Your health report was never written for you

The employee opens the report on their phone. It is six pages of test names, numbers, units and ranges. Most rows are black. One is red and marked “H”.

The report has done exactly what it was built to do. It has recorded the sample, the measured value and the laboratory’s reference interval. But the employee is not asking for any of those things. They are asking: Is this serious? Why might it have happened? What am I supposed to do now?

The PDF has no answer. So the employee searches the test name, finds a dozen possible conditions and tries to decide which one belongs to them. A routine health check has quietly handed them a clinical interpretation problem without the context needed to solve it.

The report contains the result. The uncertainty arrives with it.

The report is not failing

A conventional lab report is an important clinical document. It identifies the test, records the value, shows the unit used and compares the value with a reference interval. It gives a clinician a reliable starting point and preserves the result for future comparison.

The problem begins when that technical record is delivered directly to an employee and treated as though it were also an explanation. The conversation that makes the result useful is missing.

Laboratory results rarely stand alone. MedlinePlus, a service of the US National Library of Medicine, explains that a result outside a reference range may or may not signal a health problem, while a result inside the range does not guarantee that everything is fine. Different laboratories may also use different methods and ranges. Symptoms, medical history, medicines, examination and earlier results all affect what a value means for one person.

The red flag is forced to do too much work

For the employee, colour becomes meaning. Green feels safe. Red feels dangerous. “H” and “L” look like judgments about health rather than descriptions of where a number sits relative to a laboratory interval.

That visual shorthand can create two opposite mistakes. A person may be frightened by a small variation that needs no urgent action. Or they may feel reassured because every row appears inside range even though symptoms or history still deserve attention. The flag can draw attention. It cannot supply judgment.

This is why “abnormal” is often a poor word for the person reading the report. It sounds like a conclusion. In many cases, what the report actually knows is narrower: this value sits outside this laboratory’s reference interval. What that means comes next.

Explanation should answer the questions the report cannot

A useful explanation starts by saying what was found in everyday language. It then says how much attention the finding deserves in this person’s context—not with false certainty, but with a clear sense of proportion. Finally, it makes the next step visible: no action, watch and repeat, discuss with a doctor, or seek care sooner.

That order matters. A definition of the test is not enough. “This measures how your liver is working” may be accurate, but it still leaves the person wondering whether their result matters. Plain language is useful only when it helps someone make a better decision.

The World Health Organization describes health literacy as the ability to access, understand, assess and use health information. It also makes an important shift in responsibility: health services and information providers should make trustworthy information understandable and actionable. Confusion is not simply a weakness in the reader. It can be a failure in how the system communicates.

If the employee has to search the internet to discover whether a red flag matters, the explanation was never delivered.

AI can translate the report. It should not become the doctor

AI is well suited to some of this work. It can organise a long report, replace unfamiliar terms with everyday language, connect the current value with earlier results and keep the agreed next step visible. Used carefully, it can make the first reading of a report calmer and more useful.

But translation and judgment are different jobs. A model should not turn one marker into a diagnosis, hide uncertainty or decide urgency from a PDF alone. When the meaning depends on symptoms, history or clinical trade-offs, a physician has to remain responsible. The safest system makes that boundary obvious instead of letting fluent language disguise it.

A health check should include the handover

Corporate health programmes usually specify what tests will be included, where samples will be collected and how quickly reports will be delivered. Far less attention is given to the moment the employee opens the result. Who explains it? How quickly? What happens when the answer is not “everything is fine”?

That handover should be part of the service, not an optional layer the employee has to find. The report is the evidence. The explanation turns it into something the person can use. The next step turns that understanding into care.

What we are building

Oxology is being built so the report does not arrive alone. Results are organised around the questions an employee actually has. Important findings are explained in plain language, relevant history and trends remain available, and a physician steps in wherever clinical judgment is required. The next action stays attached to the result instead of disappearing into another app or conversation.

We are early, and good explanations will require continuous clinical review. The principle is already clear: a health report should not ask an employee to become their own interpreter.