Key result
Mobile app glucose monitoring predicts nocturnal hypoglycemia with an AUC of ~0.92.
Why the study?
The study was conducted to develop a prognostic algorithm of nocturnal hypoglycemia based on data from a glucose monitoring mobile application.
Can a prognostic algorithm based on glucose monitoring mobile application data accurately predict nocturnal hypoglycemia in patients with type 1 diabetes mellitus?
Observational (n=239)
Can a prognostic algorithm based on glucose monitoring mobile application data accurately predict nocturnal hypoglycemia in patients with type 1 diabetes mellitus?
A prognostic model using mobile application glucose monitoring data accurately predicts nocturnal hypoglycemia in patients with type 1 diabetes mellitus.
Answers are grounded in this paper and cite their sources.
May aid nocturnal hypoglycemia risk stratification via mobile apps in type 1 diabetes; hypothesis-generating and requires prospective validation.
Objective: to develop a prognostic algorithm of nocturnal hypoglycemia (NH) based on the glucose monitoring mobile application data. Material and methods. The retrospective analysis of 524 continuous glucose monitoring (CGM) profiles of patients with type 1 diabetes mellitus was performed. CGM was performed using the Medtronic iPro2 system for 6-7 days, overnight periods of CGM were analyzed to identify systematic NH. There were 239 patients included in the study, among them 65 (27.1%) were identified as having systematic NH. Models of 7-point glycemic profiles were built and their data were uploaded to the DiaLog GM mobile application to calculate standardized glucose monitoring parameters. The prognostic model of NH was developed based on the logistic regression method. Results. According to the results of regression analysis the most significant predictors of NH included in the prognostic model were: glycated hemoglobin (p=0.001), using of insulin pump therapy (p=0.001), time below target range level 1 (p<0.001), coefficient of variation for glucose (p=0.02). The area under the ROC curve for the prediction model was 0.917; the optimal cut-off point for the predicted probability of NH was 0.317, with model sensitivity of 85% and specificity of 90%. Conclusions. The developed prediction model based on data from a specialized mobile application allows to improve the existing approaches to NH risk assessment due to higher predictive ability.
No takes yet. Share an insight, caveat, or question.
Rusanov et al. (2024) conducted an observational in Type 1 diabetes mellitus (n=239). Prognostic model based on mobile application data was evaluated on Prediction of systematic nocturnal hypoglycemia (AUC 0.917). A logistic regression model based on mobile application glucose monitoring data predicted nocturnal hypoglycemia with an area under the ROC curve of 0.917, 85% sensitivity, and 90% specificity.
Synapse has enriched 10 related papers. Consider them for comparative context: