Skip to content
PharmaGuide

Methodology

The science behind the guidance.

How we collect, verify, and ship the interaction data that powers PharmaGuide — including who reviews it, what AI does and doesn't do, and where we draw the line.

01 · Why methodology matters

In a market full of opinions, we show our work.

  • 01

    Traceability

    Interaction guidance is traceable to regulatory sources, peer-reviewed publications, or professional references. No anonymous claims.

  • 02

    Accuracy

    Content is verified against primary sources and reviewed by licensed clinicians before it ships. Updates are dated and visible.

  • 03

    Accountability

    We say what we do — and what we don't. The boundaries are explicit. No marketing-speak hiding the limits.

02 · Where our data comes from

Four primary sources. No anonymous claims.

  • FDA

    FDA Resources

    • Dietary Supplement Label Database (DSLD)
    • Adverse Event Reporting System (FAERS)
    • Warning letters and safety alerts
    • Current Good Manufacturing Practice (cGMP) records

    Product identification · safety alerts · manufacturing disclosures

  • NIH

    NIH Resources

    • Office of Dietary Supplements (ODS)
    • National Center for Complementary and Integrative Health (NCCIH)
    • DailyMed (drug labeling)
    • Dietary Supplement Ingredient Database

    Ingredient information · efficacy research · safety profiles

  • PUB

    Clinical Literature

    • PubMed / MEDLINE
    • Cochrane Library systematic reviews
    • Journal of Dietary Supplements
    • Drug-nutrient interaction studies

    Interaction mechanisms · clinical evidence · contraindications

  • PRO

    Professional Resources

    • Natural Medicines Database
    • Clinical pharmacology references
    • Drug interaction databases
    • Professional society guidelines

    Cross-referencing interactions · clinical context

We do not access medical records, pharmacy systems, or any personal health data outside what you put into the app.

03 · How the engine works

From a scanned product to a verdict.

Every check runs through the same six-stage pipeline — most of it on your device. Here's what happens between the scan and the score.

  1. 01

    Catalog match

    A scan or search resolves to a specific product in a 180,000+ item catalog that lives on your device — so identification happens in milliseconds, offline, without a round-trip to a server.

    180,000+ products · on-device · sub-10ms lookup

  2. 02

    Ingredient normalization

    Active and inactive ingredients are parsed and mapped to canonical identities — and proprietary blends are decomposed into estimated per-ingredient ranges, the step most apps skip because the label hides the math.

    Actives + inactives · proprietary-blend decomposition

  3. 03

    Interaction detection

    Your stack is analyzed as a system, not one bottle at a time: supplement–supplement, supplement–drug, and drug–drug interactions, cross-product dose summation, and timing conflicts are each checked against the rules library.

    Multi-way · dose summation · timing conflicts

  4. 04

    Personal gating

    Every result is filtered through your profile — conditions, medications, age, and goals. Irrelevant warnings stay hidden; a medication that's risky given one of your conditions is surfaced as a flag.

    Condition · medication · drug↔condition flags

  5. 05

    Accumulation tracking

    Each nutrient is summed across the entire stack and measured against its Recommended Daily Allowance and Tolerable Upper Intake Level — so three products quietly stacking the same mineral don't slip past the safe ceiling.

    RDA + UL per nutrient · across the full stack

  6. 06

    Recall & safety sync

    The products you've scanned are checked against active FDA recalls and adverse-event signals, so a bottle in your stack doesn't quietly become unsafe without you hearing about it.

    FDA recalls · FAERS signals · lot-level when published

The six stages resolve to one verdict

OptimizedSolidDecentNeeds reviewUnsafe

See it in action

One real interaction, end to end.

The card below follows the same structure used by our interaction-rules pipeline — same sources, same evidence levels, same reviewer signature.

Specimen · ATORVASTATIN ↔ CoQ10

InformationalEvidence · Established

Mechanism

Statins inhibit HMG-CoA reductase, blocking not only cholesterol synthesis but also the mevalonate pathway used to synthesize CoQ10. Plasma and muscle CoQ10 levels are measurably reduced by statin therapy; CoQ10 depletion is proposed as a contributing mechanism to statin-induced myopathy.

Recommended action

CoQ10 is commonly discussed for statin-associated muscle symptoms. Ask your prescriber whether supplementation fits your situation.

Sources (2)

Reviewed by Laurie Pham, PharmD

Schema v6.1 · DB sync 2026-05-06

Want the deeper read? Read the statins + CoQ10 deep dive →

04 · How we verify content

From source to verdict, every interaction.

  1. 01

    Source identification

    We start from authoritative sources only — peer-reviewed, government-published, or recognized by professional clinical societies. No blog posts, no influencer claims, no unverified registries.

  2. 02

    Data extraction and analysis

    Information is structured into ingredient profiles with explicit citations. Each interaction is mapped to its mechanism, evidence level, and clinical context.

    Output · Structured ingredient profiles with cited sources
  3. 03

    Pharmacist verification

    A licensed clinical pharmacist reviews interaction guidance before it ships — checking the science, the framing, and the recommendation against current practice standards.

    Reviewer · Laurie Pham, PharmD
  4. 04

    Clinical advisory review

    A second reviewer reads from the patient-education angle: is the language clear, the framing accessible, the safety context unambiguous?

    Reviewer · Miriam Farez, NP
  5. 05

    Publication and ongoing monitoring

    Once published, content goes into an ongoing review cycle. FDA alerts, new research, or regulatory changes trigger immediate updates.

    Schedule · Quarterly systematic reviews + priority alerts as needed

05 · Medical advisory team

Clinical review behind the guidance.

  • Laurie Pham, PharmD

    Laurie Pham, PharmD

    Doctor of Pharmacy

    PharmD · 15+ years clinical pharmacy

    Focus · Drug-supplement interactions · pharmacovigilance · clinical accuracy review

  • Miriam Farez, NP

    Miriam Farez, NP

    Nurse Practitioner

    NP · integrative health practice

    Focus · Patient education · integrative health · content accessibility

06 · AI transparency

AI does some of the work. Humans do the rest.

We use AI to scale the parts that benefit from scale. We use clinicians for everything else. Here's exactly where the line is.

What AI does

  • Process and structure data from multiple sources
  • Identify potential interactions documented in the literature
  • Generate initial drafts of educational content
  • Power the AI Guidance Chat for in-app questions

What AI does not do

  • Generate original research or clinical recommendations
  • Diagnose medical conditions or symptoms
  • Recommend specific treatments or dosages
  • Access personal health records or medical history
  • Replace the judgment of a licensed healthcare professional

07 · Limitations and scope

Built to inform. Not to replace.

PharmaGuide is

  • An educational information platform
  • A tool to support informed conversations with your healthcare provider
  • A resource for understanding documented interactions
  • Reviewed by licensed healthcare professionals

PharmaGuide is not

  • A medical device or diagnostic tool
  • A replacement for professional medical advice
  • A source of treatment recommendations
  • A guarantee of supplement safety or efficacy

Always consult a qualified healthcare provider before starting, stopping, or changing any supplement or medication.

Want to see it in action?

The methodology is the foundation. The product is the proof.