I’m a doctor who writes the content your medical team keeps having to rewrite.

A nurse practitioner is finishing notes on the fourth patient of a busy afternoon when a banner appears across the chart: elevated sepsis risk, score 82, recommend lactate and blood cultures.

The patient is a 68-year-old woman admitted for a hip fracture. Her heart rate has been running around 104, up from 88 that morning. She is also in pain, has had nothing for it in three hours, and is anxious about surgery in the morning.

A woman arrives at an emergency department at eight in the evening with abdominal pain. She had a CT scan eleven days ago at a hospital across the city. She knows this. She says so twice.

The department cannot see the images or the report. So the registrar orders another CT. A second dose of radiation, a four-hour delay, a scanner slot another patient needed, and a bill somebody pays.

Start with the headline, because the headline is where the problem starts: “How AI Diagnoses Cancer With 99% Accuracy: The Technology That’s Replacing Radiologists”

Three claims, packed into fourteen words. None of the three survives contact with what the technology actually does.

14:32:06. A clinician approves an AI-generated recommendation. 14:32:19. She approves the next one.

Thirteen seconds apart. The audit log records two human reviews, and everything the compliance documentation promised is, technically, happening.

Here is what a quarterly blood pressure review actually measures. One reading. Taken on a machine the patient has never used before. After a journey. While they are mildly stressed about being late, in a room they do not want to be in.

That single number then stands in for ninety days of a person’s physiology.

Take control of your diabetes without spending hours at the hospital.

See a diabetes doctor from home. No waiting room, no taking a day off work, and no four-hour trip for a fifteen-minute review.

If you live with diabetes and check your blood sugar four times a day, you are working with four numbers. A continuous glucose monitor produces around 288 in the same 24 hours.

That gap is the whole story.

A1c is a good test and an average, and averages hide things. Two people sharing a result of 7.5% can be living completely different lives, one steady, the other swinging between 50 and 300 mg/dL, and only one of them is safe.

Most articles like this one are built on a quiet assumption: that people delay cardiac care because they do not know the warning signs. Tell them the signs, and they will act.

That is not what I saw in practice.

Telemedicine gets discussed as though it were one product. It is at least three, and the differences matter because they fail in different ways.