Most CGM consultations begin with the summary figure at the top of the ambulatory glucose profile: a single time-in-range percentage for the last 14 days. It is the least informative number on the page, because it averages away every pattern that could be acted on.
A better sequence takes about a minute. Open the weekly view first and find the worst day. Then open the daily overlay for that day and find the worst three-hour block. Then ask the patient what happens in that block. Almost always the answer is a specific, fixable thing: the meal that is never bolused for, the evening that runs late, the Sunday that starts at eleven. A dose change made against the fortnightly average will move the whole curve, including the parts that were already fine, and usually buys a little more time in range at the cost of a little more hypoglycaemia.
The same discipline applies to the low-glucose figure. A time-below-range of 3% spread evenly across the fortnight is a different clinical problem from the same 3% concentrated in two nights, and only the second is likely to be what wakes the patient up at three in the morning.
- Start at the weekly view, not the summary percentage; find the worst day before looking at anything else
- Drill into the worst three-hour window on that day and ask what happens then, rather than proposing a dose change
- Check whether time below range is spread out or clustered; clustered lows are the ones that cause harm and fear
- Insist on at least two full weekends of data before accepting a weekly pattern as real
- Write the identified window into the notes, so the next consultation can check whether that specific block improved
The statistics, in plain English
Time in range is a mean of proportions, so it hides distribution entirely: 70% achieved evenly across 14 days and 70% achieved through ten excellent days and four poor ones are the same number and completely different clinical situations. The same is true of time below range, where the total percentage says nothing about clustering, and clustering is what determines whether a patient has a severe event or loses their hypoglycaemia awareness. Fourteen days is the usual minimum because it captures two weekends; a seven-day report contains only one, and a single unusual weekend can then shift the whole summary by several percentage points.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free