Editorial: Choose your words carefully: a couple things technical writers can teach scientific communicators

by Geoff Hart

Previously published as: Hart, G. 2007. Editorial: Choose your words carefully: a couple things technical writers can teach scientific communicators. the Exchange 14(3):2, 10–11.

What's wrong with the phrase "take one tablet by mouth three times daily for seven days"? If you're reading this newsletter, you might not think there's anything much wrong with it—but then again, you're a highly literate technical communicator who works primarily with words, and that means you have uncommonly good reading skills. Yet an article in the July 2007 issue of Consumer Reports, citing a paper in the Annals of Internal Medicine, reported that 46% of English-speaking adults with a range of language skills could not correctly follow such instructions.

What's wrong with the drug name Lunesta? Apart from the obvious (i.e., that the name provides not the slightest indication of what the drug does), it apparently sounds too much like Neulasta—which wouldn't be a problem if both were sleeping pills, but given that Neulasta is a chemotherapy drug, there's a high downside potential should a slip of the tongue occur. The same issue of Consumer Reports reported a recent near-disaster in which a pharmacist confused the two drugs, though fortunately another pharmacist caught the problem before the drugs reached the patient. Such errors are apparently common, since syllable transpositions (new-leh versus loo-neh, in this case) are a pernicious problem that leads to the well-known and usually entertaining phenomenon of spoonerisms. I once amused a group of strangers by reporting that "I'm a grad stool", having fallen uncomfortably in the gap between "I'm a grad student" and "I'm in graduate school". The National Academy of Sciences has reported that up to 25% of medication errors may result from such problems.

Treat my description of both problems as illustrative, not authoritative, since I'm working exclusively from secondhand information instead of going direct to the source to verify the details. (Anyone willing to hunt down the original materials and provide a more detailed description of the problem? Please do! It would make for a great article in a future issue.) My point in raising these two examples is to illustrate two problems that all technical writers and editors are familiar with: First, that experts are overfamiliar with their own words, and don't examine their word choice as carefully as they should. Second, that critical text should include built-in ways to prevent errors or let readers detect errors quickly.

Can the lessons learned from technical communication help us avoid such problems in our own work? Emphatically yes. Here are some possibilities.

First, let's consider the prescription instructions. Those of us who write software documentation have learned to break instructions into discrete statements, with one statement per step, instead of stringing them together in a single long sentence. Each step can then be read, understood, and acted upon before moving to the next step, thereby eliminating the risk of overloading the reader's "channel capacity" (the amount of information they can receive simultaneously). It may also be helpful to remove useless redundancies such as "take by mouth" that don't add to the precision, leaving less text for readers to deal with; after all, there are few other ways to take a tablet, and in the absence of clear and diplomatic instructions to the contrary, few patients will assume the tablets should be used as suppositories. Using a strong single verb such as "swallow" instead of a weaker phrasal verb ("take... by mouth") would also help. Last but not least, we should focus carefully on the context of use: here, the important thing is not that the medication should be taken for a week, but rather that all pills must be consumed according to a specific schedule. On that basis, "take one tablet by mouth three times daily for seven days" might become the following, at least as a first draft we can subsequently test and refine with the help of representative patients:

Take one tablet with breakfast, one with lunch, and one with dinner.
Swallow each tablet with a full glass of water or milk.
Continue taking your medicine until all the pills are gone.

Now let's consider the confusion of drug names. We can't reasonably expect to exert any influence over the choice of drug names, which are chosen by marketing departments for reasons that surpasseth mortal understanding. But we can build some forms of error prevention and detection into the process. In contrast to the previous example, encouraging doctors to build redundancy into the information they convey would be beneficial; for example, adding "sleeping pills" or "to help patient to fall asleep" to the scrip would help to alert even a sleepy pharmacist that Neulasta is the wrong drug. In the previous example, making the instructions concrete (by listing the actual times when each pill should be taken) instead of vague (three times per day, but at what times?) provided guidance that minimizes errors. Providing redundant information such as "sleeping pills" helps in a similar way. For additional security, we could ask doctors to specify acceptable generic alternatives to brand-name drugs. If the functions of the generic drug and the name-brand version don't match, this should raise a red flag that alerts the pharmacist to a problem. (It's also an ethical approach that offers patients the chance to save money.)

Needless to say (therefore, doubly worth saying), you should not simply accept my suggestions uncritically, particularly if the consequences for human life are significant. I'm confident that my suggestions are reasonable, but my experience working with actual readers over the years has convinced me how unreliable my own impressions are. For any change to current practice for which the consequences of failure are severe, we must test our new approach carefully to ensure that reasonable also means realistic and that we haven't created any new problems. In this case, the test audience would be a sample of typical patients and pharmacists, and the testing might be simple indeed. For the rewritten prescription instructions, we could ask patients to read the instructions and tell us when they would take each dose, how many pills they would take at that time, and when they could stop taking the pills. If the answers don't match our expectations, that's a clear sign we need to probe further to find out why. For the pharmacists, we could submit a series of prescriptions for drugs whose names are likely to be confused (based on the National Academy of Sciences study results), with the prescriptions divided into two groups: first, the traditional prescription with no extra details, and second, the same prescription with clues added. We could then compare the frequencies of error for the two groups.

These two medical examples illustrate something we communicators tend to forget: that communication is not a passive activity, in which readers or listeners open their minds and we pour in pre-packaged information in perfect, industrially precise quantities. On the contrary, our audience is always an active participant in any communication, and creates their own meaning from what we give them. Because that meaning may not be what we intended, we must always ask ourselves whether it might be possible, with a little more work, to error-proof that communication. Of course, that's a purely theoretical question. In a more practical sense, my examples also remind us that as consumers of medical services, we must take responsibility for confirming that we understand a prescription, and if not, that we ask the doctor to rewrite it so that we do understand.

I suspect that many doctors will react unfavorably to efforts on our part to educate them about better ways to communicate. But it's ethical for us to try. Sometimes an appeal to selfishness works: "The costs of malpractice insurance would drop if all doctors wrote prescriptions more clearly." Sometimes an appeal to authority works: "I'm a professional communicator, and have won awards in medical writing. So when I suggest that you write things this way, you can take that as professional advice every bit as good as the advice you provide." And sometimes it's just a question of educating yourself and your friends and family to become better-informed consumers of information, medical or otherwise.

My essays on scientific communication have now been collected in the following book:

Hart, G. 2011. Exchanges: 10 years of essays on scientific communication. Diaskeuasis Publishing, Pointe-Claire, Que. Printed version, 242 p.; eBook in PDF format, 327 p.