What does this medicine mean for this patient?

Last time, we argued that professionals shouldn’t have to carry every piece of clinical knowledge in their heads, and that something should bring the right parts forward instead. But which parts are the right ones? That depends entirely on who is in front of you.

A patient is more than a record

In most healthcare software, a patient is a record: a name, a date of birth, a list of past visits. A medicine is an entry too — something that can be prescribed, dispensed, stocked or billed.

But a patient is a person with a history and a current situation that can change the meaning of every clinical decision. And a medicine carries knowledge: what it’s used for, how it should be taken, what to check first, what can make it unsafe, and how it behaves alongside other medicines.

The two only become useful together. The question is rarely what is this medicine? It’s what does this medicine mean for this patient?

Same medicine, four patients

Take ibuprofen — about as ordinary as medicines get. Here’s what it means for four people who could easily walk in on the same morning:

• A healthy 30-year-old with a sprained ankle. A sensible short course, taken with food.
• Akosua, 54, on warfarin. Taken together, these raise the risk of serious bleeding. Paracetamol is the safer painkiller — and even that needs care, because regular use can push her INR up.
• Ama, 29, twelve weeks pregnant. Best avoided in pregnancy, especially later on. Paracetamol is usually preferred, at the lowest dose that works.
• Mr Mensah, 71, with weak kidneys. Anti-inflammatories can push struggling kidneys further. Another option is needed.

Same drug, same strength, same box. Four completely different answers. None of them is unusual, and every one depends on information that sits outside the drug itself: the other medicines, the pregnancy, the kidney result.

That’s what we mean by context. It isn’t extra information. It’s the information that changes the answer.

Not everything — what matters now

There’s an easy mistake on the other side of this. If a system knows everything about a drug, it’s tempting to show everything: every interaction, every caution, every rare side effect. Anyone who has clicked past a wall of alerts knows how that ends. When everything is flagged, nothing stands out — and the warning that matters gets dismissed with the rest.

So the goal isn’t more information. It’s the right information:

• connected to the person, not shown in general
• ranked, so the serious thing comes first
• in plain language, with where it came from
• at the moment of the decision, not in another window afterwards

A good system should be able to say: for Akosua, this is the one thing that matters about ibuprofen, here’s why, and here’s what to consider instead. Then step back.

Which brings us to the point

If the knowledge lives in the system, and the system can connect it to the patient, one question is left: who decides? That’s where this series ends.

Next: The system holds the knowledge. You make the call. →