And Then You're Dead

The Autopsy / Case File · Medical Error & Preventable Care

Close Enough to Kill You

hydrOXYzine treats anxiety. hydrALAzine treats a hypertensive crisis. A tired nurse, a rushed pharmacist, or a hurried physician reading one for the other has been a documented, recurring way patients die.

The Institute for Safe Medication Practices has maintained a running list of drug names that get confused for each other since 2008 — a problem old enough, and common enough, that it has its own acronym in patient-safety literature: LASA, for look-alike, sound-alike. Estimates built from national medication-error reporting programs put LASA name confusion behind roughly a quarter of all reported medication errors, and the industry's primary defense is still a decades-old typographic trick.

And Then You're Dead · September 2026 · 8 min read

Join the discussion →
A prescription medication bottle and its capsules spilled next to it on a countertop
A prescription bottle and its contents. Reading a drug name correctly, quickly, under time pressure, is a task the entire dispensing chain — physician, pharmacist, nurse, patient — performs dozens of times a day, and errors compound at every handoff.

A patient prescribed hydrOXYzine, an antihistamine used for anxiety and itching, instead receives hydrALAzine, a blood-pressure medication that can send a normotensive patient into dangerously low blood pressure. A diabetic asks for their Humulin and is handed Humalog — same manufacturer, similar name, entirely different onset and duration of insulin action, with a dosing error that can trigger severe hypoglycemia. Neither of these mix-ups requires a careless clinician or a broken system in any dramatic sense. Both are documented, recurring patterns in medication-error reporting, and both trace back to the same root problem: two drug names similar enough that a rushed reader can mistake one for the other.

A Quarter of the Errors, By Name Alone

Medication errors have many causes — wrong dose, wrong route, wrong patient, wrong time — but name confusion occupies a specific and well-quantified slice of the total. National medication-error reporting programs, which aggregate incident reports from hospitals, pharmacies, and other care settings, have consistently found that confusion between drug names that look alike on a label or sound alike when spoken accounts for roughly a quarter of all reported medication errors, with individual analyses in the peer-reviewed literature putting the LASA-specific share of all medication errors somewhere between 7% and 25% depending on the reporting system and clinical setting studied.

0%
approximate share of medication errors reported to national error-reporting programs attributed to look-alike, sound-alike (LASA) drug-name confusion
0
the year ISMP launched its formal, continuously updated List of Confused Drug Names, consolidating years of individually reported name-pair confusion cases

The mechanism behind that share is almost entirely about human reading and hearing under time pressure, not about any single point of failure. A drug name gets misread on a handwritten or poorly formatted electronic order, misheard on a verbal order given over the phone during an emergency, or mis-selected from an alphabetized drop-down list in an electronic prescribing system where two similar names sit one entry apart. Each of those failure points sits at a different stage of the medication-use process — prescribing, transcribing, dispensing, administering — which is part of why no single fix has ever fully closed the gap: a solution built for one stage (say, computerized order entry) doesn't necessarily prevent the error at another (a pharmacist grabbing the adjacent bin on a crowded shelf).

The Pairs That Keep Recurring

The ISMP's List of Confused Drug Names isn't a hypothetical catalog — every entry on it reflects a real, reported pattern of mix-ups serious enough that the organization flagged the pair for special handling. hydrOXYzine and hydrALAzine is one of the most persistently cited examples in patient-safety literature specifically because the two drugs are used in such different clinical contexts (an antihistamine versus an antihypertensive) that a mix-up rarely gets caught by context alone — a clinician expecting to see "the itching medication" on a chart has no built-in reason to double-check that the name in front of them isn't a blood-pressure drug. Humalog and Humulin R present a related problem inside a single manufacturer's product line: both are insulins, both share the "Hum-" prefix, and the difference between them is a rapid-acting analog insulin and a regular human insulin with a meaningfully different action profile — a distinction that matters enormously for dosing timing and hypoglycemia risk, but one easy to blur under a five-second glance at two visually similar vial labels.

hydrOXYzine treats anxiety. hydrALAzine treats a hypertensive emergency. The two names differ by exactly one internal syllable — and the drugs share almost nothing else in common.

Tall Man Lettering: The Fix That's Nearly 30 Years Old

The primary tool the pharmaceutical and patient-safety world has deployed against LASA errors is tall man lettering — capitalizing the distinguishing letters within a drug name, as in hydrOXYzine and hydrALAzine, so the visual difference between two similar names is exaggerated exactly where the confusion tends to happen. Both the FDA and ISMP maintain and periodically update a list of drug-name pairs with officially recommended tall-man treatments, and most major electronic health record and pharmacy dispensing systems now render at least some of these names in mixed case by default.

The approach works, within its limits, precisely because it targets the visual read rather than trying to change clinician behavior or workflow. But it doesn't address the sound-alike half of the problem — tall man lettering does nothing for a verbal order given over the phone, where hydrOXYzine and hydrALAzine sound close enough that a distracted or rushed listener can still transpose them. ISMP's broader recommended safeguards reflect that gap: using both a drug's brand and generic name together on any order or label, stating the medication's purpose alongside its name (so "hydrALAzine for blood pressure" carries a built-in context check that a bare drug name doesn't), configuring prescribing software to require a minimum number of letters before a name auto-populates from a search field, and physically separating look-alike products on pharmacy shelves so a rushed pick from the wrong bin becomes less likely by design.

A Name Change That Wasn't Enough

The clearest illustration of how hard this problem is to engineer away comes from a drug whose manufacturer saw the collision coming and tried to dodge it before the product ever reached a pharmacy shelf. Searle originally planned to market its new arthritis medication as "Celebra." The FDA, reviewing the proposed name in advance, flagged its resemblance to Cerebyx, an anti-seizure drug, and Celexa, an antidepressant already on the market — three drugs treating three unrelated conditions, sharing enough letters that a rushed read could confuse any pair of them. The manufacturer renamed the product Celebrex before launch specifically to reduce that risk.

It wasn't enough. Within roughly the first two months after Celebrex reached the market, the FDA had logged 53 dispensing-error reports involving confusion among the three drugs, and ISMP's own error-reporting system had independently collected around 40 documented cases. In one recorded incident, an 80-year-old patient was given citalopram (Celexa), an antidepressant, in place of the arthritis medication she'd actually been prescribed, and continued taking the wrong drug for a month before the error was caught. None of the reported cases in that early wave resulted in a death, but the episode became a standard teaching example in patient-safety literature for a specific reason: a manufacturer had already done the one thing regulators most often recommend — proactively changing a drug's name pre-launch to reduce name-collision risk — and the errors happened anyway, at meaningful volume, within weeks. The remedy that followed wasn't a further rename; the FDA pushed the manufacturers to fund journal advertising specifically educating physicians and pharmacists on telling the three names apart, an acknowledgment that once a name is close enough to another to cause confusion, changing it again only shifts the collision risk to whatever the new name turns out to resemble.

Why the List Keeps Growing

ISMP's confused-drug-name list isn't shrinking as old pairs get resolved — it's expanding, with dozens of newly reported pairs added in recent revision cycles alone. Part of that growth simply reflects an ever-larger pharmaceutical name-space: as more drugs enter the market, particularly within crowded therapeutic categories that draw on similar Latin and Greek pharmacological roots for naming, the raw number of names close enough in spelling or sound to collide keeps climbing, independent of whether hospitals get any better at preventing the errors that result. A name-space problem doesn't resolve itself through vigilance alone — it requires deliberate naming conventions at the point drugs are approved, a lever regulators control more directly than any individual hospital's safety program does.

And Then You're Dead

Nobody's death certificate says "name confusion." It says the hypoglycemic event, the hypotensive crisis, whatever the wrong drug actually did once it was inside the patient — the five seconds where one syllable substituted for another leaves no trace in the vital statistics, only in the incident report that may or may not get filed afterward. The fix has existed since before most of the clinicians currently practicing were licensed. It just has to be applied correctly, every single time, by everyone in a chain that includes the one person having a bad day.

That's it. That's the whole thing.

Comments

Loading…

Comments are anonymous and moderated.

Sources

Photo: "Adderall bottle and capsules," by Ryanaxp, via Wikimedia Commons (public domain).