Polypharmacy Signal Atlas
MedTangle
List your medications and MedTangle scans every pair in the regimen, ranking which get co-reported with serious reactions in the FDA’s public data — worst first, as screening signals to raise with a pharmacist, never risk or advice.
22 drugs231 precomputed pairsopenFDA FAERS via api.fda.gov (public, no PHI)updated Jul 2, 2026
Strongest signals in the 22-drug atlas
The 114 pairs carrying the rarer interaction pattern lead the ranking. Click any to explore it, or list your own medications above.
- n=186
- n=97
- n=195
- n=188
- n=367
- n=215
- n=235
- n=753
These are screening signals from spontaneous FAERS reports — confounded by co-prescription, not measures of risk, causation, or incidence, and not medical advice.
Does it actually work?
Recovering known interactions
MedTangle independently flagged 11 of 16 textbook drug interactions clinicians are taught to watch for — from the FDA’s raw adverse-event reports alone, across a 22-drug demo slice, with no interaction rules built in. 6 reached the rarer interaction-pattern tier.
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Expected:
- Apixaban & Fluconazole
Fluconazole (CYP3A4/P-gp inhibition) raises apixaban levels and bleeding risk.
✗ not flaggedExpected: - Simvastatin & Amiodarone
Amiodarone inhibits simvastatin's metabolism (CYP3A4); the FDA warns to limit the simvastatin dose.
✗ not flaggedExpected: - Amiodarone & Ondansetron
Two QT-prolonging drugs — additive risk of a dangerous heart rhythm.
✗ not flaggedExpected: - Lisinopril & Trimethoprim
Trimethoprim has a potassium-sparing effect; with an ACE inhibitor it can raise potassium dangerously.
✗ not flaggedExpected: - Methotrexate & Trimethoprim
Both are antifolates; together they can suppress the bone marrow and drop all blood-cell counts.
✗ not flaggedExpected:
The 5it didn’t flag are cases where that exact reaction had too few co-reports in this 22-drug demo slice to clear MedTangle’s conservative threshold. It prefers a miss to a false alarm — a false negative, not a false positive. More would surface on the full FDA drug set. That restraint is the point: it is what makes the 11 it did recover worth trusting.
Reading the scale
Signal strength
Bars and the gauge encode omega 0.025 — the conservative lower bound of a disproportionality score. Calm blue-grey means near the noise floor; amber to deep coral means increasingly co-reported beyond chance.
Co-reporting signal
A reaction shows up with this drug pair more than the database would produce by chance. It is confounded by co-prescription — a reason to look closer, not a verdict.
Interaction pattern
The rarer, stronger tier, shown in violet: the co-reporting excess sits beyond what each drug shows on its own.