You bought a Stelo (or a Lingo). Now what?

What the line on the screen actually is

The sensor is not measuring your blood. A small filament sits in the fluid between the cells just under your skin, and every few minutes it produces an estimate made from that fluid. Two things follow from that, and between them they explain most of what looks odd on the chart.

It runs late. The fluid under your skin catches up with your blood a few minutes behind, so a quick rise appears on the chart after it has already happened, and so does a quick fall.

It is an estimate. Every reading carries a margin of error, so two sensors worn at the same time would not necessarily draw the same line, and neither of them is a laboratory result.

So the useful thing to read is the shape, not the individual numbers. A chart like this shows three things well: how high the line went compared with the rest of your own day, how quickly it got there, and how long it stayed up before it came back down. Those are the parts steady enough to be worth looking at.

Why one reading rarely tells you much

Here is the finding that ought to be printed on the box. When researchers fed people the same meal on different days and watched with the same brand of sensor, the response varied by close to thirty percent from one day to the next. Same food, same person, same sensor, and still close to thirty percent.

What you slept like, the walk you did or did not take, the meal before this one, being unwell, being stressed, how old the sensor is: all of it moves the line. Which means a single reading after a single meal is a fact about that particular afternoon. It is not a verdict on the food.

Two habits follow from that, and they are most of the skill.

• Compare yourself with yourself. Your line today against your own line last week is a comparison that holds up. Your line against a number you read in a forum is not.
• Repeat before you conclude. If you want to know what a particular breakfast does to you, eat it on a few different mornings and look at all of them together.

One more caveat, because the people selling sensors are quiet about it: in people who do not have diabetes, these sensors tend to read somewhat higher than a blood test would. Researchers have warned publicly that this can push people into changing how they eat over a difference that was never really there. And do not hold a reading from one brand of sensor up against a reading from another: the same study that found the day-to-day variation also found that the agreement between brands degrades. Your line is comparable with your own line, and that is the comparison worth making.

The chart on its own is weather

A week of glucose with nothing beside it is a weather report. It tells you what happened and gives you no way to find out why, because the why was in your day and your day was not written down.

So log the ordinary things alongside it: what you ate and roughly when, a dose you took, a walk, a bad night, a cold, a hard week. It does not need to be precise to be useful. “Porridge, about eight” is enough to line a meal up against the hours after it, and it is the kind of note you will still be making next month.

That last part matters more than it sounds. The reliable way to stop logging is to try to log everything. People who set out to record every field tend to find their motivation drains away, and a chart with a little context on it is worth more than an empty one with excellent intentions behind it. Pick the few things you actually want to understand, and let the rest go.

Getting your sensor data into BeeGoodHealth

Start with a file. That path works today, and this is exactly what it does and does not cover.

Export the CSV from the app that came with your sensor, or from Dexcom's Clarity dashboard if that is where you look at it, then upload it on the Import page in BeeGoodHealth. The readings land on the same timeline as everything else you log. Importing your own data is free on every plan, and it is not something we would ever put behind a payment.

What happens to it once it is in

Everything you log lands on one timeline, so a meal, a dose and a walk are visible right next to the stretch of line they might explain. That is the whole point of the thing.

Beyond the timeline, on the free plan: your time in range is compared with your own earlier days rather than with somebody else's target, and the arithmetic behind every number is shown in the app, in a panel you can open. And whatever you have put in comes back out as a CSV whenever you want it.

On Premium: meal responses carry their sample size on the face of them, and a meal stays marked provisional until you have logged it with sensor coverage on three separate days, for exactly the reason described above. There is also a report to hand a clinician, with the standard one-page glucose summary they already know how to read, and your own logged context after it.

What this guide will not do

You will have noticed that nothing above told you what your glucose ought to be, or what to do about anything on your chart. That is deliberate, and it is not us being coy.

What a given pattern means for you depends on your history, your medication and a dozen things a website does not know, and the person equipped to tell them apart is a clinician who knows you. We build the picture. Reading it is yours, and your doctor's or dietitian's if you choose to show them. If something on your chart worries you, that is a conversation with them, not with a search engine and not with us.

What we will do is show you the arithmetic behind every number, tell you how much data each one is standing on, and decline to invent one when there is not enough.