Ewout Steyerberg@ESteyerberg

Methodologist focused on prediction research from biostatistics, clinical epidemiology, and medical decision making perspectives

 Gouda, Zuid-Holland
universiteitleiden.nl/en/sta…
Joined April 2012
201 Photos and videos
Thanks for advertising. Validation / assessment of generalizability are tough topics, see also @f2harrell 's book and ncbi.nlm.nih.gov/pubmed/2598…
This tweet is unavailable
3
4
46
Very proud of my colleagues @MedicaInSpe and @CarlMoons and their consortia in obtaining two huge grants to bring AI forward in the medical area nwo.nl/en/news/four-consorti…
2
228
Ewout Steyerberg retweeted
Erik van Zwet@ErikVanZwet
19 Feb 2025
It has been claimed that there is no such thing as a validated prediction model. We provide empirical evidence for that claim in our new paper here onlinelibrary.wiley.com/doi/… with ao @Esteyerberg @DavidKlaveren see also @MaartenvSmeden @BenVanCalster @laure_wynants
1
9
31
3,519
Good to see @MaartenvSmeden talking on machine learning at the @stratosinit meeting in Leiden
16
1,347
Applications welcome, looking for two PhD candidates from outside of 🇳🇱 project 1: multidimensional statistical disease modeling of traumatic brain injury; project 2: acute stroke and subarachnoid hemorrhage. kuleuven.be/soprani/network lumc.nl/en/about-lumc/werken…
1
14
26
5,165
We are still searching candidates with interest in machine learning for time series data
2
2
1,311
Ewout Steyerberg retweeted
Ewout Steyerberg@ESteyerberg
15 Feb 2024
Applications welcome, looking for two PhD candidates from outside of 🇳🇱 project 1: multidimensional statistical disease modeling of traumatic brain injury; project 2: acute stroke and subarachnoid hemorrhage. kuleuven.be/soprani/network lumc.nl/en/about-lumc/werken…
1
14
26
5,165
Ewout Steyerberg retweeted
Erik van Zwet@ErikVanZwet
23 Dec 2023
New OA paper with Lu Tian and @robtibshirani. We show how to use the thousands of trials from the Cochrane database to improve the statistical inference for any given trial. We may expect gains that are roughly equivalent to doubling the sample size! onlinelibrary.wiley.com/doi/…

Evaluating a shrinkage estimator for the treatment effect in clinical trials

The main objective of most clinical trials is to estimate the effect of some treatment compared to a control condition. We define the signal-to-noise ratio (SNR) as the ratio of the true treatment...

onlinelibrary.wiley.com
3
27
99
17,000
Proud of @IRetelHelmrich defending her thesis on methods and applications of prognosis research. Great collaboration with @hesterlingsma, @DavidKlaveren, @CenterTBI
1
4
1,386
Enthusiastic start on an update of activities related to #AI at @LUMC_Leiden #CAIRELab and @LIACS
9
1,528
Great start of #weon2023 by @trishgreenhalgh talking about uncertainty and trustworthiness of #MachineLearning based tools for medical practice, moving on from Evidence-Based Medicine
1
7
24
6,375
Good to see the philosophical background to uncertainty highlighted. For decision making, Peirce wrote a brilliant 2 column Science paper in 1884, including Youden index and Net Benefit science.org/doi/10.1126/scie…
8
996
Ewout Steyerberg retweeted
Andrew Vickers@VickersBiostats
6 Jun 2023
Should you use confidence intervals and p-values for net benefit and decision curve analysis? (now with correct link!) diagnprognres.biomedcentral.…

Decision curve analysis: confidence intervals and hypothesis testing for net benefit

Diagnostic and Prognostic Research - A number of recent papers have proposed methods to calculate confidence intervals and p values for net benefit used in decision curve analysis. These papers are...

link.springer.com
2
16
67
10,130
We sometimes communicate risk as “we expect that out of 100 patients like you 5 will get disease X, but how do we communicate uncertainty?what if we said we base this prediction in the model on34 patients like you” #ESMDM23
1
3
14
2,168
Good start of Biennial European Conference | Society for Medical Decision Making smdm.org/meeting/18th-bienni… Humans poor at understanding randomness @Gaissmaier
1
6
1,354
Ewout Steyerberg retweeted
npj Digital Medicine
@npjDigitalMed
11 May 2023
"The generalizability of [clinical predictive] algorithms often goes untested, leaving the community in the dark on their accuracy & safety when applied to a specific medical setting." This commentary aims to provide clarity on generalizability objectives. nature.com/articles/s41746-0…
5
13
1,806
Ewout Steyerberg retweeted
CORE-MD@coremdproject
22 Mar 2023
Don't miss the third @coremdproject free webinar on Objective Performance Criteria on Monday 3rd April 2023 from 17.00 until 18.15 CET. Register here: bit.ly/core-md3
2
6
3,621
Polemic title, sensible content?@bds_lumc
Maarten van Smeden@MaartenvSmeden
24 Feb 2023
NEW PAPER Title says it all: there is no such thing as a validated prediction model doi.org/10.1186/s12916-023-0…
1
3
20
4,199
Ewout Steyerberg retweeted
Maarten van Smeden@MaartenvSmeden
13 Feb 2023
Proud of my humble friend (he would probably not announce this himself) and collaborator @BenVanCalster for becoming an Editor in Chief for the prestigious Statistics in Medicine journal
10
15
193
29,026
Great start of symposium in honor of @hesterlingsma her inaugural lecture by @Tufts_PACE on heterogeneity of treatment effect from a simple country doctor's perspective 😂@ErasmusMCPH
2
1
25
3,652
Ewout Steyerberg retweeted
Hein Putter@hein_putter
30 Nov 2022
Additive hazards model: ever wondered what to do about negative hazards and why the default method is ordinary least squares and not maximum likelihood? Then this will be of interest! @bds_lumc Maximum likelihood estimation in the additive hazards model onlinelibrary.wiley.com/doi/…

Maximum likelihood estimation in the additive hazards model

The additive hazards model specifies the effect of covariates on the hazard in an additive way, in contrast to the popular Cox model, in which it is multiplicative. As the non-parametric model,...

onlinelibrary.wiley.com
4
17
Load more