FDA Drug Names Drifted Toward Stranger Spellings Over Four Decades
By the 2020s, approved drug names had moved far from the familiar shape of ordinary English. Jesduvroq, Omvoh, Zurzuvae, and Xolremdi do not read like the older brand names that once anchored the market, such as Cardene or Claritin. In this preprint analysis of FDA approvals and related naming datasets, proprietary names became steadily more orthographically distinctive over time, with letter sequences growing less predictable and less similar to earlier names already on the market.
The finding is not just cosmetic. Drug naming sits at the intersection of branding, regulation, and patient safety, and a name that is too close to an existing product can trigger prescribing or dispensing confusion. The same forces that push a name away from ordinary English may also push it into unfamiliar terrain for clinicians, pharmacists, and patients. A drug name can be made more unique, yet still become harder to read, pronounce, and remember.
Orthographic Distinctiveness Rose in Every Decade Measured
The analysis examined 1,321 FDA new molecular entity names from 1985 through 2024, 2,769 FDA Orange Book trade names from 1982 through 2026, 177,337 US pharmaceutical trademarks from 1980 through 2026, and 1,273 generic names from 1982 through 2025. Each name was scored for character trigram surprisal, a measure of how unexpected its three-letter sequences were relative to earlier names and, in a parallel baseline, ordinary English. The study also measured nearest-neighbor similarity using edit distance, which captures how closely a name resembles the most similar existing names.
For new molecular entity names, mean trigram surprisal increased in every decade from the 1980s to the 2020s, rising from 7.21 to 8.50 nats. That represented an 18% increase, with a positive time trend of b = 0.038, p < .001, and R² = 0.34. In practical terms, newer names contained more unusual letter combinations and less familiar orthographic patterns than earlier ones. Half of names approved in the 2020s contained a letter sequence found in no common English word, compared with 7% in the 1980s.
The progression was visible in the examples themselves. Older names tended to resemble recognizable English forms. Newer names more often presented clusters such as “mvo” or “voh,” sequences that rarely appear in everyday vocabulary and can make a name feel more invented at first glance.
A sharp rise in distinctiveness does not guarantee safer communication.
Similarity to Existing Names Fell as the Namespace Crowded
The edit-distance analysis told the same story from a different angle. Nearest-neighbor similarity declined after the early 1990s, suggesting that new names were increasingly selected to avoid sounding or looking too close to existing products. The trademark data mirrored this trend, with a similar rise in orthographic distinctiveness over time, b = 0.016, p < .001. Even the enormous trademark set, which included many names that were never marketed, tracked the same directional shift seen in FDA-approved products.
That pattern is consistent with namespace crowding. Every new brand name has to stand apart from a growing inventory of existing trademarks, approved products, and established generic names. The more crowded the naming field becomes, the more pressure falls on manufacturers to invent unusual spellings, novel consonant clusters, and atypical letter combinations. Regulatory scrutiny adds another layer, because proposed names must clear safety review for confusing similarity before approval.
The study also connects this trend to the mechanics of medication error prevention. Look-alike and sound-alike name pairs remain a recognized source of confusion in pharmacy practice and prescribing systems. Prior observation studies have found wrong-drug dispensing and interception rates near 1 in 1,000 in real-world settings, far below the much larger counts seen in error-reporting systems. Experimental work has also shown that visually degraded or highly similar names are more likely to be misidentified. In that context, distinctiveness is not an abstract branding feature; it is a design variable with direct consequences for human error.
Generic Names Shifted Too, but More Modestly
The secondary analysis of generic names found a weaker but still positive trend. Character trigram surprisal rose over time for International Nonproprietary Names and United States Adopted Names, with b = 0.011 and R² = 0.063. The pattern was less pronounced than in proprietary names, which makes sense given the structural constraints of the WHO stem system. Generic names are not free to range as widely as brand names, because related drugs must share common suffixes such as -tinib or -sartan.
Even so, the direction of travel was similar. As the pharmaceutical lexicon expanded, generic names also drifted toward more distinctive character patterns, though not to the same degree as trade names. The comparison is useful because it separates pure branding behavior from broader pressures on the naming system as a whole. Brand names appear to be under the strongest push toward novelty, but the whole namespace has become more crowded.
Why Strange Spellings Can Cut Both Ways
Orthographic distinctiveness may reduce one kind of error while creating another. A name that stands apart from competitors is less likely to be confused with a near neighbor, but it may also be less familiar to readers and harder to retrieve from memory. Uncommon letter sequences can slow recognition, and difficult spelling can interfere with pronunciation, recall, and transmission across clinical workflows. The result is a trade-off: fewer close look-alikes on one side, more cognitive friction on the other.
The study cannot isolate which force is most responsible for the trend. Regulatory pressure, commercial strategy, and market crowding almost certainly all contribute, but the data do not allow those causes to be disentangled. What the analysis does establish is narrower and more concrete: over four decades, FDA-related drug names became statistically more distinctive in spelling, and they did so across multiple independent datasets.
The preprint also leaves a bedside question unresolved. More unusual names may help prevent some confusion errors, but orthographic novelty is not the same as safety. A medication name that is visually exotic can still mislead, stall recognition, or slip through memory in busy clinical settings. The study ends there, with the naming system now more crowded, more inventive, and more difficult to navigate than it was four decades ago.