The issue has little effect on the glucose measurements doctors normally use. However, during very high or very low blood sugar episodes, CGMs may not be able to show exactly how high or low blood sugar levels become.
the study analysed nearly 47 million glucose readings from 948 people with type 1 diabetes, making it the largest investigation of this issue to date.
The research was inspired by a personal connection to diabetes. Dr John Mulley, of ¶º±ÆÖ±²¥'s School of Environmental and Natural Sciences, who’s expertise is Evolutionary biology said,
"I became interested in continuous glucose monitoring after a close family member was diagnosed with type 1 diabetes. Like many families, we quickly realised how transformative these devices can be. They provide an extraordinary amount of information and have changed the way people manage diabetes. At the same time, I started wondering what happens when glucose goes beyond the limits that the sensor is actually able to measure."
Most CGM systems can measure glucose only within a fixed range, typically 40–400 mg/dL (2.2–22.2 mmol/L). When glucose falls outside that range, the sensor reports only that its upper or lower limit has been reached rather than the true glucose concentration. Although these limits are well known, their real-world impact has never been systematically investigated.
The study found that between 94% and 100% of participants experienced at least one measurement-limit reading during the monitoring period. Many episodes of high glucose remained at the upper measurement limit for an hour or longer, meaning that although clinicians and patients could see that glucose was extremely high, the sensor could no longer distinguish between values of 400 mg/dL, 500 mg/dL or even higher.
Importantly, the research also found that these measurement limits do not substantially distort the standard summary measures used routinely in diabetes care, including mean glucose, time in range (TIR), glucose management indicator (GMI), glucose variability and standard deviation. Instead, the principal consequence is the loss of information about the severity of the most extreme glucose excursions.
"This isn't a criticism of continuous glucose monitoring," said Dr Mulley. "CGMs are one of the most important advances in diabetes care over the last two decades, and these results shouldn't discourage anyone from using them. Rather, the study highlights something that is easily overlooked: once the sensor reaches its measurement limit, it can tell you that glucose is extremely high or extremely low, but it cannot tell you exactly how high or how low."
The research also found that measurement-limit readings were particularly common in younger people and in individuals with higher HbA1c levels, suggesting that the issue is most relevant in groups who experience more frequent episodes of marked hyperglycaemia.
The findings suggest that future CGM software could be improved by routinely reporting how often sensor measurement limits are reached, and for how long. Such information would help clinicians and researchers recognise when glucose measurements have become censored by the device's measurement range, even though the standard summary metrics remain reliable.
Dr Mulley added: "Every scientific instrument has a measurement range, and continuous glucose monitors are no exception. Understanding those limits helps us interpret the data more appropriately and reminds us that, during the most extreme glucose events, there is still information that current sensors simply cannot capture."
Read the research here:
