Guide

Time in range by hour of day

One number for the whole day hides which hours go wrong. Split it by the clock, and the hours that cost the most stand out.

The short answer

For each hour of the day, take every CGM reading from that hour over the last few weeks and count the share between 70 and 180 mg/dL (3.9–10.0 mmol/L), the consensus target range. That gives 24 numbers instead of one. The lowest hours are where most of the time out of range happens. Each 1% of a day is about 14 minutes, so an hour that sits at 50% while the rest of the day sits at 80% is easy to see.

Do weekdays and weekends apart. School, work and sleep move by hours on a Saturday, and an average of both puts breakfast spikes at two different times into one blur.

Compare with your own best days

A per-hour number says where time goes, not how much of it you could get back. A target of 100% in every hour is not useful. A fairer bar is your own good days. This is how the Where to focus card in Glide's Insights does it:

  1. Each day gets a score: the average of its hourly time in range. A day needs at least 12 hours with 6 or more readings, so a day with a long sensor gap is left out.
  2. The best 1 in 4 days (the top quarter of scores) are kept apart. Weekdays and weekends each need 8 days before the card shows them.
  3. Two lines per group: the typical hour (all days) and the same hour on the best 1 in 4.
  4. The 3-hour stretch with the biggest gap between the two lines is the window to look at.
  5. The gain for the whole week weighs weekdays 5 and weekends 2.
Glide's Where to focus card: both together 78% to 80% time in range if the hours below went like your best 1 in 4 days. Weekdays 7 am to 10 am, 58% in range typically, 70% on the best 1 in 4. Weekends 8 pm to 11 pm, 63% typically, 74% on the best 1 in 4. Each with a typical line and a dashed best line by hour.
Made-up data. Weekday mornings and weekend evenings are the two windows here.

Some of every good day is luck, so the gain is the most you could expect, not a promise. It also does not say what to change. It says where to look.

Why some hours stand out

With pump data: what costs the time

If a Tandem pump or a pump that uploads to Glooko is connected (beta), Insights also sorts every reading out of range by what came before it. Each reading goes to the first rule it meets:

RowRule
Rebound after a lowAbove 180 within 3 hours after a low ends (15 minutes or more below 70)
High after a bolusAbove 180 within 4 hours after a bolus given by hand
High with no bolus beforeAbove 180 with no bolus in the 4 hours before
Low after a bolusBelow 70 within 4 hours after a bolus given by hand
Other lowsEvery other reading below 70
Glide's What costs time in range card: in range 78%, high after a bolus 15.0%, high with no bolus before 2.3%, rebound after a low 2.1%, low after a bolus 1.1%, other lows 0.8%, pump data not in yet 0.2%.
Made-up data. Automatic pump insulin does not count as a bolus here.

A match is an association, not proof of the cause. Talk through patterns with your care team before you change a setting.