Abstract
BACKGROUND: Adherence to annual OGTT screening for cystic fibrosis-related diabetes (CFRD) is poor. In this mixed methods study we assessed the accuracy, feasibility, and acceptability of alternative approaches to glucose measurements during an OGTT.
METHODS: Standard OGTT with plasma glucose sampling at 0, 60, and 120-min was conducted in 14 youth ≥ 10 years of age with CF. A self-administered OGTT kit (Digostics, GTT@home) measuring capillary glucose and Dexcom G7 continuous glucose monitoring (CGM) were performed concomitantly with OGTT. CGM glucose values were recorded at 5-minute intervals for 20-minutes after each time point to account for lag. Plasma OGTT glucose and glucose tolerance categories [normal (NGT), impaired (IGT), indeterminate (INDET), and CFRD] were compared with these same outcomes as defined by the self-administered kit and CGM. Younden's index was used to determine the optimal CGM timepoint for categorization of glucose tolerance, and ROC curves were used to identify the optimal glucose thresholds. Participants and their parents were interviewed to understand their experience with alternative testing approaches.
RESULTS: Based on plasma glucose, participants were characterized as NGT (n = 8), IGT (n = 5), and CFRD (n = 1). Whereas the GTT@home correctly categorized glucose tolerance in 6 participants (43 %), CGM correctly categorized 13 (93 %). The CGM glucose at 125 min was identified as the timepoint at which the measured glucose yielded maximum discrimination for all categories of glucose tolerance (AUC for NGT = 0.979, IGT = 0.867, CFRD = 1.0). A CGM glucose threshold of 154 mg/dL demonstrated 100 % sensitivity and 87.5 % specificity for identifying NGT; for IGT a threshold of 182 mg/dL exhibited 80 % sensitivity and 88.9 % specificity. Parent-child dyads valued convenience during OGTT screening, but expressed concerns about glucose measurement accuracy and performing medical procedures in the home.
CONCLUSIONS: CGM glucose values during OGTT may offer an accurate assessment of glucose tolerance categories, though this approach may require further refinement for patient acceptability.