How We Write About Blood Tests

Reference Sources

Bloodwork uses reference ranges and clinical thresholds from peer-reviewed medical guidelines and authoritative health organizations. These same sources form the foundation of our proprietary AI training data. Where a standard lab reference range differs from an optimal or functional range — as defined by the relevant clinical body — we explain both and why the distinction matters.

• NIH

  National Institutes of Health (NIH)

  National Library of Medicine and MedlinePlus reference data

• WHO

  World Health Organization (WHO)

  Clinical guidelines and international diagnostic criteria

• MCL

  Mayo Clinic Laboratories

  Reference interval database for all major biomarkers

• ADA

  American Diabetes Association (ADA)

  Glucose, HbA1c, and metabolic thresholds

• AAS

  American Association for the Study of Liver Diseases (AASLD)

  Liver enzyme and hepatic marker interpretation

• ATA

  American Thyroid Association (ATA)

  TSH, Free T3, and Free T4 reference ranges

• ACC

  American College of Cardiology / American Heart Association (ACC/AHA)

  Lipid panels and cardiovascular risk markers

How We Write

Our content is written with the precision expected of a clinical reference, not a general wellness article. These principles govern both our editorial team and the AI models we use to generate content and action plans.

• Specific thresholds, not approximations

  We do not write "levels above normal." We write "levels above 300 ng/mL" and cite the guideline by name.

• Mechanism over label

  We do not just name a cause. We explain how it raises or lowers the marker at a physiological level, in plain English.

• Clinical nuance preserved

  When a marker can be elevated for two very different reasons with different implications, we explain both and how to distinguish them.

• Every claim is traceable

  Sources are named in the references section of each page. We do not cite unnamed "experts" or "studies."

AI & Content Creation

Bloodwork uses proprietary AI models to generate educational content and build personalized biomarker action plans. These models are trained specifically on the clinical literature, medical guidelines, and reference interval databases listed in this document.

This is not general-purpose AI. Our models are built specifically for blood test interpretation — trained to surface what your individual results actually mean, not generic wellness information.

• Trained on clinical sources, not the open web

  Our AI models are trained exclusively on peer-reviewed medical guidelines, reference interval databases, and specialty association publications listed in this document — not general internet text.

• Personalized to your results

  Action plans are built by comparing your individual biomarker results against clinically established reference thresholds and surfacing only the interventions directly relevant to what your labs show.

• Held to strict editorial standards

  Model outputs are governed by the same writing framework described above — enforcing specific thresholds, named sources, and plain-language explanations. The model cannot produce vague claims or cite unnamed sources.

• Reviewed and refined by our team

  AI-generated content is continuously reviewed and corrected by our editorial team. As clinical guidelines are updated, training data and model outputs are refined to reflect the current consensus.

Medical Review

As content undergoes physician review, the reviewer's credentials, specialty, and review date will appear on each relevant page. Review status will be clearly indicated. Content not yet marked as reviewed should be treated as informational only.

Scope of Content

Bloodwork's content explains what blood test results mean in the context of established clinical guidelines. It is educational information, not medical advice, and does not replace a consultation with a qualified healthcare provider.

Action plans and recommendations generated by our AI models are informational tools based on population-level reference thresholds. They do not incorporate your full medical history, symptoms, medications, or the clinical context that a licensed physician would consider when interpreting your results.

If your results fall outside the ranges described, or if you have symptoms that concern you, speak to your healthcare provider before making any changes to your diet, supplements, or medications.

Content Updates

Clinical guidelines evolve. We review and update our content — and the training data underlying our AI models — when guideline changes are published by the relevant medical association, when new research meaningfully shifts the clinical consensus, or when inaccuracies in our content are identified. The last-reviewed date will be displayed on each educational page.

Corrections

If you believe a specific claim on Bloodwork — whether written or AI-generated — is inaccurate or contradicts current clinical guidelines, please contact us at support@bloodwork.app. We take every substantive correction request seriously and respond to credible clinical concerns.

Affiliates & Sponsorship

Some product recommendations surfaced by Bloodwork — including supplements, tests, or wellness products — may be subject to affiliate arrangements, under which Bloodwork earns a commission or fee if you make a purchase through a recommendation link.

Commercial relationships do not influence our AI recommendation logic. Products are recommended based solely on your biomarker results and the clinical guidelines described in this document. Bloodwork does not recommend a product because of an affiliate arrangement — affiliate arrangements are only established with products that already meet our editorial standards.