A Methotrexate Mix-Up Can Turn Fatal

Some drug name mix-ups cause a delayed refill or a phone call. Methotrexate mix-ups are different. When the wrong product gets selected — or the right product gets dosed the wrong way — the outcome can be a life-threatening event, sometimes weeks after the prescription left the pharmacy.

Writing in Pharmacy Times, medication safety pharmacist Michael J. Gaunt documented fatal errors involving methotrexate and discussed why look-alike, sound-alike drug names remain one of the most persistent hazards in dispensing. The cases are sober, not sensational — and that is exactly what makes them instructive for pharmacy teams.

These are not stories about careless people. They are stories about a selection step that gives a tired human very little to distinguish between very different medications — and about verification steps that often check the selected product against the wrong reference.

Two reported errors, two deaths

Gaunt reported two cases. In the first, a prescription for Paxil (paroxetine) 10 mg daily was reportedly misheard as Trexall (methotrexate), and Trexall 10 mg daily was dispensed. An older patient took the medication for seven days, became extremely ill, and later died.

In the second, methotrexate was dispensed instead of metolazone for a patient in an assisted-living facility. The medication was administered daily, and the patient later died.

In both cases the intended therapy was common and unremarkable: an antidepressant, a diuretic. It was the substitution at the product selection step — and the daily dosing of a drug meant to be taken weekly — that made the difference.

How a few letters turn into a fatal error

The failure chain usually starts before the pharmacist ever makes a decision. A prescription is entered with a partial search — Gaunt noted that entering only “MET” can surface methimazole, metolazone, methotrexate, and other similar names together, sometimes with overlapping strengths and dosage forms. The products look adjacent on screen, so the distance between the right choice and the wrong one is a single click.

Once the wrong product is selected, the rest of the workflow is mostly confirmatory. Barcode scanning checks the scanned item against the label. The dispensing system checks the fill against itself. Downstream steps tend to validate the selected item — is the NDC right, is the count right, does it match what was entered — rather than independently comparing it back to the original prescription: was the product chosen actually what the prescriber intended?

That distinction is the heart of the problem. A chain that verifies internally can be perfectly executed at every step and still deliver the wrong medication, because the comparison that mattered happened at selection — and nothing downstream re-examined it against the prescription as written.

What ISMP recommends

Citing the Institute for Safe Medication Practices, Gaunt laid out safeguards specifically for methotrexate:

These are practical, well-supported steps, and every pharmacy dispensing methotrexate should have them in place. But notice what they have in common: most act at or near the selection step, inside the same workflow where the pressure and the look-alike lists exist. They reduce the probability of error. They do not add an independent check after the fact.

An independent layer that compares prescribed to dispensed

This is the class of mismatch SAV E-Rx, our system for reviewing e-prescription product selections, is designed to identify. SAV E-Rx independently compares what was prescribed to what was dispensed — drug, strength, dosage form — and sends meaningful discrepancies to a pharmacist for review. It does not trust the label to be right because the system accepted it; it goes back to the original prescription and asks whether the two actually match.

In the reported cases here, a Paxil prescription filled with Trexall, or a metolazone prescription filled with methotrexate, is precisely the kind of prescribed-versus-dispensed difference such a comparison surfaces. We cannot claim SAV E-Rx would have prevented these deaths — we were not there, and the outcomes depended on many factors beyond product selection. What it does add is an opportunity to detect the mismatch and intervene before harm, with the outcome depending on workflow timing and how quickly the alert is reviewed and acted on.

SAV E-Rx complements, and does not replace, ISMP safeguards, pharmacist verification, and patient counseling. The safest pharmacy is one where careful humans and an independent comparison layer are both looking at the same risk from different angles.

What pharmacies can do now

Drawing on the safeguards in the source article, here is a starting checklist for your team:

• Set oral methotrexate to default to weekly dosing in your system, with exceptions limited to specific cancer indications.
• Make sure inadvertent daily methotrexate dosing triggers an alert rather than a silent fill.
• Search prescriptions with more letters of the drug name plus the strength so look-alike entries are separated.
• Capture and display the indication on prescriptions where possible, and use it during product selection.
• Verify dose, frequency, and indication before dispensing, and counsel patients clearly on how and when to take methotrexate.
• Add a post-selection check that compares the prescribed medication to the dispensed product — automatically if possible, so it does not depend on catching the difference by eye.

None of these steps requires new infrastructure to begin. The last one is where most pharmacies have a genuine gap today, and it is the gap that makes selection errors invisible until the patient is already harmed.