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

Pivoting approaches for bulk extraction of Entity-Attribute-Value data.

Valentin Dinu1, Prakash Nadkarni, Cynthia Brandt

  • 1Center for Medical Informatics, Yale University School of Medicine, PO Box 208009, New Haven, CT 06520-8009, USA. Valentin.Dinu@yale.edu

Computer Methods and Programs in Biomedicine
|March 25, 2006
PubMed
Summary
This summary is machine-generated.

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

Modulation of miRNA Signature in Human Adipose Tissue After 3 Months of ω-3PUFA Supplementation.

Cells·2026
Same author

Transmission Electron Microscopy Corneal Ultrastructure Study in Hematocornea of Corneal Transplant Graft.

Diagnostics (Basel, Switzerland)·2026
Same author

Tibial Artery Hemodynamics Predict Foveal Avascular Zone Enlargement in Type 2 Diabetes.

Diagnostics (Basel, Switzerland)·2026
Same author

Three Unrelated Children With Childhood Apraxia of Speech: Exome Sequencing and Functional Gene Analysis Imply a Role of Laminin-511 in Early Neurodevelopment.

Case reports in genetics·2026
Same author

A Scoping Review of Surgical Care for People Experiencing Homelessness With Cancer.

Journal of surgical oncology·2026
Same author

Leveraging the Electronic Health Record for Early Detection of Pancreatic Cancer Among 9.4 Million US Veterans.

Clinical and translational gastroenterology·2026

Transforming clinical patient data from Entity-Attribute-Value (EAV) format to a usable one-column-per-parameter structure is essential for analysis. This study benchmarks three pivoting algorithms, finding static SQL suitable for fewer attributes and hash tables for improved performance with more data.

Area of Science:

  • Biomedical Informatics
  • Clinical Data Management
  • Database Engineering

Background:

  • Clinical patient data is often stored in Entity-Attribute-Value (EAV) format, which is not directly compatible with many analytical tools.
  • Existing methods for pivoting EAV data are not well-documented or are outdated, posing challenges for data utilization.

Purpose of the Study:

  • To describe and benchmark three distinct algorithms for pivoting clinical data within a clinical study data management system.
  • To evaluate the performance and feasibility of different data transformation approaches for EAV clinical data.

Main Methods:

  • Development and implementation of three alternative algorithms for pivoting EAV data.
  • Benchmarking of the algorithms using a clinical study data management system.

Related Experiment Videos

  • Performance evaluation considering factors like the number of attributes and database load.

    • Main Results:

    • Static SQL is a feasible approach for pivoting when the number of attributes is relatively small.
    • A more complex algorithm utilizing hash tables and sufficient random-access memory offers improved performance by reducing database server strain.
    • The choice of algorithm depends on the specific data characteristics and system resources.

    Conclusions:

    • Pivoting EAV clinical data is crucial for enabling downstream analysis.
    • Static SQL provides a viable solution for simpler pivoting tasks.
    • Advanced techniques involving hash tables offer a scalable and performant alternative for handling larger attribute sets, optimizing database performance.