Medical Content
What Health-Tech Companies Get Wrong About Medical Content
By Miracle Olajuyigbe 4 min read
A marketing lead once sent me a draft with a note attached. “A doctor already checked this one, so it should be quick.”
The doctor had checked it. Every fact in it was correct. The piece was about starting insulin, and it was also, from the point of view of a frightened newly diagnosed patient, close to unreadable. Nine paragraphs before anyone mentioned that injecting insulin does not really hurt. That is the single thing most people want to know first, and it was sitting on page two underneath a section about pharmacokinetics.
Nothing in that draft was wrong. It just did not work.
I see some version of this constantly, and after a few years of it, the mistakes have started to look less like individual errors and more like a pattern. Here are the ones that come up most.
Hiring for one half of a two-part problem
Medical content needs two competencies that rarely live in the same person.
The first is clinical judgment: knowing what a claim actually commits you to, spotting the sentence that quietly overstates what a device does, understanding why “detects cancer” and “flags a suspicious region” are separated by a regulatory canyon.
The second is the writing itself. Structure, rhythm, knowing what a reader needs in the first twenty seconds, being able to explain something hard without either dumbing it down or hiding behind vocabulary.
Companies usually solve for one and hope the other sorts itself out. Hire a brilliant content writer and you get pages that read beautifully and occasionally claim things you cannot support. Hire a clinician who writes on the side and you get accuracy wrapped in prose that nobody finishes. Then you bolt the other one on at the end as a reviewer, which is the most expensive point to discover a structural problem.
Treating clinical review as a final checkpoint
Review that happens after the draft exists can only catch errors. It cannot fix the fact that the article was framed wrongly from the first line.
By the time a piece reaches a reviewer, the angle is set, the structure is built, the headline has been signed off, and the reviewer’s realistic options are to approve it or to detonate a week of work. Most reviewers, sensibly, pick their battles. They fix the outright inaccuracies and leave the deeper problem, which is that the piece answers a question no patient was asking. Clinical thinking belongs at the brief stage, when someone can still say “this framing is going to scare people unnecessarily” or “the actual search intent here is not what we think it is.” That conversation costs an hour. The same realisation after publication costs a rewrite and sometimes a retraction.
Confusing credentials with clarity
The “medically reviewed by Dr So-and-so” badge has become standard, and it does real work for trust. It has also quietly become a substitute for the harder question.
A physician signing off means the content is not wrong. It does not mean a worried thirty-four-year-old will understand it, act on it correctly, or come away less anxious than when they arrived. Those are different tests, and only one of them is being run.
If you want to know whether your content works, the badge will not tell you. Give the page to someone outside healthcare, ask them to read it once, then ask them what they would do next. The answer is usually illuminating and occasionally alarming.
Forgetting who else is reading
A lot of health-tech content is written squarely at the buyer. Hospital administrators, procurement teams, investors. Fair enough, that is who signs the contract.
But your marketing pages are public, and patients find them. So do the clinicians you are selling into, which is its own hazard. I have read vendor blog posts announcing the replacement of radiologists, published by companies whose customers are radiology departments. Somebody wrote that, somebody approved it, and nobody in the chain asked what a radiologist would feel reading it over coffee.
You do not need separate content for every audience. You do need to know, for each page, who might land on it, and to write in a way that does not embarrass you in front of any of them.
Mistaking hedging for safety
This is the failure mode I find most frustrating, because it comes from good intentions.
Everyone is nervous about liability, so qualifiers get layered on. “May,” “in some cases,” “individual results vary,” “consult your healthcare provider.” Each one is defensible. Twenty of them in one article produce a page that says nothing at all, and a reader who leaves no better informed than they arrived.
Vagueness is not actually safe. A patient who cannot extract a clear action from your page will go and find a clearer answer somewhere with lower standards than yours. You have not protected them. You have outsourced them to a forum.
The alternative is not recklessness. It is specificity about the thing you can be specific about, and honesty about the rest. “Most people notice a difference within two weeks. If you do not, that is worth mentioning to your doctor rather than waiting for your next appointment” is both safer and more useful than “results may vary.”
Publishing and walking away
Medical claims age. Guidelines change, a drug gets a new indication, a trial you cited gets superseded, a device gets a clearance that widens or narrows what you can say about it.
Most companies have no mechanism for this. Content goes up, gets forgotten, and then somebody encounters a three-year-old page during due diligence or a regulatory conversation and everyone spends a bad fortnight. The fix is unglamorous: a claims inventory, a review cycle, and a rule that anything citing a specific study or statistic gets a date stamp and a next-review date.
What good actually looks like
Bring clinical thinking in at the brief, not the review. Write the first two hundred words for the most frightened person likely to read them. Be specific where you can be and honest where you cannot. Test comprehension on someone outside the field. Keep a list of what you have claimed and when you last checked it.
None of that is complicated. It is just a different order of operations from the one most teams default to, where the medicine gets checked last and the writing gets checked never.