No sensor: does time in range still work for me?

The definition of time in range is clear. Less clear is the data it assumes: a record without gaps. I have no sensor, three or four fingersticks a day.

The 2019 international consensus defines it as the percentage of readings and of time in 70–180 mg/dL. Pregnancy has its own targets; otherwise the target for adults with type 1 or type 2 diabetes is above 70%, above 50% for older or high-risk individuals, about 60% under 25 when the A1C goal is 7.5%. These targets come from continuous sensor data; a person’s own target is set by their doctor.

It sets a data condition too: 14 days, the sensor active at least 70% of the time. Hand-entered readings cannot meet it: nothing exists between two fingersticks.

A sensor measures duration; a fingerstick measures a moment. Sensor readings come at equal intervals, so the percentage of readings matches the percentage of time; with three or four readings a day it does not.

In DCCT data from people with type 1 diabetes, Beck and colleagues computed a ratio from seven-point fingerstick profiles, calling it time in range. A lower ratio went with a higher risk of retinopathy progression; they argued it should be an acceptable clinical trial end point. Those samples came one day every three months, measured at a central laboratory; mine come from a home meter, no such profile.

ForMyGluco calculates a different figure: my hand-entered readings over 7, 14 or 30 days, against a target range I picked from a list, not necessarily the consensus 70–180. The card carries the metric’s name; its ⓘ text says: the percentage of readings inside the target range.

Two questions for an appointment: is this ratio meaningful at my testing frequency, and what should my target range be?