Page 216 - Glucose Monitoring Devices

CHAPTER

11
Retrofitting CGM traces

1
2
2
Simone Del Favero, PhD , Andrea Facchinetti, PhD , Giovanni Sparacino, PhD ,
Claudio Cobelli, PhD 2
1 2
Assistant Professor, Department of Information Engineering, Padova, Italy; Department of
Information Engineering, University of Padova, Padova, Italy
Introduction
Continuous glucose monitoring (CGM) technology has been constantly improving
since its first appearance two decades ago, and CGM is now spreading in clinical
practice [1]. CGM has been profitably used as an addition to self-monitoring blood
glucose (SMBG) measurements to improve glucose control by using the glucose
trends-in-time measured by the device to adjust insulin dosing in real time [2].
More recently, some CGM models have been approved by the US Food and Drug
Administration (FDA) as a substitute to SMBG devices (nonadjunctive CGM use)
insulin dosing [3]. Beside real-time use of CGM data for T1D treatment, the
recorded CGM traces can be downloaded and used retrospectively for many
purposes, for instance, to analyze glucose patterns and adjust standard therapy
parameters (e.g., basal pattern, carbohydrate-to-insulin ratio, etc.) [4], to assess
glucose control achieved in a clinical trial [5,6], to estimate physiological model
parameters [7], and to identify glucose-insulin models [8,9].
In the present chapter, we consider the retrospective use of CGM data and review
a “retrofitting” algorithm we originally proposed in Ref. [10], a technique designed
to improve a posteriori both precision and accuracy of a CGM trace by using a few
SMBG measurements collected in parallel to CGM. By merging information of
CGM (high temporal resolution) and SMBG (sparse in time but more accurate
than CGM), the retrofitting method produces a continuous-time BG profile which
is more accurate than the original CGM data. Having a more accurate CGM trace
is beneficial for the above-mentioned retrospective CGM applications.

Chapter organization
In section The retrofitting algorithm we review the retrofitting algorithm presented
in Ref. [10].
In section Retrofitting outpatient study data we show that the retrofitting
algorithm is very effective in enhancing precision and accuracy of Dexcom SEVEN
PLUS CGM sensor, in the setup we encountered in some of our outpatient clinical

Glucose Monitoring Devices. https://doi.org/10.1016/B978-0-12-816714-4.00011-9 219
Copyright © 2020 Elsevier Inc. All rights reserved.