Improved individual and population-level HbA1c estimation using CGM data and patient characteristics

Joshua Grossman; Andrew Ward; Jamie L Crandell; Priya Prahalad; David M Maahs; David Scheinker

doi:10.1016/j.jdiacomp.2021.107950

Improved individual and population-level HbA1c estimation using CGM data and patient characteristics

J Diabetes Complications. 2021 Aug;35(8):107950. doi: 10.1016/j.jdiacomp.2021.107950. Epub 2021 May 17.

Authors

Joshua Grossman¹, Andrew Ward¹, Jamie L Crandell², Priya Prahalad³, David M Maahs⁴, David Scheinker⁵

Affiliations

¹ Department of Management Science and Engineering, Stanford School of Engineering, Stanford, CA, USA.
² School of Nursing, Department of Biostatistics, UNC Chapel Hill, Chapel Hill, NC, USA.
³ Division of Pediatric Endocrinology, Stanford School of Medicine, Stanford, CA, USA; Lucile Packard Children's Hospital, Stanford, CA, USA.
⁴ Division of Pediatric Endocrinology, Stanford School of Medicine, Stanford, CA, USA; Lucile Packard Children's Hospital, Stanford, CA, USA; Stanford Diabetes Research Center, Stanford School of Medicine, Stanford, CA, USA; Department of Health Research and Policy, Stanford School of Medicine, Stanford, CA, USA.
⁵ Department of Management Science and Engineering, Stanford School of Engineering, Stanford, CA, USA; Division of Pediatric Endocrinology, Stanford School of Medicine, Stanford, CA, USA; Lucile Packard Children's Hospital, Stanford, CA, USA; Stanford Diabetes Research Center, Stanford School of Medicine, Stanford, CA, USA; Clinical Excellence Research Center, Stanford School of Medicine, Stanford, CA, USA. Electronic address: dscheink@stanford.edu.

Abstract

Machine learning and linear regression models using CGM and participant data reduced HbA1c estimation error by up to 26% compared to the GMI formula, and exhibit superior performance in estimating the median of HbA1c at the cohort level, potentially of value for remote clinical trials interrupted by COVID-19.

Keywords: Continuous glucose monitoring; HbA1c estimation.

Publication types

Research Support, N.I.H., Extramural

MeSH terms

Adolescent
Blood Glucose / analysis
Blood Glucose / metabolism
Blood Glucose Self-Monitoring / instrumentation
Blood Glucose Self-Monitoring / methods
Blood Specimen Collection / methods
Blood Specimen Collection / standards
Blood Specimen Collection / statistics & numerical data*
COVID-19 / epidemiology*
Child
Cohort Studies
Diabetes Mellitus, Type 1 / blood*
Diabetes Mellitus, Type 1 / epidemiology*
Female
Glycated Hemoglobin / analysis*
Glycated Hemoglobin / metabolism
Health Services Accessibility / statistics & numerical data
Humans
Machine Learning / standards
Male
Pandemics
Remote Sensing Technology / methods
Remote Sensing Technology / standards
SARS-CoV-2 / physiology
Statistics as Topic / instrumentation
Statistics as Topic / methods
Young Adult

Substances

Blood Glucose
Glycated Hemoglobin A

Grants and funding

R18 DK122422/DK/NIDDK NIH HHS/United States