Real-World Ketone Testing Patterns Among People Using FreeStyle Libre

OBJECTIVE: Ketone testing behavior in relation to glucose control is not well characterized. We aimed to describe ketone testing patterns and assess glucose control among individuals using a FreeStyle Libre (FSL) reader and blood ketone test strips in a real-world dataset of over 2.5 million ketone readings. RESEARCH DESIGN AND METHODS: Historical glucose and ketone data from September 2014 to September 2024 were analyzed. As data are recorded for individual readers, the analysis is performed by readers. Ketone excursions were categorized by initial ketone level: ≤0.5, 0.6-1.4, 1.5-2.9, and ≥3.0 mmol/L. Glucose concentrations were evaluated 24 h before and after the initial ketone test. Ketone test frequency and concentrations within 72 h post-test were also assessed. RESULTS: A total of 92.4% of readers never recorded a ketone test, and only 5% of readers performed a ketone test when glucose exceeded 250 mg/dL. Users with ketone levels ≥3.0 mmol/L had higher glucose concentrations in the 24 h preceding the ketone test. Among those with ketones ≥1.5 mmol/L, ketone testing was typically delayed by 5-8 h after glucose surpassed 250 mg/dL, and retesting ketones within 24-72 h was rare. CONCLUSIONS: Most readers did not demonstrate adherence to standard-of-care recommendations for ketone monitoring and infrequently recheck ketones after elevated readings. Incorporation of continuous ketone monitoring into FSL would address limitations with current ketone testing methodologies and adherence to ketone monitoring guidelines.