CocktailCheck

Drug interaction checker · Evidence-gated

Flag meaningful drug interactions — not alert noise

Check two or more medications for statistically gated interaction signals. Built for clinics and product teams that need fewer, clearer alerts.

CocktailCheck Live check

Type at least 2 letters to search the drug catalog.

Search and add at least two medications, then run a check.

Live check against the production API — only detected signals are shown (no “not found” noise).

Pair-wise
Every ingredient pair in the regimen
Pre-gated
Only interaction signals backed by enough real-world reports
30 days
Pilot — API or web app, days not months

How CocktailCheck works

One clear loop from medication list to ranked signals

  1. 01

    Enter medications

    Search brand or product names. We resolve each drug to its active ingredients.

  2. 02

    Check all pairs

    Every ingredient pair is evaluated against a pre-gated interaction dataset.

  3. 03

    See ranked signals

    Results return condition names and reporting frequency — only pairs that passed statistical gates.

The problem we solve

Traditional interaction databases flood clinicians with low-confidence alerts. Teams ignore them. CocktailCheck only surfaces pairs that already passed strict evidence gates — then ranks by reporting frequency so the workflow stays usable.

Statistical gates, not speculation

What must be true before a signal can appear

Gate Rule Meaning
Co-reported cases At least 5 reports Enough co-reports to avoid one-off noise
Proportional reporting ratio (PRR) ≥ 2× At least double the expected reporting rate
Lower confidence bound Above 1.0 Confidence interval stays above chance
Reporting frequency filter ≥ 10% of reports (default) Shows prominent signals; adjustable per request
High-confidence mode (optional) ≥10 co-reports and bound ≥1.5 Optional stricter filter for cautious workflows

Data foundations

Interaction signals come from public adverse-event co-reporting data (FDA FAERS). The drug catalog uses Health Canada product records. We do not claim separate FDA label validation.

CocktailCheck is a reference tool, not a substitute for clinical judgment. It is not a certified medical device and is not intended for use as clinical decision support.

Less noise than traditional lookups

Same safety intent — stricter evidence bar

Criteria Traditional databases CocktailCheck
Alert threshold Many low-confidence alerts ≥2× risk ratio, ≥5 co-reports, confidence bound above chance
What you see Severity labels from mixed rules Condition + reporting frequency from gated data
Delivery today Standalone lookup tools REST API + web demo; embed patterns via API
Time to try Long procurement cycles Free 30-day pilot

Integrate via REST API

Available today: drug search + bulk interaction check. Widget, EHR panels, and chatbots are integration patterns on the same API — built with your team, not turnkey connectors.

Available now

Web demo app

Ready checker UI for pilots and clinical evaluation.

Available now

REST API

GET /api/v1/drugs · POST /interactions/check — OpenAPI in the service.

Build with our API

Embed / portal

Call the API from your physician portal or chart sidebar.

Build with our API

EHR workflow

Trigger checks on prescribe/save via your integration layer.

Build with our API

Assistant / chatbot

Return structured interaction cards from the same endpoints.

Example response shape

Anonymous processing: send drug names or ingredient IDs only — no patient identifiers required.

{
  "interactions": [
    {
      "firstId": "ing_warfarin",
      "secondId": "ing_aspirin",
      "conditionName": "Haemorrhage",
      "reportingFrequency": 0.18,
      "caseCount": 28,
      "generic": false
    }
  ],
  "coverage": "FULL"
}

Privacy & buyer trust

What healthcare teams ask before a pilot

No patient health data in the check

Analysis runs on drug names and ingredient IDs only. We do not need patient charts to return signals.

Privacy & compliance (PHIPA / PIPEDA)

We discuss Canadian privacy expectations, data residency intent, and vendor review packs during pilot onboarding.

Reference tool — not clinical decision support

Statistical evidence gates for lookup and integration. Not generative AI. Not a certified medical device.

Pilot pricing

Start with a free 30-day pilot. Scale when the workflow proves out.

Pilot

Free

30 days

  • Web app + API access
  • Onboarding & support
  • Workflow setup help
Start pilot

Enterprise

Custom

API volume

  • Volume + SLA discussion
  • Security & compliance review
  • Partner-built EHR embeds
Talk to us

Request a pilot

Tell us your clinic, product, or integration path. We respond within one business day.