Discuss a pilot

RESEARCH & PUBLICATIONS · LDL-C

What does an LDL-C result support?

Our peer-reviewed studies examine how calculation methods and measurement origin affect LDL-C classifications near clinical thresholds.

These studies examine the underlying problem. They do not clinically validate GOVERN.

Browse the studies

FEATURED STUDY

A label on a laboratory record is not a record of how the value was made.

Detecting Friedewald-Substituted “Direct” LDL Cholesterol in a National Multi-Site Electronic Health Record Cohort: Prevalence and Benchmarking Impact

The question
Whether records labeled as direct LDL-C measurements in a large electronic health record cohort could contain calculated values.
Why it matters
A reported laboratory number may not fully describe how it was produced. Preserving measurement origin matters when clinical evidence is evaluated.
What it does not show
Retrospective study · Production-record authentication remains distinct
Read the paper The Journal of Applied Laboratory Medicine · 2026 · DOI 10.1093/jalm/jfag098

HOW WE WORK

From scientific questions
to clinical tools.

Each stage has a different job, and the claims we make depend on which stage the work has reached.

  1. 01

    Investigate

    Identify gaps in how laboratory evidence supports clinical decisions, using real-world data.

    Published · 3 studies
  2. 02

    Publish

    Put methods and findings through peer review, with limitations stated alongside.

    Ongoing
  3. 03

    Translate

    Build practical tools, with further validation before any claim of clinical effectiveness.

    In development

MORE PUBLISHED RESEARCH

The scientific work
behind the problem.

Each study is listed with the question it asked and what it does not show.

Peer-reviewed researchPublished finding

LDL-C equation disagreement as a threshold-integrity flag for orthogonal lipid confirmation

Journal of Clinical Lipidology · 2026

The question
Whether disagreement between established LDL-C equations near a treatment threshold can signal that a result deserves closer review.
Why it matters
Near a threshold, an LDL-C classification can depend on how the value was calculated.
What it does not show
Retrospective study · Routine comparator, not BQ
Read the paper DOI: 10.1016/j.jacl.2026.09.004
Peer-reviewed researchPublished finding

Equation Disagreement as a Zero-Cost Uncertainty Signal for LDL-C Classification: An Interpretable Regime-Aware Calibration Model Rescues Misclassification at Treatment Thresholds

The Journal of Applied Laboratory Medicine · 2026

The question
How often LDL-C classification at treatment thresholds changes with the calculation used, in large retrospective datasets.
Why it matters
Treatment thresholds are applied to single numbers. Knowing when that classification is fragile helps clinicians interpret results near the line.
What it does not show
Retrospective study · Routine comparator, not BQ
Read the paper DOI: 10.1093/jalm/jfag139
Independent editorialIndependent commentary

When the Reference Is Not a Reference: Hidden Substitution Bias in LDL-C Measurement and Its Implications for Electronic Health Record-Based Research and Risk Prediction

The Journal of Applied Laboratory Medicine · 2026

The question
Written by independent authors, not GOVERN: what hidden substitution of measurement methods means for LDL-C records used in research and risk prediction.
Why it matters
Independent experts describe why measurement provenance deserves attention in electronic health record research.
What it does not show
Editorial · Not an independent validation of GOVERN
Read the paper DOI: 10.1093/jalm/jfag136

AN HONEST EVIDENCE STANDARD

Three levels of evidence,
labeled as such.

A paper about a problem does not automatically validate software built to address it.

  1. Level 1

    Published scientific findings

    Peer-reviewed studies of the underlying problem.

    Available
  2. Level 2

    Capabilities under evaluation

    Functions tested against a defined method, with limitations stated alongside.

    Planned
  3. Level 3

    Clinically validated functionality

    Labeled this way only after evaluation, with results open to review.

    Not yet claimed
PRODUCT PILOT PARTNERSHIPS

Bring us the evidence question
your team needs answered.

Discuss a pilot, a research collaboration, or an evaluation with your team.

Discuss a clinical pilot Start with a conversation. No patient data required.