# Sensor coverage

How much of each day the sensor actually reported; a day with half its readings missing is not a day’s average.

Type: cgm_metric · Unit: % of expected readings · Domain: Metabolic Health · Accuracy: Calculated · Importance: Supporting

Also called: CGM coverage, data coverage, missing readings, sensor wear time, data sufficiency

## In one sentence

How much of the period your sensor actually reported, which is the number that says whether every other number on this shelf means anything.

## Why it matters

A sensor that reads every five minutes should give 288 readings a day. Readings go missing when the phone is out of range for longer than the sensor's memory, during the warm-up hour, when VT removes an impossible value, and in the gap between one sensor and the next. A day's mean built from a morning and an evening is not a day's mean; a week's time in tight range with the weekend missing is a weekday figure. The international consensus asks for at least 70 percent of possible readings over at least 14 days before a period's figures are taken as representative (Battelino 2019; Danne 2017), and that is the floor VT uses.

## What moves it

You, mostly: how far the phone is from your arm and for how long; whether the GS3 app is running; a sensor that failed early; a swim long enough for the signal to drop. The sensor itself stores 14 days of readings and fills the gap when the phone returns, so most gaps close on their own.

## How to keep it high

Keep the phone within about ten meters most of the time, keep the sensor app installed and allowed to run in the background, and change the sensor as soon as the old one expires rather than the next morning; VT reminds you twelve hours before. Check the coverage figure on the day you look at anything else.

## How Vitality Tracker uses it

Shown with every summary, as a percentage for the period and a bar for the day; a day or period below the 70 percent floor is shown as partial, and the glucose management indicator is not computed until 14 days with 70 percent coverage exist.

## How far to trust the number

Calculated exactly: received readings over expected readings.

## Method

Coverage = readings received ÷ readings expected × 100, capped at 100, where expected is the period's length divided by your sensor's usual interval between readings (five minutes for most).

## Sources

1. Battelino T, et al. Diabetes Care 2019;42:1593–603. [doi:10.2337/dci19-0028](https://doi.org/10.2337/dci19-0028)
2. Danne T, et al. Diabetes Care 2017;40:1631–40. [doi:10.2337/dc17-1600](https://doi.org/10.2337/dc17-1600)
3. Sibionics GS3 product page, read 19 October 2026.

## Related

- [What a CGM measures](https://vitalitytracker.app/knowledge/what-a-cgm-measures)
- [Why a month](https://vitalitytracker.app/knowledge/why-a-month)
- [Mean glucose](https://vitalitytracker.app/knowledge/mean-glucose)
- [Glucose management indicator](https://vitalitytracker.app/knowledge/glucose-management-indicator)
- [Reading a CGM without fooling yourself](https://vitalitytracker.app/knowledge/reading-a-cgm-without-fooling-yourself)

Reviewed October 2026 · v1

Source: https://vitalitytracker.app/knowledge/sensor-coverage