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

Automatic identification of variables in epidemiological datasets using logic regression.

Matthias W Lorenz1, Negin Ashtiani Abdi2, Frank Scheckenbach3

  • 1Department of Neurology, University Clinic Frankfurt, Schleusenweg 2-16, D-60528, Frankfurt/Main, Germany. Matthias.lorenz@em.uni-frankfurt.de.

BMC Medical Informatics and Decision Making
|April 15, 2017
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

Comparative safety of lipid-lowering drugs alone or in combination: insights from a systematic review and network meta-analysis.

Nutrition, metabolism, and cardiovascular diseases : NMCD·2026
Same author

Low-density lipoprotein cholesterol management and goal attainment at 1 year in patients with diabetes, pre-diabetes and normoglycaemia: results from the European, prospective, observational SANTORINI study.

BMJ open·2026
Same author

Efficacy and Safety of Bempedoic Acid in Patients Aged ≥75 Years Stratified by Varying Statin Exposure: Results From Phase 3 Studies of Bempedoic Acid.

Journal of the American Heart Association·2026
Same author

European Medicines Agency versus Italian Medicines Agency indications for icosapent ethyl: a long and winding road.

Nutrition, metabolism, and cardiovascular diseases : NMCD·2026
Same author

Personalized polygenic profiling based on the genetic architecture of lipid metabolism in the Russian population.

Frontiers in cardiovascular medicine·2026
Same author

Potential impact on cholesterol goal achievement and predicted cardiovascular risk by the addition of ezetimibe and bempedoic acid on top of statins: a simulation from the SANTORINI study.

American journal of preventive cardiology·2026

Automated variable identification for individual participant data (IPD) meta-analyses using logic regression showed poor performance. While feasible, the method requires backup strategies due to low positive predictive values for matching variables.

Area of Science:

  • Biostatistics
  • Epidemiology
  • Data Science

Background:

  • Individual participant data (IPD) meta-analyses require consistent data formatting across multiple datasets.
  • Manual data transformation is time-consuming and prone to errors.
  • Automated or semi-automated variable identification can improve efficiency and data quality.

Purpose of the Study:

  • To assess the feasibility and performance of logic regression for semi-automated variable identification in large-scale IPD meta-analyses.
  • To evaluate the effectiveness of using Boolean combinations of simple rules for matching variables.

Main Methods:

  • Manually created simple rules for each target variable.
  • Employed logic regression to find optimal Boolean combinations of rules on a construction subset of data.
Keywords:
Data managementEpidemiologyLogic regressionMeta-analysis

Related Experiment Videos

  • Validated the identified rule combinations on a separate validation subset.

    • Main Results:

    • Logic regression achieved an average positive predictive value (PPV) of 34% and a negative predictive value (NPV) of 95% in the construction sample.
    • In the validation sample, PPV was 33% and NPV was 94%.
    • PPV was 50% or less for a majority of variables (63% in construction, 71% in validation).

    Conclusions:

    • Logic regression is a feasible approach for data management in large epidemiological IPD meta-analyses.
    • The algorithm's performance, particularly its low PPV, is suboptimal.
    • Backup strategies are necessary to address the limitations of this method.