Blood Journal
Leading the way in experimental and clinical research in hematology

Enough already of the word “robust”!

  1. David P. Steensma

Lately, I have noticed that the word “robust” has become one of the overused terms in biomedical science, and we hematologists are among the worst offenders. This abuse of “robust” is a relatively recent phenomenon; the number of biomedical articles published annually with the word “robust” in title or abstract has increased 40-fold since 1982 (Figure 1A). The occurrence of “robust” in Blood articles has gone from nil in the early 1990s to more than 100 references in 2003 (Figure 1B). While other terms have also experienced dramatic growth (eg, p53), such usage changes have occurred because of scientific discovery, not linguistic conformity and imitation.

Figure 1.

Growth of usage of the word “robust” in recent publications. (A) The number of biomedical publications each year containing the word “robust” in the title or abstract, indexed by the National Library of Medicine and accessed via PubMed at on October 20, 2003. Publications that included the word “robust” only in the author field (eg, publications by Drs Robusto, Robustellini, Robustelli, and Robustova) were excluded. Although PubMed searches also retrieve a small subset of paleontology and anthropology articles, reading the recent abstracts verified that the growth of articles containing the word “robust” has not come because research concerning Australopithecus robustus and related East African hominids is proceeding at an astonishing pace. The total annual number of PubMed-indexed publications increased less than 2-fold from 1982 to 2002. Figures for 2003 are projected based on occurrences of robust through October 1. (B) The number of articles published in Blood each year that have included the word “robust” in the abstract, title, or text of the article. The full text of all Blood articles published since January 1990 was searched at on October 20, 2003.

Dictionary definitions of “robust” (derived from a Latin word for “oak”) include hale and hearty synonyms like “vigorous” and “firm.” Describing an object or idea as “robust” implies that it has the power to withstand physical or intellectual challenge. But in modern medical parlance, this once useful code word is becoming meaningless jargon. “Robust” used to designate a treatment strategy or laboratory technique that actually worked most of the time, in contrast to those that failed frequently and for no apparent reason. When a PhD described an assay as “robust,” the audience could safely assume that the technique was so straightforward, even a clinically trained MD could perform it successfully. (Note: The author is a clinically trained MD.) But now robust is degenerating into merely a trendy way of saying “good.”

At a recent meeting I counted 8 consecutive speakers who used the word “robust” in their presentations, as if infectious robustitis were spreading from one to the next like a meme, a “virus of the mind.”1 The speaker who mercifully broke the “robust” string spent most of her talk struggling with the unfamiliar data projector (a robust and universal standard for these is desperately needed) and also had laryngitis, forcing minimalist language.

It is possible to make a robust point without using the word “robust.” Literary standards such as the complete works of Shakespeare (37 plays and 154 sonnets), the King James Bible, and Bulfinch's mythology2 do not use the word “robust” even once. Despite plenty of robust structures in the human body, there is only a single “robust” descriptor buried in the 1396 pages of Henry Gray's anatomical classic.3 Bartlett's Quotations does not contain one aphorism with the word “robust,” proving that witty and clever sayings can exist in a robust-less world.

In contrast to this parsimony, among the 5739 abstracts submitted for the 2002 American Society of Hematology (ASH) annual meeting, a whopping 53 contained the word “robust”; in 2001 there were 36. Interestingly, there appears to be an acceptance bias in favor of abstracts containing the word “robust”: in 2002, 83% (44 of 53) of ASH abstracts containing “robust” were chosen for presentation, whereas only 60% of all submitted abstracts escaped the stigma of “publication only.” In 2001, the same trend existed (78% “robust” accepted vs 66% overall). In contrast to words like “robust” and “molecular” (76% presentation rate in 2001 and 73% in 2002), the term “descriptive”4 is the kiss of death for an ASH abstract: a 42% accept rate in 2001-2002, and almost all of the accepted abstracts in this group used “descriptive” to refer to statistics, not science. The take-home message is crystal clear: all my future ASH abstracts will gratuitously use the words “robust” and “molecular” and will avoid “descriptive” like the plague.

This tiresome use of “robust” is not unique to hematology. The American Society of Clinical Oncology suffers from the same disease, although at an earlier stage: 47 “robust” meeting abstracts spread over the last 3 years. The American College of Cardiology suffered 10 “robust” abstracts this year, while “Digestive Disease Week 2003” featured 16 “robust” abstracts among the nearly 5000 presented. Surprisingly, orthopedic surgery, the specialty of choice for Olympic athletes and football linebackers seeking a career change and blessed with many physically robust individuals, remains unaffected: at their big annual meeting, only 1 orally presented abstract in the last 3 years has been “robust.”

If we are to rescue this word before it becomes as cliché as “proof of principle,”“elegant,” and “intriguing,” we must act soon. One way of highlighting and remedying the overuse of the word “robust” might be to declare a “Robust-Free Day” at the next ASH annual meeting. On this day, all speakers caught using the word “robust” would be required to buy a drink for the first 3 rows of the audience. The author welcomes other robust suggestions; you will find me at the front of the room in the plenary sessions at ASH, waiting to collect my free drinks.


D.P.S is a Mayo Foundation Research Scholar supported by the Mayo Foundation, Rochester, MN, USA. The author has no potential conflicts of interest to report.


  • Correspondence: David P. Steensma, MRC Molecular Haematology Unit, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, University of Oxford, Headington, Oxford OX3 9DU, United Kingdom; e-mail: david.steensma{at} or steensma.david{at}