Drug-Disease Interactions: Examples, and the Checks a Drug Interaction Checker Misses

The short answer: A drug-disease interaction is when a medication that is perfectly safe in general becomes harmful because of a condition the patient already has. Ibuprofen is unremarkable until the patient has heart failure or chronic kidney disease. Pseudoephedrine is unremarkable until the blood pressure is uncontrolled. Most interaction checkers will not flag any of this, because they compare drugs against other drugs, not drugs against the problem list.

That blind spot matters more than the drug-drug one in practice, because drug-drug interactions are the thing everybody already checks. Drug-disease interactions get missed quietly, in patients who look straightforward, and they are a recognized driver of avoidable hospitalization in older adults. This page defines the term, gives a working table of the pairs worth knowing by heart, explains why the software usually stays silent, and sets out how to check for them in the time you actually have.

What is a drug-disease interaction?

A drug-disease interaction occurs when a drug worsens an existing condition, or when that condition changes how the drug behaves in the body. The medication does not have to be inappropriate in the abstract. It only has to be inappropriate for this patient, given what else is true about them.

The mechanisms fall into two broad groups. In the first, the drug directly aggravates the disease: a non-selective beta-blocker causing bronchoconstriction in asthma, or a dopamine antagonist worsening the motor symptoms of Parkinson disease. In the second, the disease alters the drug's handling, usually by impairing clearance, so a standard dose produces a higher exposure than intended. Reduced kidney or liver function is the common example, and it is why renal and hepatic dose adjustment is really a drug-disease problem wearing different clothes.

You will also see the terms drug-condition interaction and drug-disease state interaction. They mean the same thing. Some sources treat drug-age and drug-pregnancy considerations as a related category, since the same logic applies: something about the patient, rather than something in the medication list, is what makes the choice wrong.

What is an example of a drug-disease interaction?

The single clearest example is an NSAID in chronic kidney disease. Ibuprofen or naproxen inhibits prostaglandin-mediated afferent arteriolar dilation, the kidney loses a compensatory mechanism it was relying on, and glomerular filtration drops. In a healthy patient this is inconsequential. In a patient whose kidneys are already compromised, or who is also on an ACE inhibitor and a diuretic, it can precipitate an acute injury. Nothing on the medication list looks alarming, which is exactly the problem.

Here are the pairs worth recognizing on sight. This is a working list rather than a complete one, and every entry should be checked against the current labeling for the specific agent before you act on it.

Ciprofloxacin comes up often enough to call out separately, since people search for its drug-disease interactions specifically. Beyond the myasthenia gravis warning above, fluoroquinolone labeling flags tendinitis and tendon rupture (higher in older patients, those on corticosteroids and transplant recipients), QT prolongation in patients with existing risk factors, CNS effects including seizure lowering in patients with a seizure history, and aortic aneurysm and dissection risk in patients with existing aneurysm or relevant risk factors. Most of those are drug-disease interactions rather than drug-drug ones.

How is a drug-disease interaction different from a drug-drug interaction?

A drug-drug interaction involves two things you can both see on the medication list. Something inhibits or induces the metabolism of something else, or two agents push the same physiology in the same direction. Both halves are in the same document, which is why software handles them well: it is a lookup between two structured fields.

A drug-disease interaction has one half on the medication list and the other half somewhere else entirely, in the problem list, in a discharge summary, in a specialist letter, or only in the patient's memory. The software has to join two different data sources that were never designed to be joined, and the second one is frequently incomplete. That structural difference, not clinical difficulty, is why one gets flagged and the other does not.

Why most interaction checkers miss them

Open a typical consumer or reference drug interaction checker and it asks you to enter medications. That is the entire input. There is nowhere to say the patient has heart failure, so there is no mechanism by which it could warn you about the NSAID. It is not that the tool is wrong. It was never asked the question.

The tools that can check drug-disease pairs, including EHR-integrated clinical decision support, depend on the condition being coded in the chart. If heart failure is documented in a scanned cardiology letter but never added to the structured problem list, the rule does not fire. This is why so much of the value in drug-disease checking is really a documentation problem: when a new patient's history arrives as a stack of faxed records and PDFs, those comorbidities only reach the problem list if someone pulls the structured data out of the documents and enters it. Alert logic can only reason over what was recorded.

The third failure is alert fatigue. Systems that do fire drug-disease alerts often fire far too many, at too low a severity threshold, and clinicians override them reflexively within weeks. A warning that is dismissed without being read is functionally identical to no warning at all, which is a problem we look at in more detail in our piece on EHR drug interaction alerts and why they get ignored.

The Beers Criteria are the best published starting point

If you want an authoritative, maintained list rather than one clinician's table, the American Geriatrics Society Beers Criteria include an explicit section on potentially inappropriate medications due to drug-disease or drug-syndrome interactions in older adults. It is where the dementia, falls, heart failure, delirium and urinary retention pairings are set out formally, with the strength of evidence behind each one.

One correction worth making, because searches for it are consistently wrong: the current edition is the 2023 update. The publication that appeared in July 2025 was a separate companion document on alternative treatments to selected medications in the 2023 criteria, not a new edition. We keep a working version of the list in our Beers Criteria medication list, and the broader approach to older patients on long medication lists is covered in managing polypharmacy in older adults.

The Beers Criteria are explicitly a prompt for judgment rather than a prohibition. A medication on the list can still be the right call for a specific patient with a specific goal of care. What the criteria give you is the obligation to have thought about it.

How to check for drug-disease interactions

Four habits catch most of what matters, and none of them takes long.

Read the problem list before the medication list. The order matters. If you start from the drug, you evaluate it in isolation. If you start from the conditions, the risky choices announce themselves. Kidney function, heart failure, airway disease, cognitive status and bleeding history are the five that change prescribing most often.

Treat renal and hepatic function as a drug-disease check, not a dosing footnote. Most US labels state renal thresholds by Cockcroft-Gault creatinine clearance, while the eGFR your lab reports automatically is CKD-EPI indexed to body surface area. They are not interchangeable, and the gap is widest in very small, very large, elderly and low-muscle-mass patients, exactly the group where the decision is tightest. Our renal dose adjustment and hepatic dosing pages go through this properly.

Ask about the conditions the chart tends to miss. Falls in the last year, urinary symptoms in older men, glaucoma, and any airway disease. These rarely make it into a structured problem list and each one rules out a common drug class.

Check the whole picture at once, at the point you sign. Running an interaction lookup, then a contraindication check, then a renal dose check as three separate steps is where the process breaks down under time pressure, because the third step is the one that gets skipped.

That last point is what Prescriber.io is built for. You enter the drug or the scenario once and it checks for drug-drug interactions, flags contraindications and allergy blockers, surfaces renal and hepatic dose adjustments, and suggests guideline-based alternatives, together, with a cited source on each flag so you can confirm it. It is decision-support, not autonomous prescribing: every flag is a prompt for a licensed clinician who reviews it, verifies against official sources, and signs. If contraindication checking specifically is what you came for, our contraindication checker page covers how that works, and the drug interaction checker covers the drug-drug half.