Artificial intelligence in diabetes care requires better benchmarks
Clinician-curated standards can help assess glucose readings.
April 30, 2026
By
Matthew Solan
2 min read
Clinical Scorecard: Artificial intelligence in diabetes care requires better benchmarks
At a Glance
Category Detail
Condition Diabetes management using continuous glucose monitoring (CGM) Key Mechanisms Interpretation of CGM data influenced by patient-specific factors and clinician judgment Target Population Individuals with diabetes using CGM Care Setting Clinical practice
Key Highlights
Need for validated clinician-informed benchmarks for AI in diabetes care Variability in CGM data interpretation among clinicians Glycemia Risk Index (GRI) as a composite metric for glycemic quality AI tools require training on datasets reflecting expert clinical reasoning Development of benchmarks tailored to specific patient subpopulations Guideline-Based Recommendations
Diagnosis
Utilize standardized metrics from CGM for assessing glycemic control Management
Incorporate clinician assessments in the interpretation of CGM data Monitoring & Follow-up
Regularly evaluate CGM data with a focus on time in range, hypoglycemia, and hyperglycemia Risks
Recognize variability in CGM interpretation may lead to inconsistent clinical decisions Patient & Prescribing Data
Patients with diabetes utilizing continuous glucose monitoring
AI tools must be informed by clinician-derived benchmarks for safety and efficacy
Clinical Best Practices
Develop gold standard benchmarks from clinician interpretations of CGM data Use GRI for evaluating glycemic quality in diabetes management Tailor benchmarks to specific diabetes subpopulations for improved outcomes References
AACE Endocrine AI
is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE ® ). The ideas and opinions expressed in
AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies .