Checking a drug interaction

The situation: a new prescription and a possible interaction

A pharmacist checking a new prescription, or a nurse giving a first dose on a ward, notices that the patient already takes a drug that may interact with it. The interaction checker in the dispensing or prescribing system has flagged it, often alongside several lower-priority alerts, and the question is what to do: whether the combination is contraindicated or only needs monitoring, what the mechanism is, which alternative avoids it, and how to explain it to the prescriber and the patient. A classic example is a patient on simvastatin who is prescribed clarithromycin, a strong CYP3A4 inhibitor that raises simvastatin levels and with them the risk of myopathy and, rarely, rhabdomyolysis. Other common questions involve QT-prolonging combinations, serotonergic drugs, anticoagulants, drugs cleared by the kidney and medicines whose levels are monitored. They often sound like these:

• Patient on simvastatin is prescribed clarithromycin: what is the mechanism, what does the label say, and what are the alternatives?
• Two QT-prolonging drugs together: which patient factors raise the risk, and what monitoring is recommended?
• Which herbal products interact with common antidepressants, and how?

What goes wrong without a cited source

Without a cited source, an interaction decision rests on an alert or on memory, and both have known weaknesses. Interaction alerts in clinical systems are numerous, and many clinicians override most of them, a pattern known as alert fatigue, so a serious alert can be dismissed with the trivial ones. Different interaction resources can grade the same pair differently, and an alert rarely explains the mechanism or the alternatives. Memory covers the classic pairs but not the newer drugs or the local brand a patient brought from another country. A general web search returns consumer pages, and a general AI chatbot may give a fluent explanation without naming the label it relies on. The cost is either a missed interaction that harms the patient, or an unnecessary change that confuses the prescriber and the patient. A label citation settles what the manufacturer and regulator say.

• Alert fatigue. The serious alert is dismissed with the trivial ones.
• Conflicting gradings. Resources disagree on how severe the same pair is.
• Unknown local brands. A brand name from another country hides the active ingredient.

How AskMed helps with an interaction

AskMed explains an interaction with sources, from the labels first. For a drug question, AskMed is instructed to cover the class, mechanism, uses, common and serious adverse effects, named interacting drugs, contraindications and monitoring. Before writing, it searches a fixed list of medical sources: FDA and DailyMed drug labels; guideline bodies and agencies such as NICE, WHO and CDC; journals such as NEJM, JAMA and The BMJ; and references such as StatPearls and the MSD Manual. For local brand names, AskMed also looks in official drug registries from many countries, such as the EMA and the UK's medicines.org.uk, to identify the active ingredients. Every claim links to its source, and the source list labels each one Drug label, Guideline, Research or Reference. AskMed explains how dosing is adjusted for an interaction in principle, but never gives amounts.

• Drug labels only. Turn off other source types in the filters when you want the label view.
• Counselling. A follow-up such as "explain this for the patient in plain words" keeps the topic and the sources.
• Related tools. The QTc calculator and eGFR calculator help with QT and kidney questions.
• AskMed works on iPhone and iPad, Android and the web, with one account on all of them.

Checklist: checking a drug interaction

The checklist below is a structured way to check an interaction and act on it, used alongside your pharmacy's interaction database rather than instead of it. It starts with a full medicine list, because many missed interactions involve a product nobody thought to mention, such as an over-the-counter painkiller, St John's wort or a medicine bought abroad. It then moves from the alert to the evidence: the mechanism, what the label says and how strong the evidence is, then to the patient: their kidney and liver function, age and other risk factors, and finally to a decision and counselling. AskMed fits the evidence steps and works on iPhone, iPad, Android and the web, so it is available at the dispensary computer or on a ward round. Every step that involves an amount ends at the prescribing information or your formulary, never at AskMed.

1. Take a full medicine list. Prescriptions, over-the-counter products, supplements, herbal products and medicines bought abroad.
2. Run it through your interaction resource. Note every flagged pair and how it is graded.
3. Ask AskMed about the pairs that matter. Mechanism, label wording, severity, alternatives and monitoring, with sources.
4. Open the label citation. Check whether the label says contraindicated, avoid or monitor.
5. Weigh the patient's risk factors. Kidney and liver function, age, QT risk and other drugs on the same pathway.
6. Agree and counsel. Contact the prescriber with a sourced recommendation, then explain any change to the patient.

Limits for interaction checks

AskMed has limits that matter for interaction checks. It is not a dedicated interaction database: it does not screen a full medicine list automatically or grade every pair, so use it alongside your pharmacy's interaction resource, not instead of it. AskMed never gives doses, dose adjustments in numbers or schedules, even when the label states them; take those from the current prescribing information or your national formulary. Answers are written by an AI language model from the sources AskMed finds, with no human review before you see them, so an answer can miss an interaction or misread a label; open the label before you act. AskMed searches a fixed list of sources, so local formularies and hospital policies will not appear. It is a reference for healthcare professionals and students, it does not decide treatment for a person, and it is not for emergencies such as a suspected overdose.

Related

• ToolQTc calculatorCorrected QT by Bazett, Fridericia, Framingham and Hodges, graded by sex.
• TooleGFR calculatorEstimated GFR from creatinine, age and sex with CKD-EPI 2021, plus the KDIGO GFR category.
• ScenarioA night shift with no senior to askCheck a cited answer at 3 a.m., and know when to escalate instead.
• ScenarioA patient brings in a studyFind the real paper, judge the evidence and explain it in plain words.
• ScenarioChecking a guideline between patientsConfirm a current recommendation and its year in the minute before the next patient.