Drug Interaction Checker

What the Drug interaction checker does

This tool checks whether any pair of medications and supplements you select is flagged as interacting in a hand-built list focused on women's hormonal health. You tick items from a fixed list of 20 entries — oral and transdermal estradiol, micronized progesterone, semaglutide, tirzepatide, spironolactone, metformin, levothyroxine, sertraline, paroxetine, fezolinetant, warfarin, tamoxifen, lithium, and supplements including St. John's Wort, black cohosh, soy phytoestrogens, vitamin D, calcium, and iron. Select at least two, and it looks up every possible pair against 20 interaction entries and shows the ones it finds.

It is a fixed lookup table, not a model or a live drug database. There is no algorithm predicting interactions, no personalization, and no connection to a pharmacy record. Each flagged pair was written in advance with four fields: a severity tier, the effect, the mechanism, and a plain 'what to do' note — for example, the tool pairs oral estradiol with warfarin as Major, explaining that oral estrogen raises clotting-factor synthesis and can blunt warfarin, and suggesting transdermal estradiol or weekly INR monitoring as points to raise with a prescriber.

Every result is educational. The tool does not diagnose, does not tell you to start, stop, or change a dose, and does not replace a pharmacist or prescriber. Its own note recommends taking anything missing or uncertain to a pharmacist consult or a full database such as DrugBank or Lexicomp.

How to read your result

Results are sorted by severity, most serious first, across four tiers the tool defines: Major, Moderate, Minor, and Monitor. Major flags a pair the tool treats as one to avoid or manage closely (for example paroxetine with tamoxifen, where the tool notes CYP2D6 inhibition can reduce tamoxifen activation). Monitor means the pair is usually workable with a check — such as separating timing or rechecking a lab. Each card gives the effect, the mechanism, and a concrete next step, all written to be discussed with a clinician rather than acted on alone.

An empty result is not a clearance. If you select two items and see nothing, it only means that specific pair is not among the 20 entries in the list — not that the combination is safe. The tool says as much in its no-results message and points you to a pharmacist. Treat every flag, and every silence, as a prompt for a conversation, not a decision.

The most useful move after a flag is to bring the exact pair, the severity, and the suggested step to whoever prescribes or dispenses your medications. Where the tool suggests a route difference — for instance transdermal versus oral estradiol with warfarin or St. John's Wort — that is a question to ask, not a change to make on your own.

What the result is based on

The output is computed entirely from two hardcoded lists in the source: a set of 20 selectable medications and supplements, and 20 predefined interaction entries. When you select items, the code compares every pair against those 20 entries in both directions and returns exact matches only. There is no fuzzy matching, no dose input, and no scoring — a pair either appears in the list or it does not.

Severity, effect, mechanism, and recommendation text are fixed strings written into each entry, not generated per user. The four severity tiers (Major, Moderate, Minor, Monitor) exist only to label and order the pre-written entries. The page states its reference basis as FDA labels, Lexicomp, and peer-reviewed literature, and describes its own coverage as common hormonal-therapy interactions rather than a full drug database.

Because matching is pairwise, the tool never evaluates three-way or higher combinations as a group — it only reports the individual two-drug pairs among your selections that happen to be in the list.

What it cannot tell you

• Coverage is limited to 20 named items and 20 specific pairs. Any medication, supplement, or over-the-counter product not on the list — and any real interaction between two listed items that simply was not entered — will show nothing, which the tool explicitly warns is not proof of safety.
• It ignores dose, formulation strength, and how long you have been taking something. A flag like calcium or iron reducing levothyroxine absorption is the same text regardless of your doses; the tool cannot tell you how much your own levels are affected.
• It evaluates only pairs. If three or more of your selections interact as a combination, or if a third drug changes the risk of a flagged pair, the tool does not model that — it lists each two-drug match separately.
• It knows nothing about you — kidney or liver function, pregnancy, genetics (such as CYP2D6 status behind the tamoxifen flags), other conditions, or lab values — so it cannot say whether a listed interaction actually matters for your situation.
• The recommendation text is fixed guidance written into each entry (timing gaps, lab rechecks, route swaps). It is a starting point for a clinician conversation, not personalized medical advice, and the tool does not update it to your circumstances.
• Some entries reflect areas the field itself treats as uncertain — for example the soy phytoestrogen and tamoxifen entry is labeled controversial and black cohosh is described as a theoretical concern — so a flag can represent caution rather than a settled fact.