Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Prognostic modelling with logistic regression analysis: a comparison of selection and estimation methods in small

E W Steyerberg1, M J Eijkemans, F E Harrell

  • 1Center for Clinical Decision Sciences, Department of Public Health, Erasmus University, Rotterdam, The Netherlands. steyerberg@mgz.fgg.eur.nl

Statistics in Medicine
|May 3, 2000
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Navigating Fairness in AI-based Prediction Models: Theoretical Constructs and Practical Applications.

medRxiv : the preprint server for health sciences·2025
Same author

Prediction of transforaminal epidural injection success in sciatica (POTEISS): a protocol for the development of a multivariable prediction model for outcome after transforaminal epidural steroid injection in patients with lumbar radicular pain due to disc herniation or stenosis.

BMC neurology·2024
Same author

Methodological guidance for the evaluation and updating of clinical prediction models: a systematic review.

BMC medical research methodology·2022
Same author

Recall bias in pain scores evaluating abdominal wall and groin pain surgery.

Hernia : the journal of hernias and abdominal wall surgery·2022
Same author

Instrumental variable analysis to estimate treatment effects: a simulation study showing potential benefits of conditioning on hospital.

BMC medical research methodology·2022
Same author

A tailored intervention does not reduce low value MRI's and arthroscopies in degenerative knee disease when the secular time trend is taken into account: a difference-in-difference analysis.

Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA·2022

Developing prognostic models with limited data is challenging. Shrinkage methods in full models, incorporating external information, significantly improve prediction accuracy for outcomes like 30-day mortality in acute myocardial infarction patients.

Area of Science:

  • Biostatistics
  • Clinical Epidemiology
  • Health Informatics

Background:

  • Logistic regression models are used for dichotomous outcomes.
  • Selecting appropriate covariables and estimating coefficients in small datasets presents challenges.
  • Shrinkage methods can improve prediction accuracy in prognostic models.

Purpose of the Study:

  • To compare the performance of various selection and shrinkage methods for prognostic models in small datasets.
  • To identify optimal strategies for predicting 30-day mortality in acute myocardial infarction (AMI) patients.
  • To evaluate the impact of external information on model stability and quality.

Main Methods:

  • Compared backward stepwise selection (with various alpha levels and AIC) and external information for covariable selection.

Related Experiment Videos

  • Evaluated estimation methods including maximum likelihood, linear shrinkage, penalized likelihood, Lasso, and external information.
  • Assessed model performance on independent data using small datasets of AMI patients.

    • Main Results:

    • Stepwise selection with low alpha levels (e.g., 0.05) resulted in poorer model performance on independent data.
    • Full models with shrinkage methods applied to regression coefficients demonstrated substantially better performance.
    • Incorporating external information for both selection and estimation enhanced prognostic model stability and quality.

    Conclusions:

    • Shrinkage methods are recommended for constructing prognostic models in small datasets.
    • Full models with prespecified predictors and shrinkage are superior to stepwise selection.
    • External information integration improves the robustness and accuracy of prognostic models, especially with limited data.