 2025 confirmed!** Methods for Evaluating Models, Tests & Biomarkers Uni of Birmingham, UK, 29th April - 1st May 2025 29th April: pre-conference courses 30th April: conference day 1 & evening event 1st May: conference day 2 Save the date! Abstract call begins July 2024
All papers now out in the BMJ series 'Evaluation of Clinical Prediction Models' part 1: from development to external validation tinyurl.com/yrpfhkuntinyurl.com/4jv7bnxetinyurl.com/n4v3ky73
ICYMI BMJ series: Evaluation of Clinical Prediction Models part 1: from development to external validation www.bmj.com/content/384/...tinyurl.com/4jv7bnxe part 3: calculating the sample size for external validation (out Monday!)
Agreed. Might be due to curricula of medical / grad schools that historically teach Sn/Sp/PPV/NPV so people are used to binary classifications
Just published: "Two‐stage or not two‐stage? That is the question for IPD meta‐analysis projects" - some claims pro 1-stage approach are misleading - 2-stage will often suffice unless data sparse - here we provide guidance for a broad audience Hope helpful! onlinelibrary.wiley.com/doi/10.1002/...
NEW #openaccesstinyurl.com/468m8buw Registration & protocols rare, data & code sharing uncommon, prediction models largely unavailable
Too bad I couldn't attend but this is an amazing development. We also need methods for proper calibration of such models.
Multiple postdocs in health data science at Stanford! I look for potential & skills (either algorithmic bias and fairness or microsimulation modeling); elite degrees & fancy publications not required. Pay range: $107,000-$115,000 goto.stanford.edu/rosepostdocs #stats #healthpolicy #academicsky
Value-of-information analysis provides a unique answer to the issue of the consequences of finite sample size when examining a prediction model in a new population. journals.sagepub.com/doi/full/10....