Statistical Packages and Algorithms for the Analysis of Continuous Glucose Monitoring Data: A Systematic Review

Abstract

Background:

Continuous glucose monitoring (CGM) measures glucose levels every 1 to 15 minutes and is widely used in clinical and research contexts. Statistical packages and algorithms reduce the time-consuming and error-prone process of manually calculating CGM metrics and contribute to standardizing CGM metrics defined by international consensus. The aim of this systematic review is to summarize existing data on (1) statistical packages for retrospective CGM data analysis and (2) statistical algorithms for retrospective CGM analysis not available in these statistical packages.

Methods:

A systematic literature search in PubMed and EMBASE was conducted on September 19, 2023. We also searched Google Scholar and Google Search until October 12, 2023 as sources of gray literature and performed reference checks of the included literature. Articles in English and Danish were included. This systematic review is registered with PROSPERO (CRD42022378163).

Results:

A total of 8731 references were screened and 46 references were included. We identified 23 statistical packages for the analysis of CGM data. The statistical packages could calculate many metrics of the 2022 CGM consensus and non-consensus CGM metrics, and 22/23 (96%) statistical packages were freely available. Also, 23 statistical algorithms were identified. The statistical algorithms could be divided into three groups based on content: (1) CGM data reduction (eg, clustering of CGM data), (2) composite CGM outcomes, and (3) other CGM metrics.

Conclusion:

This systematic review provides detailed tabular and textual up-to-date descriptions of the contents of statistical packages and statistical algorithms for retrospective analysis of CGM data.

Keywords

continuous glucose monitoring statistical algorithms statistical packages statistics systematic review

Introduction

Continuous glucose monitoring (CGM) is an established technology in research and clinical practice.^1,2 Continuous glucose monitoring provides information on glucose levels collected automatically every 1 to 15 minutes.

Consensus on reporting CGM outcomes is established.³ However, handling large amounts of CGM data is challenging. The manual entering of summary statistics from the ambulatory glucose profile (AGP) reports from CGM manufacturers into databases before analysis is time-consuming and prone to error, although AGP report summary statistics are often used in clinical studies.^4,5 Continuous glucose monitoring manufacturers use different algorithms to compute CGM summary statistics and graphical displays of the glycemic status in AGP reports.⁶ Consequently, comparison of AGP reports between CGM manufacturers can be challenging. Also, specific research questions or clinical situations might require different summary statistics not offered by AGP reports from CGM manufacturers, for example, different glycemic variability metrics⁷ or different hypoglycemic and hyperglycemic cut-offs.⁸

A recent review of 12 statistical packages has been published.⁹ However, no systematic review of statistical packages and statistical algorithms for retrospective analysis of CGM data exists. The objective of the present review is to summarize existing data on (1) statistical packages to retrospectively analyze CGM data and (2) statistical algorithms for retrospective CGM data analysis not currently available in these statistical packages. The present review has the potential to contribute to the standardization of CGM reporting³ or, where appropriate, to contribute to more individualized statistical approaches in CGM analysis useful to researchers and clinicians in need of tools capable of analyzing CGM data retrospectively.

Methods and Materials

A systematic literature search was performed on September 19, 2023 (no lower publication date limit) in PubMed and Embase (Figure 1). We also searched Google Scholar and Google Search until October 12, 2023 as important sources of gray literature and performed reference checks of the included literature. An information specialist assisted in defining the search strings, see Supplementary Material. Important search words include: (“statistics” OR “analysis” OR “algorithm” OR “tools” OR “software”) AND (“continuous glucose monitoring” OR “CGM”). The titles and abstracts were screened for eligibility by reviewer 1 (MTO) and reviewer 2 (CKK) in Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia). Disagreements were resolved by discussion or by consulting reviewer 3 (PLK). This systematic review is registered with PROSPERO (CRD42022378163) and follows the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guideline.¹⁰

Figure 1.

PRISMA diagram.

Eligibility Criteria

Inclusion criteria: Statistical packages and/or statistical algorithms for retrospective analysis of CGM data.

Exclusion criteria: Statistical packages and/or statistical algorithms for glucose level predictions or statistical algorithms relying on an artificial pancreas setup. Statistical packages and/or statistical algorithms for denoising (ie, “smoothing”) of CGM traces, sensor drift, interstitial-to-plasma glucose time delay, and calibration of CGM signals.

Data Extraction and Presentation

Results on statistical packages (Tables 1 and 2) and statistical algorithms (Table 3) were tabulated. The corresponding authors were contacted for missing information.

Table 1.

Information on Statistical Packages for Retrospective Analysis of Continuous Glucose Monitoring Data.

	A	B	C	D	E	F	G
	Package	Summary of content	Graphical user interface	Data structure (file types, input unit)	Possibility of exporting results	Open source	Public availability
1	Rodbard²⁸	Calculation of SD (several subtypes, eg, variability within days (SDw), between days (SDb), between daily means (SDdm), within time series (SDws) for various durations, and/or with selected starting points (eg, nocturnal, daytime), CV, CONGA, MODD, mean glucose, minimum, maximum, and MAGE	No	Manual data entering, mmol/L or mg/dL	Yes	Yes Available as Appendix 2²⁸	Available at: https://doi.org/10.1089/dia.2009.0015 Available as Appendix 2²⁸
2	Baghurst²⁹	Calculation of glycemic variability metrics: MAGE, J-index, MODD, CONGA, and GRADE	NA	NA	NA	No	By request to corresponding author²⁹
3	Fritzsche et al³⁰	Computation of MAGE, mean glucose, SD, minimum, and maximum	Yes	.txt, mmol/L or mg/dL	Yes	No	Available at: https://sddgf.de/daten/mage.zip (Microsoft Windows only)
4	Moscardó et al³³ Package name: EasyGV	Commonly used glycemic metrics. Missing data imputation option Previous version available³¹	Yes	.xlsm, mmol/L	Yes	No	Available at: www.phc.ox.ac.uk/research/technology-outputs/easygv
5	Rawlings et al³² Package name: CGM-GUIDE	Commonly used glycemic metrics	Yes	.xls, mg/dL	No	No	No
6	Marics et al²² Package name: GVAP	Commonly used glycemic metrics. Missing data imputation option	Yes	.xls, mg/dL	No	Yes Available at: https://sourceforge.net/projects/glyvariab/	Available as webpage at: https://sourceforge.net/projects/glyvariab/
7	Neinstein et al²⁰ Package name: Tidepool Platform	Commonly used glycemic metrics. Video tutorial on how to use at: https://www.tidepool.org/	Yes	Device-specific, mmol/L or mg/dL	Yes	Yes Available at: https://github.com/tidepool-org/blip	Available as webpage at: https://github.com/tidepool-org/blip
8	Hall et al³⁴	Clustering patients in three motifs according to glycemic variability	Yes	.csv or .txt, mg/dL	Yes	Yes Available at: at: https://github.com/abreschi/shinySpecClust	Available as webpage at: https://adaychen.shinyapps.io/shinyspecclust/
9	Zhang et al⁶⁵ Package name: CGManalyzer	Commonly used glycemic metrics. Multiple group comparison and missing data imputation options	No	Device-specific, mmol/L or mg/dL	Yes	Yes Available at: https://cran.r-project.org/web/packages/CGManalyzer/index.html	Available as R package at: https://cran.r-project.org/web/packages/CGManalyzer/index.html
10	Yu et al²¹ Package name: MAGECAA	Computation of MAGE (with subtypes) and SD	Yes	.csv mmol/L	Yes	Yes	By request to corresponding author.
11	Vigers et al¹⁷ Package name: cgmanalysis	Commonly used glycemic metrics. Missing data imputation option	No	.csv, mg/dL	Yes	Yes Available at: https://cran.r-project.org/web/packages/cgmanalysis/index.html	Available as R package at: https://cran.r-project.org/web/packages/cgmanalysis/index.html
12	Millard et al⁶⁰ Package name: GLU	Commonly used glycemic metrics. Missing data imputation option	No	.csv, mmol/L	Yes	Yes Available at: at: https://github.com/MRCIEU/GLU	Available as R package at: https://github.com/MRCIEU/GLU
13	Bent et al⁹⁸ Package name: cgmquantify	Commonly used glycemic metrics. Missing data imputation option	No	.csv, mg/dL	Yes	Yes Available at: at: https://github.com/brinnaebent/cgmquantify	Available as R package at: https://github.com/brinnaebent/cgmquantify
14	Broll et al¹⁵ Package name: iglu	Commonly used glycemic metrics	Yes	.csv, mg/dL	Yes	Yes Available at: https://cran.r-project.org/web/packages/iglu/index.html	Available as R package at: https://cran.r-project.org/web/packages/iglu/index.html Available as webpage at: https://irinagain.shinyapps.io/shiny_iglu/
15	Kitae et al⁹⁹ Package name: KAMOGAWA	Commonly used glycemic metrics	Yes	.csv, mg/dL	Yes	No	Package available at Microsoft Access. Information on how to access⁹⁹
16	Hangelbroek¹⁰⁰ Package name: CGM Shiny	Commonly used glycemic metrics	No	.csv or .txt, mmol/L or mg/dL	Yes	Yes Available at: at: https://github.com/rolandhangelbroek/cgmshiny/tree/main	Available as R package at: https://github.com/rolandhangelbroek/cgmshiny/tree/main
17	Matabuena et al⁶⁶ Package name: Glucodensity	Density curves on distribution of glucose levels. Multiple group comparison option	Yes	.csv, mg/dL	Yes	Yes Available at: https://cran.r-project.org/web/packages/biosensors.usc/index.html	Available as R package at: https://cran.r-project.org/web/packages/biosensors.usc/index.html Available as webpage at: https://prediccion.shinyapps.io/functional_diabetes/
18	Attaye et al⁶⁷ Package name: CGDA	Commonly used glycemic metrics. Multiple group comparison and missing data imputation options	No	.csv or .txt, mmol/L or mg/dL	Yes	Yes Available at: https://github.com/EvdVossen/CGDA	Available as R package at: https://github.com/EvdVossen/CGDA
19	De Livera et al⁶⁸ Package name: CGMStatsAnalyser	Commonly used glycemic metrics and multiple group comparisons option	Yes	.csv, mmol/L	Yes	No	Available as webpage at: https://baker-biostats.shinyapps.io/CGMStatsAnalyser/
20	Shao et al¹⁰¹ Package name: CGMTSA	Commonly used glycemic metrics, including outlier detection and missing data imputation option	No	.csv or .txt, mmol/L or mg/dL	Yes	Yes Available at: https://github.com/RyanJ-Shao/CGMTSA	Available as R package at: https://github.com/RyanJ-Shao/CGMTSA
21	Olawsky et al⁶¹ Package name: rGV	Commonly used glycemic metrics	Yes	.csv, mmol/L or mg/dL	Yes	Yes Available at: https://cran.r-project.org/web/packages/rGV/index.html	Available as R package at: https://cran.r-project.org/web/packages/rGV/index.html Available as webpage at: https://shiny.biostat.umn.edu/GV/
22	Cappon et al¹⁶ Package name: AGATA	Commonly used glycemic metrics. Multiple group comparison and missing data imputation options	Yes	.mat or .csv or .xls or .xlsx, mmol/L or mg/dL	Yes	Yes Available at: https://github.com/gcappon/agata	Available at: https://github.com/gcappon/agata Available as stand-alone solution at: https://github.com/gcappon/agata/tree/master/standalone
23	Chrzanowski et al¹⁹ Package name: GlyCulator 3.0	Commonly used glycemic metrics. Missing data imputation option Previous versions in the works of Pagacz et al¹⁰² and Czerwoniuk et al¹⁰³ Video tutorial on how to use at: https://glyculator.btm.umed.pl/user_manual	Yes	.csv or .txt or .xls or .xlsx; mmol/L or mg/dL	Yes	No Available for v. 2.0 at on: https://github.com/kpagacz/glyculator	Available as webpage at: https://glyculator.btm.umed.pl

Table 1 provides an overview of the statistical packages with (A) authors of the statistical packages and publication year, (B) a summary of the statistical package content and previous package versions, (C) graphical user interphase (yes/no), (D) the data structure required when uploading CGM data for package analysis, (E) possibility of exporting/downloading results from the package analysis (yes/no), (F) open-source code (yes/no), and (G) publicly available (yes/no).

Abbreviations: AGATA, automated glucose dATa analysis; AUC, area under curve; CGDA, continuous glucose data analysis; CGM-GUIDE, continuous glucose monitoring graphical user interface with diabetes evaluation; CGMTSA, continuous glucose monitoring time series data analysis; CONGA, continuous overall net glycemic action; CV, coefficient of variation; EF, excursion frequency; GMI, glucose management indicator; GRADE, glycemic risk assessment diabetes equation; GVAP, glycemic variability analyzer program; iglu, interpreting blood GLUcose; KAMOGAWA, Kyoto Auto MAGE Of Glucose cAlculator With isCGM Application; MAD, median absolute deviation; MAGE, mean amplitude of glycemic excursions; MODD, mean of daily differences; NA, not available; SD, standard deviation; sGVP: standardized glycemic variability percentage; TAR, time above range; TBR, time below range; TIR, time in range.

Table 2.

Consensus Metrics,³ Additional Metrics, Imputation Options, and Multiple Group Comparison Options for Statistical Packages for Retrospective Analysis of Continuous Glucose Monitoring Data.

Statistical packagesAvailable metrics and options		A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W
Statistical packagesAvailable metrics and options		Rodbard²⁸	Baghurst²⁹	Fritzsche et al³⁰	Easy GV³³ (previous version³¹)	CGM-GUIDE³²	GVAP²²	Tidepool Platform²⁰	Hall et al³⁴	CGManalyzer⁶⁵	MAGECAA	cgmanalysis¹⁷	GLU⁶⁰	Cgmquantify⁹⁸	iglu¹⁵	KAMOGAWA⁹⁹	CGM Shiny	Glucodensity⁶⁶	CGDA⁶⁷	CGMStatsAnalyser⁶⁸	CGMTSA¹⁰¹	rGV⁶¹	AGATA¹⁶	GlyCulator¹⁹ (previous versions^102,103)
	Statistical package version (NA if no version number available)	NA	NA	NA	10	NA	1.1	2.53.0	NA	1.3	1.0	2.7.6	0.2.0	0.1.0	3.4.2	NA	0.0.0.9	NA	0.8.2	0.1.0	0.1.0	0.0.4	2.1.1	3.0
CONSENSUS GLYCEMIC METRICS
1	Sensor data obtained (percentage)³							X				X			X		X		X				X	X
2	Sensor data obtained (days)³											X			X							X	X	X
3	Scanning frequency (for flash glucose monitoring)³
4	TIR (time in range), 3.9-10·0 mmol/L or 70-180 mg/dL³				X	X	X^a	X				X	X		X		X		X	X	X	X	X	X
5	TITR (time in tight range), 3.9-7.8 mmol/L or 70-140 mg/dL³				X	X	X^a	X				X	X		X							X	X	X
6	TBR (time below range), < 3.9 mmol/L or < 70 mg/dL³				X	X	X^a	X				X	X		X	X	X^a		X	X^a		X	X	X
7	TBR (time below range), 3.0-3.8 mmol/L or 54-69 mg/dL (level 1)³				X	X	X^a	X				X	X		X		X		X	X		X	X	X
8	TBR (time below range) < 3.0 mmol/L or < 54 mg/dL (level 2)³				X	X	X^a	X				X	X		X		X		X	X		X	X	X
9	TAR (time above range) > 10.0 mmol/L or > 180 mg/dL³				X	X	X^a	X				X	X		X	X	X^a		X	X^a		X	X	X
10	TAR (time above range) 10.1-13.9 mmol/L or 181-250 mg/dL (level 1)³				X	X	X^a	X				X	X		X		X		X	X		X	X	X
11	TAR (time above range) > 13.9 mmol/L or >2 50 mg/dL (level 2)³				X	X	X^a	X				X	X		X		X		X	X		X	X	X
12	Mean sensor glucose³	X			X	X	X	X		X		X		X	X	X	X		X	X	X	X	X	X
13	GMI (glucose management indicator)³							X				X		X	X	X			X		X	X	X	X
14	GRI (glycemia risk index)^3,18														X								X
15	CV (coefficient of variation)³	X			X			X				X		X	X	X			X	X	X	X	X	X
16	Hyperglycemic events and extended hyperglycemic events^3,24											X			X								X
17	Hypoglycemic events and extended hypoglycemic events^3,24											X			X							X	X
18	Total consensus glycemic metrics computable	2	0	0	10	9	9	12	0	1		15	8	3	16	5	9	0	11	9	4	13	16	13
NON-CONSENSUS GLYCEMIC METRICS
		A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W
19	ADRR (average daily risk range)¹⁰⁴				X										X							X	X
20	AUC (area under curve) total(Area under the CGM curve)					X^b	X^b					X	X		X		X		X^b	X^b		X	X^b	X
21	AUC following an event, for example, meals, and time between event start and the glucose peak (event start and stop required)⁶⁰												X
22	Autocorrelation and partial autocorrelation coefficients¹⁰¹																				X
23	BGRI (blood glucose risk index)¹¹																						X
24	COGI (composite glucose index)^105,106														X								X
25	CONGA (continuous overall net glycemic action)¹⁰⁷	X	X	X	X	X	X			X		X			X				X	X		X	X
26	CVGA (control variability glucose analysis)¹⁰⁸																						X
27	Date of replacement of CGM																		X
28	Days with at least 80% data availability																		X
29	Distance travelled^c,92 (Sum of absolute differences in CGM glucose levels)																					X
30	DR (dynamic risk)¹⁰⁹																						X
31	eA1c (estimated A1c)¹¹⁰											X		X	X	X			X	X
32	Excursion number^d (Number of glucose excursions)						X					X							X
33	Fasting proxy^e,60 (Requires knowledge on night-time glucose levels)												X
34	Glucotype classification³⁴								X
35	GRADE (glycemic risk assessment diabetes equation)⁸⁹		X		X^f										X^f							X^f	X^f	X
36	GVP (glycemic variability percentage)¹¹¹				X										X							X
37	HBGI (high blood glucose index)⁶⁴				X							X		X	X					X	X	X	X	X
38	Hypoglycemic and hyperglycemic indices⁴⁰														X								X
39	Hyperglycemic events per day⁶¹														X
40	Hypoglycemic events per day⁶¹														X							X
41	Hyperglycemic duration per day⁶¹														X
42	Hypoglycemic duration per day⁶¹														X
43	IGC (index of glycemic control)⁴⁰				X										X								X
44	IQR (interquartile range)									X^g		X^g		X^g	X				X^g				X
45	J-index³⁷		X	X	X							X		X	X					X		X	X	X
46	LI (liability index)¹¹²				X																	X
47	LBGI (low blood glucose index)^63,64				X							X		X	X					X	X	X	X	X
48	MAG (mean absolute glucose change per unit time)¹¹³				X										X							X
49	MAD (median absolute deviation)⁶⁰									X			X		X		X
50	MAGE (mean amplitude of glycemic excursions)³⁵		X	X	X	X	X				X	X			X	X	X		X	X	X	X	X	X
51	MAGE upward and downward excursions^h						X				X				X								X
52	Maximum	X		X				X		X		X		X	X	X			X
53	First quartile									X		X		X	X		X		X
54	Median									X		X		X	X		X		X				X	X
55	Second quartile									X		X		X	X		X		X
56	Minimum	X		X				X		X		X		X	X	X			X
57	Meal time data^i,114														X
58	Mean sensor glucose outside range^j,98													X
59	Mean sensor glucose inside range^j,98													X
60	MODD (mean of daily differences)³⁶	X	X	X	X	X	X			X		X			X		X		X		X	X	X
61	Multiscale entropy⁶⁵									X
62	M value¹¹⁵				X										X							X	X	X
63	Number of sensor glucose observations	X																	X
64	Number of missing values(Requires information on sensor measuring frequency)				X					X
65	PGS (personal glycemic state)¹¹⁶				X
66	Rate of change incl. SD⁶¹														X								X
67	sGVP (standardized glycemic variability percentage)^60,117												X
68	SD (standard deviation) of glucose levels²⁴	X^k			X	X	X	X		X	X	X		X	X	X	X		X	X	X	X	X^k	X
69	Time in/outside range of 1 SD⁹⁸													X
70	Total non-consensus glycemic metrics computable	6	5	6	16	5	7	3	1	12	3	16	5	14	31	5	8	0	16	8	6	16	21	9
GRAPHICAL PLOTS
		A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W
71	ADO (aggregate daily overlay) plot^l,6											X			X				X				X
72	AGP (ambulatory glucose profile) report⁶														X								X
73	Antenna plot⁶⁵									X
74	Autocorrelation and partial autocorrelation plots¹⁰¹																				X
75	Bar charts of glucose level ranges¹²							X							X				X					X
76	Boxplots of glucose levels									X
77	CGM trace plot (time series)			X		X	X	X	X	X			X	X	X				X	X	X	X	X
78	Clarke error grid^118,119																						X
79	CVGA (control variability glucose analysis) plot¹⁶																						X
80	Glucodensity graphs⁶⁶																	X
81	Heat map of the correlation between the metrics⁶⁸																			X
82	Histogram of glucose distribution⁶⁰												X
83	Histogram of glucose rate of change¹⁶														X								X
84	Glucotype graphs³⁴								X
85	GRI (glycemia risk index) plot^16,18																						X
86	Lasagna plot¹⁵														X
87	MAGE (mean amplitude of glycemic excursions) plot¹⁵						X				X
88	Mean glucose per subject (color coded plot)¹⁷											X
89	Multiscale entropy plot^m,65									X
90	Poincaré plot (return map)⁶⁰												X
91	Rate of change of glucose levels plot⁶¹					X									X							X	X
92	Symmetrized glucose values over time plot⁶¹																					X
93	Transition density profile^n,32					X
94	Three-dimensional plot of date, time and, glucose level¹⁰¹																				X
95	Time series decomposition plot¹⁰¹																				X
96	Violin plot⁶⁸																			X
97	Total graphical plots computable	0	0	1	0	3	2	2	2	4	1	2	3	1	7	0	0	1	3	3	4	3	8	1
CGM DATA IMPUTATION
		A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W
98	Options for imputation of missing CGM data				X		X			X		X	X	X	X				X		X		X	X
MULTIPLE GROUP COMPARISON
		A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W
98	Options for multiple group comparison									X								X	X	X			X

Abbreviations: AGATA, Automated Glucose dATa Analysis; CGDA, continuous glucose data analysis; CGM-GUIDE, continuous glucose monitoring graphical user interface with diabetes evaluation; CGMTSA, continuous glucose monitoring time series data analysis; GVAP, glycemic variability analyzer program; iglu, interpreting blood GLUcose; KAMOGAWA, Kyoto Auto MAGE Of Glucose cAlculator With isCGM Application; NA, not available.

Times in different ranges are not directly provided. For the GVAP package,²² percentages above and below a customizable target range are computed. For the CGMStatsAnalyser⁶⁸ and CGM Shiny packages, TBR < 3.9 mmol/L or < 70 mg/dL is the sum of TBR level 1 and TBR level 2, and TAR > 10.0 mmol/L or > 180 mg/dL is the sum of TAR level 1 and TAR level 2.

Several subtypes of AUC calculations are possible in the CGDA package,⁶⁷ for example, AUC average per day. The packages CGM-GUIDE,³² GVAP,²² CGMStatsAnalyser,⁶⁸ and AGATA¹⁶ have customizable lower and upper glycemic limits for AUC calculation.

Sum of the absolute difference in glucose levels from successive CGM measurements over a given interval of time.

The GVAP²² package counts excursions (n) above a customizable glycemic amplitude. The packages cgmanalysis¹⁷ and CGDA⁶⁷count excursions (n) in customizable ranges of glucose levels with customizable thresholds for the number of minutes an excursion has to persist to be counted.

GLU package⁶⁰ gives a measure of fasting glucose levels without the knowledge of mealtimes. Calculation is based on the mean of the lowest glucose levels in 30 consecutive minutes during the night time.

The packages Easy GV,³³ iglu,¹⁵ rGV,⁶¹ and AGATA¹⁶ provide computation of the relative contribution of hypoglycemia, euglycemia, and hyperglycemia to the GRADE score.

Interquartile ranges are not directly provided but can be calculated from subtracting the first quartile from the second quartile.

Separates MAGE into MAGE for upward and downward glucose excursions.

Includes glucose rise to peak (ie, difference in glucose levels before meal till glucose peak during the meal), time to peak (ie, difference in time before meal till glucose peak during meal), and % baseline recovery one hour after glucose level peak.

Mean of sensor glucose levels in- and outside range with default settings of 1 SD of mean sensor glucose as inside range.

Several subtypes of within and between days SD calculations are offered in Rodbard²⁸ and SD of within-day mean index and mean of within-day SD index in AGATA.¹⁶

Combination of CGM traces from different days and/or different subjects and combines them in one plot with median and percentiles displayed as a single day as known from the AGP report.⁶ AGATA package¹⁶ also plots CGM traces over multiple days with superimposed aggregated glucose values to facilitate the identification of meaningful hyperglycemic and hypoglycemic events.

Plot the mean and standard error or standard deviation of multiscale entropy by groups.

Frequency of glucose transitions per day as they move between critical glycemic zones.

Table 3.

Statistical Algorithms for Retrospective Analysis of Continuous Glucose Monitoring Data Not Included in Any Statistical Packages.

A		B	C	D	E
Statistical algorithm		Content of the statistical algorithm	Core statistical methods used to develop the statistical algorithm	Subjects (n) and CGM days (n) to illustrate the statistical algorithm	CGM type(s) used to test the statistical algorithm
CGM DATA REDUCTION
1	Hovorka et al¹²⁰	Construction of “hourly circadian sensor modal days” representing superimposed daily profiles with 24 consecutive means of hourly CGM glucose values	Simple arithmetic.	Subjects: n = 16 T1DMCGM days: n = 48	Medtronic MiniMed Continuous Glucose Monitoring System (Medtronic, Northridge, CA)
2	Kovatchev et al¹¹	LBGI/HBGI plot to assess temporal glucose variability. Glucose variability assessed by its Markov chain	Simple arithmetic and Markov process	Not relevant	Not relevant
3	Miller et al¹²¹	Construction of a new CGM trace by reducing the influence of small excursions to allow trends to be visualized in the CGM trace	Discrete Fourier transform methodology	Subjects: n = 90 T1DMCGM days: n = 90	NA
4	Rodbard¹²	Visual representation of glucose levels by stacked bar charts, pie-charts, frequency histograms, and so on for glucose level ranges	Simple arithmetic	Not relevant	Not relevant
5	Rodbard¹³	Graphical scatterplots of several glucose metrics fitted against each other	Simple arithmetic	Not relevant	Not relevant
6	Contreras et al⁷⁷	Clustering of T1DM patients into five motifs based on CGM data and clinical features	Cluster analyses	Subjects: n = 10 T1DM, n = 10 virtual T1DMCGM days: n = 1800 (from real-life patients)	Medtronic Enlite sensors(Medtronic, Northridge, CA)
7	Biagi et al¹²²	Graphical categorization of intra-patient variability of daily glucose profiles into different groups	Compositional data analysis	Subjects: n = 6 T1DMCGM days: n = 330	MiniMed Paradigm REAL-Time Veo System and Enlite glucose sensor (Medtronic, Northridge, CA)
8	Derozier et al⁷⁸	Graphical display with qualitative color scales that exhibits the patient’s overall glycemic state.	Cluster analyses	Subjects: n = 62CGM days: NA	NA
9	Kahkoska et al⁷⁹	Clustering of T1DM patients into three motifs based on eight CGM metrics of hypoglycemia, hyperglycemia, and glycemic variability	Cluster analyses	Subjects: n = 234CGM days: n = 1.600	iPro2 with Enlite sensor (Medtronic, Northridge, CA)
10	Tao et al⁸⁰	Clustering of T2DM patients into four motifs based on CGM-derived metrics, c-peptide index, insulin-sensitivity, and insulin-resistance	Cluster analyses	Subjects: n = 908 T2DMCGM days: n = 2.700	iPro2 with Enlite sensor (Medtronic, Northridge, CA)
11	Herrero et al¹²³	Identification of whether CGM data of 15 days belong to an individual with T1DM or not from nine commonly used CGM metrics	Support vector machine	Subjects: n = 205CGM days: n = 18.400	Dexcom G4 (Dexcom Inc., San Diego, USA)
12	Lim et al⁸¹	Clustering of CGM data by representative daily profiles and glycemic variability metrics	Cluster analyses (variational encoders [VAEs])	Subjects: n = 112 T1DM, n = 20 virtual T1DMCGM days: n = 20.160 from T1DM, n = 400 days from virtual T1DM patients	Dexcom G6 (Dexcom Inc., San Diego, USA)
13	Lobo et al⁸²	Clustering of CGM data into eight motifs based on mean glucose level, TIR, TAR, TBR, SD, CV, LBGI, and HBGI	Cluster analyses	Subjects: n = 611 T1DM, n = 158 T2DMCGM days: n = 42.000	Dexcom G4, G5, and G6 (Dexcom Inc., San Diego, USA)
COMPOSITE CGM OUTCOMES
	A	B	C	D	E
14	Thomas et al⁹⁰	The glucose pentagon, combining mean glucose level, SD, time per day, and AUC with hyperglycemia > 8.9 mmol/L (> 160 mg/dL), and HbA1c	Simple arithmetic	Subjects: n = 3CGM days: n = 9	NA
15	Augstein et al⁹⁵	Q-score, defining glycemic status as very good, good, satisfactory, fair, or poor accounting for mean glucose, MODD, time above range > 8.9 mmol/L (> 160 mg/dL), time below range < 3.9 mmol/L (< 70 mg/dL), and total glucose range	Simple arithmetic	Subjects: n = 48 T1DM, n = 1514 T2DMCGM days: n = 140 from T1DM, n = 4500 T2DM	Medtronic Gold system (Medtronic, Northridge, CA)
16	Vigersky¹⁴	Composite metric of three variables: change in HbA1c, absolute severe hypoglycemia rate (ie, events per 100 person-years), and percentage change in severe hypoglycemia rate	Simple arithmetic	Not relevant	Not relevant
17	Thomas et al¹²⁴	Two composite metrics: Intensity of hypoglycemia (IntHypo) and overall hypoglycemic risk called the “hypoglycemia risk volume” (HypoRV) combining AUC, duration of hypoglycemia and frequency of excursions per day	Simple arithmetic	Subjects: n = 247CGM days: n = 22.000	Enlite glucose sensors (Medtronic, Northridge, CA)
18	Vigersky et al¹²⁵	Comprehensive Glucose Pentagon (CGP) combining mean glucose, CV, severity of hypoglycemia and hyperglycemia, and time out of range (the inverse of time in range)	Simple arithmetic	Subjects: n = 732CGM days: n = 29.000	Enlite glucose sensors (Medtronic, Northridge, CA)
19	Zout et al¹²⁶	Glycemic deviation index (GDI) based on mean glucose index and SD of the glucose index. GDI takes values between 0 and 10 with normal range of 0 to 1	Simple arithmetic	Subjects: n = 30 T2DM and several in-silico simulationsCGM days: n = 30	iPro2 with Enlite sensor (Medtronic, Northridge, CA)
OTHER CGM METRICS
	A	B	C	D	E
20	Marling et al⁹²	CGM traces were scored for MAGE, excursion frequency, and distance traveled and used to make a glycemic variability metric based on perceived variability judged by clinicians	Naive Bayes classifiers	Subjects: n = 11 T1DMCGM days: n = 350	Guardian REAL-Time CGM System (Medtronic, Northridge, CA)
21	Marling et al⁹³	A Consensus Perceived Glycemic Variability (CPGV) was developed by physicians rating CGM traces as exhibiting low, borderline, high, or extremely high GV. Ratings were averaged and put into machine learning algorithms	Multilayer perceptrons and support vector machines for regression	Subjects: n = 19 T1DMCGM days: n= 250	NA
22	Henriques et al¹²⁷	A graphical plot termed glucose-at-a-glance, to visualize short-term and long-term glycemic variability	Density delay maps	Subjects: n = 2 non-diabetes individuals, n = 6 with diabetes (unknown types)CGM days: n = 30	NA
23	Lai et al¹²⁸	Glycemic variability assessment by decomposing the CGM data into separate characteristic traces. Characterizes the entropy, ie, complexity of CGM data	Ensemble empirical mode decomposition and refined composite multiscale entropy	Subjects: n = 98 T2DMCGM days: n = 294	CGMs (Medtronic, Northridge, CA), not further specified

Table 3 provides information on the statistical algorithms with (A) authors of the statistical algorithm and publication year, (B) summary of content, (C) core statistical methods used to develop the algorithm, (D) testing of the algorithm in (n) subjects and on (n) CGM days, and (E) CGM type(s) used to illustrate the algorithm.

Abbreviations: AUC, area under curve; CGM, continuous glucose monitoring; CGP, comprehensive glucose pentagon; CONGA, continuous overall net glycemic action; CPGV, consensus perceived glycemic variability; CV, control variability; GCI, glycemic deviation index; HBGI, high blood glucose index; LBGI, low blood glucose index; MAGE, mean amplitude of glycemic excursions; MODD, mean of daily differences; NA, not available; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; TAR, time above range; TBR, time below range; TIR, time in range.

Results

Statistical Packages

We identified 23 statistical packages (Table 1) and have tabulated the computable CGM metrics (Table 2). The packages can compute the metrics listed in Table 2 regardless of glucose reporting frequency specific for the CGM device, often every 1, 5, or 15 minutes. In total, 14 (61%) of the packages had a graphical user interface, input formats were .csv, .txt, .mat, .xls, or .xlsx and glucose level units were in mmol/L and mg/dL (manual conversions are possible between formats and units). Furthermore, 20 (87%) of the statistical packages could export/download results, 16 (70%) were open source, 22 (96%) were publicly available, 11 (48%) could impute missing CGM data, and five (24%) could perform multiple group comparisons.

Statistical Algorithms Not Included in Statistical Packages

We identified 23 statistical algorithms not included in any statistical packages but still relevant for retrospective analysis of CGM data (Table 3). The statistical algorithms are grouped by: (1) CGM data reduction (eg, clustering of CGM data), (2) composite CGM outcomes, and (3) other CGM metrics.

In total, 19 (83%) of the statistical algorithms were illustrated with use of CGM data in the original articles. The remaining four (17%) of the statistical algorithms by Kovatchev et al,¹¹ Rodbard,¹² Rodbard,¹³ and Vigersky¹⁴ are conceptual and independent of specific CGM types and/or patient characteristics and not illustrated with use of CGM data.

Discussion

Statistical Packages

Usability and compatibility

This systematic review identified 23 statistical packages for the retrospective analysis of CGM data (Table 1). The packages iglu,¹⁵ AGATA,¹⁶ and cgmanalysis¹⁷ could calculate the most CGM metrics (n = 47, n = 37, and n = 31, respectively) and also the most of the 2022 CGM consensus metrics (n = 16, n = 16, and n = 15, respectively) (Table 2) which makes these packages highly usable. The iglu¹⁵ and AGATA¹⁶ packages were the only ones to calculate glycemia risk index (GRI).¹⁸ None of the statistical packages could estimate the scanning frequency (for flash glucose monitoring) nor the several secondary endpoints (binary outcomes) or composite endpoints of the 2022 CGM consensus.³

GlyCulator 3.0¹⁹ and Tidepool Platform²⁰ provide a graphical user interface and video tutorial on how to use them and may be preferable for basic researchers not skilled in, for example, RStudio. Tidepool also has the advantage that patients can share access to glucose data with their health care professionals or relatives (including data from insulin pumps, blood glucose meters, and ketone monitors) and can therefore be used for telemedicine. The packages AGATA,¹⁶ MAGECAA,²¹ and GVAP²² require MATLAB to run, which might limit the usability of those solutions. The additional statistical packages (Table 1) run in RStudio, in Excel, or as interactive web Shiny apps.

Imputation of missing data

When working with CGM data, censored and missing data frequently occur.²³ At least, 70% of data availability over 14 days is recommended.^3,24 Recently, a greater awareness of missing data in research has emerged.^25-27 Currently, no consensus exists on how to manage missing CGM data.²³ In the present review, 11 (48%) of the statistical packages could impute missing data (Table 2), including the previous mentioned packages iglu,¹⁵ AGATA,¹⁶ and cgmanalysis.¹⁷

Glycemic variability

A strong focus on assessing glycemic variability more precisely with CGM than is possible by POC (point-of-care) glucose data are reflected in the first statistical packages released by Rodbard,²⁸ Baghurst,²⁹ Fritzsche et al,³⁰ Hill et al,³¹ and Rawlings et al³² (Table 1). Interestingly, the top 3 cited packages according to Google per year, Rodbard²⁸ (average of 17.3 citations yearly), EasyGV³³ (average of 34.0 citations yearly), and Hall et al³⁴ (average of 37.6 citations yearly), all focus on calculating glycemic variability.

Glycemic variability was conceptually introduced many years before CGM as evidenced by the mean amplitude of glycemic excursions (MAGE) in 1970,³⁵ the mean of daily differences (MODD) in 1972,³⁶ and the J-index in 1995.³⁷ Today, the most popular metric for reporting glycemic variability is the coefficient of variation (CV). It has been recommended that CV is less than or equal to 36% for patients with type 1 diabetes.³ Several other metrics of glycemic variability are also validated and are implemented in statistical packages (Table 2). Nearly, all glycemic variability metrics are highly correlated with standard deviation (SD) of glucose distribution.^38-40 Recent reviews have been published on the clinical implications of glycemic variability in diabetes.^41-47

Graphical display options

Continuous glucose monitoring summary statistics do not, in general, consider that CGM data are time series data.⁴⁸ For example, consider CGM glucose data ordered in time from 3.0 to 3.5 to 4.0 mmol/L (54-63-72 mg/dL) and from 4.0 to 3.5 to 3.0 mmol/L (72-63-54 mg/dL). These glucose sequences are clinically very different situations regarding hypoglycemia but yield identical summary statistics, for example, mean glucose level and CV. The example demonstrates the need for clinical judgment and visual interpretation of CGM data.^49-52 The statistical packages iglu¹⁵ and AGATA¹⁶ provide the most graphical display options (Table 2) alongside summary statistics.

Standardization vs customization

The AGP^4,53 is the basis for the first consensus on CGM metrics and visualization released in 2012 by the International Diabetes Center.³⁸ The AGP report has three major components: (1) glucose statistics and time in different glucose ranges, (2) a single 24-hour “modal” day that combines all daily CGM glucose data, referred to as AGP, and (3) daily glucose profiles.⁶ Commercial tools for retrospective analysis and visualization of CGM data are based on AGP principles (eg, Glooko,⁵⁴ Dexcom Clarity,⁵⁵ Eversense Data Management System,⁵⁶ Medtronic Carelink,⁵⁷ and Abbot LibreView⁵⁸). These commercial solutions compute only a subset of CGM metrics, often time in different glucose ranges, CV, mean glucose levels, and glucose management indicator (GMI). While simple, standardized overviews of CGM data are often useful in clinical practice, researchers often need different metrics and graphical display options of CGM data to answer specific research questions and/or to align with the 2022 CGM consensus.³ Standardization must be done in a manner that permits diversity of clinical situations and fosters research innovations.⁵⁹

Both the statistical packages (Tables 1 and 2) and the statistical algorithms (Table 3) can complement the standardized AGP report and generate information not currently provided in the AGP report, for example, possible changes in glycemic target ranges more suitable for the patients prone to hypoglycemia, elderly patients or during pregnancy. Customizable thresholds for time in different ranges are possible to compute by the packages: EasyGV,³³ CGM-GUIDE,³² GVAP,²² Tidepool²⁰ cgmanalysis,¹⁷ GLU,⁶⁰ iglu,¹⁵ rGV,⁶¹ and GlyCulator 3.0.¹⁹

It has been proposed⁶² that the AGP report should also include (1) hypoglycemic and hyperglycemic events (which is possible by the packages iglu¹⁵ and AGATA¹⁶), (2) details on postprandial excursions (which is possible by packages GLU⁶⁰ and iglu¹⁵), (3) account for the skewed distribution of CGM data (which is possible by several of the statistical packages (Table 2) by blood glucose risk index [BGRI],¹¹ low blood glucose index [LBGI],^63,64 and high blood glucose index [HBGI]⁶⁴), and (4) provide methods to compare CGM metrics between groups or individuals, which is possible by the statistical packages^16,65-68 (Table 1).

Statistical Algorithms

This systematic review identified 23 statistical algorithms (Table 3) for the retrospective analysis of CGM data but not included in any of the 23 statistical packages (Tables 1 and 2).

CGM data reduction

The purpose of most of the reviewed algorithms is to cluster data. Subgroups or “clusters” of patients with diabetes might foster precision medicine initiatives aiming to characterize homogeneous subgroups based on the individuals’ characteristics.^69-71 In time, this could lead to the development of targeted interventions for disease prevention.^72,73 Recently, data-driven analytical methods have identified distinctive subgroups of individuals with type 2 diabetes or prediabetes and their associations with diabetes complications.^74-76 A barrier to implementing some of the statistical algorithms in daily clinical practice is that the clustering algorithms reviewed here^77-82 might require initial training on data in similar setups used by their developers. Therefore, the algorithms are probably primarily useful in research.

Composite CGM outcomes

Several composite CGM metrics have been proposed (Table 3). There are clinical advantages to composite metrics.⁸³ Composite metrics giving weight to hypoglycemia and hyperglycemia⁸⁴ might be an advantage since some of the widely used CGM metrics, for example, CV, time in range (TIR), and mean glucose level, are more correlated with hyperglycemia due to the skewed glucose level distribution.^84-88 Algorithms accounting for skewed glucose distributions have been developed, for example, BGRI,¹¹ LBGI,^63,64 HBGI,⁶⁴ index of glycemic control (IGC),⁴⁰ and glycemic risk assessment diabetes equation (GRADE).⁸⁹ However, these metrics are often neglected, possibly due to the difficulty of computing them⁶¹ and/or interpreting them.⁸⁴ Today, several statistical packages offer easy calculations of these metrics (Table 2). The statistical algorithms by Thomas et al⁹⁰ and Vigersky¹⁴ include HbA1c as one component in the composite metric and are not exclusively CGM-based. Reviews on CGM-based composite metrics have been published.^88,91 Recently, the composite metric GRI¹⁸ has been included in the 2022 CGM consensus.³

Other CGM metrics

The additional CGM metrics identified (Table 3) all focus on glycemic variability. Interestingly, the metrics developed by Marling et al^92,93 are based on clinicians’ perceived risk regarding glycemic variability from CGM glucose traces and not based on more traditional analytical data approaches as cluster analyses or constructing composite metrics from already established glycemic metrics as reviewed above. More established CGM metrics also taking advantage of such perceived risks include the GRADE score,⁹⁴ the Q-score,⁹⁵ the IGC,⁴⁰ and the GRI,¹⁸ all based on a clinical perception of the risk of adverse outcomes associated with a glucose profile.⁸⁹ In relation to the GRADE score, the perceived risk differs from health care professionals and patients.⁹⁴ These metrics are developed to capture what clinicians perceive and treat as dangerous glucose levels and might supplement more traditional ways of interpreting CGM data.

A gold standard metric or algorithm

Recent work has been performed to identify gold standard metrics and algorithms to optimize the use of CGM data by the discriminant ratio.⁹⁶ The discriminant ratio is a way to compare different metrics to determine the ability of each metric to discriminate between different people with diabetes. Mean absolute glucose (MAG), LBGI, and TIR have the highest discriminant ratio in the tested metrics.⁹⁶ However, more research is needed to establish correlations between these metrics and short-term and long-term diabetes complications. Therefore, clinicians and researchers continue to balance between not having too many CGM metrics that they become confusing⁹⁷ and still having enough metrics to match the sophistication of CGM technology.

Conclusion

This systematic review identified 23 statistical packages (Table 1) for retrospective analysis of CGM data with textual and tabular (Table 2) overviews of computable CGM metrics, plots, and graphs in the statistical packages and whether imputation of missing CGM data is possible. An additional 23 statistical algorithms were identified (Table 3) that are relevant for retrospective analysis of CGM data not included in the statistical packages. The statistical packages and algorithms supplement CGM manufacturers’ CGM reports with additional CGM metrics, plots, and graphs according to specific research questions or clinical situations.

Future Perspectives

Statistical methods to handle censored CGM data are needed. Options for group comparisons and graphical user phase solutions for non-programmers should be expanded since only five (22%) and 14 (61%) of the statistical packages have these options, respectively.

Supplemental Material

sj-docx-1-dst-10.1177_19322968231221803 – Supplemental material for Statistical Packages and Algorithms for the Analysis of Continuous Glucose Monitoring Data: A Systematic Review

Supplemental material, sj-docx-1-dst-10.1177_19322968231221803 for Statistical Packages and Algorithms for the Analysis of Continuous Glucose Monitoring Data: A Systematic Review by Mikkel Thor Olsen, Carina Kirstine Klarskov, Arnold Matovu Dungu, Katrine Bagge Hansen, Ulrik Pedersen-Bjergaard and Peter Lommer Kristensen in Journal of Diabetes Science and Technology

Footnotes

Acknowledgements

The authors are grateful for the advice on the search strings by information specialist Jette Melby at Copenhagen University Hospital—North Zealand, Denmark.

Abbreviations

AGP, ambulatory glucose profile; BGRI, blood glucose risk index; CGM, continuous glucose monitoring; GMI, glucose management indicator; GRADE, glycemic risk assessment diabetes equation; GRI, glycemia risk index; HbA1c, hemoglobin a1c; HBGI, high blood glucose index; IGC, index of glycemic control; LBGI, low blood glucose index; MAG, mean absolute glucose; SD, standard deviation; TIR, time in range.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Novo Nordisk Foundation (grant no. NNF20SA0062872).

ORCID iDs

Mikkel Thor Olsen

Carina Kirstine Klarskov

Supplemental Material

Supplemental material for this article is available online.

References

Díez-Fernández

Rodríguez-Huerta

Mirón-González

Laredo-Aguilera

Martín-Espinosa

NM.

Flash glucose monitoring and patient satisfaction: a meta-review of systematic reviews. Int J Environ Res Public Heal. 2021;18(6):3123. doi:10.3390/ijerph18063123

Galindo

Umpierrez

Rushakoff

, et al. Continuous glucose monitors and automated insulin dosing systems in the hospital consensus guideline. J Diabetes Sci Technol. 2020;14(6):1035-1064. doi:10.1177/1932296820954163

Battelino

Alexander

Amiel

, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol. 2022;11(1):42-57. doi:10.1016/S2213-8587(22)00319-9

Mazze

Strock

Wesley

, et al. Characterizing glucose exposure for individuals with normal glucose tolerance using continuous glucose monitoring and ambulatory glucose profile analysis. Diabetes Technol Ther. 2008;10(3):149-159. doi:10.1089/dia.2007.0293

Mazze

Strock

Morgan

Wesley

Bergenstal

Cuddihy

Diurnal glucose patterns of exenatide once weekly: a 1-year study using continuous glucose monitoring with ambulatory glucose profile analysis. Endocr Pract. 2009;15(4):326-334. doi:10.4158/EP09046.ORR

Johnson

Martens

Criego

Carlson

Simonson

Bergenstal

RM.

Utilizing the ambulatory glucose profile to standardize and implement continuous glucose monitoring in clinical practice. Diabetes Technol Ther. 2019;21(S2):S217-S225. doi:10.1089/dia.2019.0034

Olsen

Dungu

Klarskov

Jensen

Lindegaard

Kristensen

PL.

Glycemic variability assessed by continuous glucose monitoring in hospitalized patients with community-acquired pneumonia. BMC Pulm Med. 2022;22(1):83. doi:10.1186/s12890-022-01874-7

Klarskov

Windum

Olsen

, et al. Telemetric continuous glucose monitoring during the COVID-19 pandemic in isolated hospitalized patients in Denmark: a randomized controlled exploratory trial. Diabetes Technol Ther. 2022;24(2):102-112. doi:10.1089/dia.2021.0291

Piersanti

Giurato

Göbl

Burattini

Tura

Morettini

Software packages and tools for the analysis of continuous glucose monitoring data. Diabetes Technol Ther. 2023;25(1):69-85. doi:10.1089/dia.2022.0237

10.

Page

Mckenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;29(372):n71. doi:10.1136/bmj.n71

11.

Kovatchev

Clarke

Breton

Brayman

McCall

Quantifying temporal glucose variability in diabetes via continuous glucose monitoring: mathematical methods and clinical application. Diabetes Technol Ther. 2005;7(6):849-862. doi:10.1089/dia.2005.7.849

12.

Rodbard

Display of glucose distributions by date, time of day, and day of week: new and improved methods. J Diabetes Sci Technol. 2009;3(6):1388-1394. doi:10.1177/193229680900300619

13.

Rodbard

Evaluating quality of glycemic control: graphical displays of hypo- and hyperglycemia, time in target range, and mean glucose. J Diabetes Sci Technol. 2015;9(1):56-62. doi:10.1177/1932296814551046

14.

Vigersky

RA.

Escaping the hemoglobin A1c-centric world in evaluating diabetes mellitus interventions. J Diabetes Sci Technol. 2015;9(5):1148-1151. doi:10.1177/1932296815576363

15.

Broll

Urbanek

Buchanan

, et al. Interpreting blood GLUcose data with R package iglu. PLoS One. 2021;16(4):e0248560. doi:10.1371/journal.pone.0248560

16.

Cappon

Sparacino

Facchinetti

. AGATA: a toolbox for automated glucose data analysis [published online ahead of print January 5, 2023]. J Diabetes Sci Technol. doi:10.1177/19322968221147570

17.

Vigers

Chan

Snell-Bergeon

, et al. Cgmanalysis: an R package for descriptive analysis of continuous glucose monitor data. PLoS One. 2019;14(10):e0216851. doi:10.1371/journal.pone.0216851

18.

Klonoff

Wang

Rodbard

, et al. A glycemia risk index (GRI) of hypoglycemia and hyperglycemia for continuous glucose monitoring validated by clinician ratings. J Diabetes Sci Technol. 2023;17(5):1226-1242. doi:10.1177/19322968221085273

19.

Chrzanowski

Grabia

Michalak

, et al. GlyCulator 3.0: a fast, easy-to-use analytical tool for CGM data analysis, aggregation, center benchmarking, and data sharing. Diabetes Care. 2023;46(1):e3-e5. doi:https://doi.org/10.2337/dc22-0534

20.

Neinstein

Wong

Look

, et al. A case study in open source innovation: developing the Tidepool Platform for interoperability in type 1 diabetes management. J Am Med Inform Assoc. 2016;23(2):324-332. doi:10.1093/jamia/ocv104

21.

Lin

Shen

, et al. Calculating the mean amplitude of glycemic excursions from continuous glucose data using an open-code programmable algorithm based on the integer nonlinear method. Comput Math Methods Med. 2018;2018:6286893. doi:10.1155/2018/6286893

22.

Marics

Lendvai

Lódi

, et al. Evaluation of an open access software for calculating glucose variability parameters of a continuous glucose monitoring system applied at pediatric intensive care unit. Biomed Eng Online. 2015;14:37. doi:10.1186/s12938-015-0035-3

23.

Cuesta-Frau

Novák

Burda

, et al. Characterization of artifact influence on the classification of glucose time series using sample entropy statistics. Entropy. 2018;20(11):871. doi:10.3390/e20110871

24.

Battelino

Danne

Bergenstal

, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care. 2019;42(8):1593-1603. doi:10.2337/dci19-0028

25.

European Medicines Agency. ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials. 2020. https://www.ema.europa.eu/en/documents/scientific-guideline/ich-e9-r1-addendum-estimands-sensitivity-analysis-clinical-trials-guideline-statistical-principles_en.pdf. Accessed April 24, 2023.

26.

European Medicines Agency. Guideline on missing data in confirmatory clinical trials. 2010. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-missing-data-confirmatory-clinical-trials_en.pdf. Accessed May 16, 2022.

27.

National Research Council, Panel on Handling Missing Data in Clinical Trials. The prevention and treatment of missing data in clinical trials. 2010. https://nap.nationalacademies.org/catalog/12955/the-prevention-and-treatment-of-missing-data-in-clinical-trials. Accessed May 16, 2022.

28.

Rodbard

New and improved methods to characterize glycemic variability using continuous glucose monitoring. Diabetes Technol Ther. 2009;11(9):551-565. doi:10.1089.dia.2009.0015

29.

Baghurst

PA.

Calculating the mean amplitude of glycemic excursion from continuous glucose monitoring data: an automated algorithm. Diabetes Technol Ther. 2011;13(3):296-302. doi:10.1089/dia.2010.0090

30.

Fritzsche

Kohnert

Heinke

Vogt

Salzsieder

The use of a computer program to calculate the mean amplitude of glycemic excursions. Diabetes Technol Ther. 2011;13(3):319-325. doi:10.1089/dia.2010.0108

31.

Hill

Oliver

Choudhary

Levy

Hindmarsh

Matthews

DR.

Normal reference range for mean tissue glucose and glycemic variability derived from continuous glucose monitoring for subjects without diabetes in different ethnic groups. Diabetes Technol Ther. 2011;13(9):921-928. doi:10.1089/dia.2010.0247

32.

Rawlings

Shi

Yuan

Brehm

Pop-Busui

Nelson

PW.

Translating glucose variability metrics into the clinic via continuous glucose monitoring: a graphical user interface for diabetes evaluation (CGM-GUIDE). Diabetes Technol Ther. 2011;13(12):1241-1248. doi:10.1089/dia.2011.0099

33.

Moscardó

Giménez

Oliver

Hill

NR.

Updated software for automated assessment of glucose variability and quality of glycemic control in diabetes. Diabetes Technol Ther. 2020;22(10):701-708. doi:10.1089/dia.2019.0416

34.

Hall

Perelman

Breschi

, et al. Glucotypes reveal new patterns of glucose dysregulation. PLoS Biol. 2018;16(7):e2005143. doi:10.1371/journal.pbio.2005143

35.

Service

Molnar

Rosevear

Ackerman

Gatewood

Taylor

WF.

Mean amplitude of glycemic excursions, a measure of diabetic instability. Diabetes. 1970;19(9):644-655. doi:10.2337/diab.19.9.644

36.

Molnar

Taylor

MM.

Day-to-day variation of continuously monitored glycaemia: a further measure of diabetic instability. Diabetologia. 1972;8(5):342-348. doi:10.1007/BF01218495

37.

Wójcicki

JM.

“J”-index. A new proposition of the assessment of current glucose control in diabetic patients. Horm Metab Res. 1995;27(1):41-42. doi:10.1055/s-2007-979906

38.

Bergenstal

Ahmann

Bailey

, et al. Recommendations for standardizing glucose reporting and analysis to optimize clinical decision making in diabetes: the ambulatory glucose profile (AGP). Diabetes Technol Ther. 2013;15(3):198-211. doi:10.1089/dia.2013.0051

39.

Fabris

Facchinetti

Sparacino

, et al. Glucose variability indices in type 1 diabetes: parsimonious set of indices revealed. Diabetes Technol Ther. 2014;16(10):644-652. doi:10.1089/dia.2013.0252

40.

Rodbard

Interpretation of continuous glucose monitoring data: glycemic variability and quality of glycemic control. Diabetes Technol Ther. 2009;11(suppl 1):S55-S67. doi:10.1089/dia.2008.0132

41.

Monnier

Colette

Owens

DR.

Glycemic variability: the third component of the dysglycemia in diabetes. Is it important? how to measure it?

J Diabetes Sci Technol. 2008;2(6):1094-1100. doi:10.1177/193229680800200618

42.

Suh

Kim

JH.

Glycemic variability: how do we measure it and why is it important?

Diabetes Metab J. 2015;39(4):273-282. doi:10.4093/dmj.2015.39.4.273

43.

Ceriello

Monnier

Owens

Glycaemic variability in diabetes: clinical and therapeutic implications. Lancet Diabetes Endocrinol. 2019;7(3):221-230. doi:10.1016/S2213-8587(18)30136-0

44.

Candido

Which patients should be evaluated for blood glucose variability. Diabetes Obes Metab. 2013;15(suppl 2):9-12.

45.

Rodbard

Glucose variability: a review of clinical applications and research developments. Diabetes Technol Ther. 2018;20(S2):S25-S215. doi:10.1089/dia.2018.0092

46.

Monnier

Colette

Owens

DR.

The application of simple metrics in the assessment of glycaemic variability. Diabetes Metab. 2018;44(4):313-319. doi:10.1016/j.diabet.2018.02.008

47.

Monnier

Colette

Glycemic variability. Should we and can we prevent it?

Diabetes Care. 2008;31(suppl 2):150-154. doi:10.2337/dc08-s241

48.

Kovatchev

Breton

Clarke

Analytical methods for the retrieval and interpretation of continuous glucose monitoring data in diabetes. Methods Enzymol. 2009;454:69-86. doi:10.1016/S0076-6879(08)03803-2

49.

Pearson

Bergenstal

Fine-tuning control: pattern management versus supplementation: view 1: pattern management: an essential component of effective insulin management. Diabetes Spectr. 2001;14(2):75-78.

50.

Ritholz

Atakov-Castillo

Beste

, et al. Psychosocial factors associated with use of continuous glucose monitoring. Diabet Med. 2010;27(9):1060-1065. doi:10.1111/j.1464-5491.2010.03061.x

51.

Scheiner

CGM retrospective data analysis. Diabetes Technol Ther. 2016;18(suppl 2):S214-S222. doi:10.1089/dia.2015.0281

52.

Bowen

Rumana

Killgore

Gong

A user-centered glucose-insulin data display for the inpatient setting. Stud Heal Technol Inform. 2018;245:684-688.

53.

Mazze

Lucido

Langer

Hartmann

Rodbard

Ambulatory glucose profile: representation of verified self-monitored blood glucose data. Diabetes Care. 1987;10(1):111-117. doi:10.2337/diacare.10.1.111

54.

Glooko, Inc. Glooko. n.d. https://glooko.com/providers/. Accessed July 17, 2023.

55.

Dexcom, Inc. Dexcom clarity. n.d. https://www.dexcom.com/en-gb. Accessed July 17, 2023.

56.

Senseornics, Inc. Eversense. n.d. https://global.eversensediabetes.com/patient-education/eversense-user-guides/. Accessed July 17, 2023.

57.

Medtronic. Medtronic CareLink. n.d. https://www.medtronicdiabetes.com/customer-support/carelink-software-support/carelink-reports. Accessed July 17, 2023.

58.

Abbott Diabetes Care, Inc. LibreView. n.d. https://www.freestyle.abbott/ch-it/prodotti/libre-view.html. Accessed July 17, 2023.

59.

Rodbard

Standardization versus customization of glucose reporting. Diabetes Technol Ther. 2013;15(5):439-443. doi:10.1089/dia.2013.0116

60.

Millard

LAC

Patel

Tilling

Lewcock

Flach

Lawlor

. GLU: a software package for analysing continuously measured glucose levels in epidemiology. Int J Epidemiol. 2020;49(3):744-757. doi:10.1093/ije/dyaa004

61.

Olawsky

Zhang

Eberly

Helgeson

Chow

LS.

A new analysis tool for continuous glucose monitor data. J Diabetes Sci Technol. 2022;16(6):1496-1504. doi:10.1177/19322968211028909

62.

Rodbard

The ambulatory glucose profile: opportunities for enhancement. Diabetes Technol Ther. 2021;23(5):332-341. doi:10.1089/dia.2020.0524

63.

Kovatchev

Cox

Gonder-Frederick

Young-Hyman

Schlundt

Clarke

Assessment of risk for severe hypoglycemia among adults with IDDM: validation of the low blood glucose index. Diabetes Care. 1998;21(11):1870-1875. doi:10.2337/diacare.21.11.1870

64.

Kovatchev

Straume

Cox

Farhy

LS.

Risk analysis of blood glucose data: a quantitative approach to optimizing the control of insulin dependent diabetes. J Theor Med. 2000;3:208936. doi:10.1080/10273660008833060

65.

Zhang

Wang

CGManalyzer: an R package for analyzing continuous glucose monitoring studies. Bioinformatics. 2018;34(9):1609-1611. doi:10.1093/bioinformatics/btx826

66.

Matabuena

Petersen

Vidal

Gude

Glucodensities: a new representation of glucose profiles using distributional data analysis. Stat Methods Med Res. 2021;30(6):1445-1464. doi:10.1177/0962280221998064

67.

Attaye

van der Vossen

EWJ

Mendes Bastos

Nieuwdorp

Levin

Introducing the continuous glucose data analysis (CGDA) R package: an intuitive package to analyze continuous glucose monitoring data. J Diabetes Sci Technol. 2022;16(3):783-785. doi:10.1177/19322968211070293

68.

De Livera

Shaw

Cohen

Reutens

Salim

An interactive web application for the statistical analysis of continuous glucose monitoring data in epidemiological studies. J Diabetes Sci Technol. 2022;16(2):397-400. doi:10.1177/1932296820985570

69.

Fitipaldi

McCarthy

Florez

Franks

PW.

A global overview of precision medicine in type 2 diabetes. Diabetes. 2018;67(10):1911-1922. doi:10.2337/dbi17-0045

70.

Prasad

Groop

Precision medicine in type 2 diabetes. J Intern Med. 2018;285(1):40-48. doi:10.1111/joim.12859

71.

Méndez

Zhou

Lagerros

, et al. Characterization of data-driven clusters in diabetes-free adults and their utility for risk stratification of type 2 diabetes. BMC Med. 2022;20(1):356. doi:10.1186/s12916-022-02551-6

72.

Khoury

Iademarco

Riley

WT.

Precision public health for the era of precision medicine. Am J Prev Med. 2016;50(3):398-401. doi:10.1016/j.amepre.2015.08.031

73.

Bonnefond

Froguel

Clustering for a better prediction of type 2 diabetes mellitus. Nat Rev Endocrinol. 2021;17(4):193-194. doi:10.1038/s41574-021-00475-4

74.

Cheng

W-Y

Glicksberg

, et al. Identification of type 2 diabetes subgroups through topological analysis of patient similarity. Sci Transl Med. 2015;7(311):311ra174. doi:10.1126/scitranslmed.aaa936

75.

Safai

Ali

Rossing

Ridderstråle

Stratification of type 2 diabetes based on routine clinical markers. Diabetes Res Clin Pract. 2018;141:275-283. doi:10.1016/j.diabres.2018.05.014

76.

Ahlqvist

Storm

Käräjämäki

, et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol. 2018;6(5):361-369. doi:10.1016/S2213-8587(18)30051-2

77.

Contreras

Quirós

Giménez

Conget

Vehi

Profiling intra-patient type I diabetes behaviors. Comput Methods Programs Biomed. 2016;136:131-141. doi:10.1016/j.cmpb.2016.08.022

78.

Derozier

Arnavielhe

Renard

Dray

Martin

How knowledge emerges from artificial intelligence algorithm and data visualization for diabetes management. J Diabetes Sci Technol. 2019;13(4):698-707. doi:10.1177/1932296819847739

79.

Kahkoska

Adair

Aiello

, et al. Identification of clinically relevant dysglycemia phenotypes based on continuous glucose monitoring data from youth with type 1 diabetes and elevated hemoglobin A1c. Pediatr Diabetes. 2019;20(5):556-566. doi:10.1111/pedi.12856.Identification

80.

Tao

, et al. Multilevel clustering approach driven by continuous glucose monitoring data for further classification of type 2 diabetes. BMJ Open Diabetes Res Care. 2021;9(1):e001869. doi:10.1136/bmjdrc-2020-001869

81.

Lim

Cho

Kim

Multi-task disentangled autoencoder for time-series data in glucose dynamics. IEEE J Biomed Health Inform. 2022;26(9):4702-4713. doi:10.1109/JBHI.2022.3175928

82.

Lobo

Farhy

Shafiei

Kovatchev

A data-driven approach to classifying daily continuous glucose monitoring (CGM) time series. IEEE Trans Biomed Eng. 2022;69(2):654-665. doi:10.1109/TBME.2021.3103127

83.

Ferreira-González

Busse

Heels-Ansdell

, et al. Problems with use of composite endpoints in cardiovascular trials: systematic review of randomised controlled trials. BMJ. 2007;334(7597):786. doi:10.1136/bmj.39136.682083.AE

84.

Rodbard

. Metrics to evaluate quality of glycemic control: comparison of time in target, hypoglycemic, and hyperglycemic ranges with ‘“risk indices..” Diabetes Technol Ther. 2018;20(5):325-334. doi:10.1089/dia.2017.0416

85.

Rodbard

A semilogarithmic scale for glucose provides a balanced view of hyperglycemia and hypoglycemia. J Diabetes Sci Technol. 2009;3(6):1395-1401. doi:10.1177/193229680900300620

86.

Kovatchev

Cox

Gonder-Frederick

Clarke

Symmetrization of the blood glucose measurement scale and its applications. Diabetes Care. 1997;20(11):1655-1658. doi:10.2337/diacare.20.11.1655

87.

Brown

Basu

Kovatchev

BP.

Beyond HbA 1c: using continuous glucose monitoring metrics to enhance interpretation of treatment effect and improve clinical decision-making. Diabet Med. 2019;36(6):679-687. doi:10.1111/dme.13944

88.

Rodbard

Quality of glycemic control: assessment using relationships between metrics for safety and efficacy. Diabetes Technol Ther. 2021;23(10):692-704. doi:10.1089/dia.2021.0115

89.

Hill

Hindmarsh

Stevens

Stratton

Levy

Matthews

DR.

A method for assessing quality of control from glucose profiles. Diabet Med. 2007;24(7):753-758. doi:10.1111/j.1464-5491.2007.02119.x

90.

Thomas

Schönauer

Achermann

, et al. The “glucose pentagon”: assessing glycemic control of patients with diabetes mellitus by a model integrating different parameters from glucose profile. Diabetes Technol Ther. 2009;11(6):399-409. doi:10.1089/dia.2008.0119

91.

Nguyen

Han

Spanakis

Kovatchev

Klonoff

DC.

A review of continuous glucose monitoring-based composite metrics for glycemic control. Diabetes Technol Ther. 2020;22(8):613-622. doi:10.1089/dia.2019.0434.A

92.

Marling

Shubrook

Vernier

Wiley

Schwartz

FL.

Characterizing blood glucose variability using new metrics with continuous glucose monitoring data. J Diabetes Sci Technol. 2011;5(4):871-878. doi:10.1177/193229681100500408

93.

Marling

Struble

Bunescu

Shubrook

Schwartz

FL.

A consensus perceived glycemic variability metric. J Diabetes Sci Technol. 2013;7(4):871-879. doi:10.1177/193229681300700409

94.

Hill

Thompson

Bruce

Matthews

Hindmarsh

Glycaemic risk assessment in children and young people with Type 1 diabetes mellitus. Diabet Med. 2009;26(7):740-743. doi:10.1111/j.1464-5491.2009.02763.x

95.

Augstein

Heinke

Vogt

, et al. Q-Score: development of a new metric for continuous glucose monitoring that enables stratification of antihyperglycaemic therapies. BMC Endocr Disord. 2015;15:22. doi:10.1186/s12902-015-0019-0

96.

Moscardó

Herrero

Reddy

Hill

Georgiou

Oliver

Assessment of glucose control metrics by discriminant ratio. Diabetes Technol Ther. 2020;22(10):719-726. doi:10.1089/dia.2019.0415

97.

Rodbard

The challenges of measuring glycemic variability. J Diabetes Sci Technol. 2012;6(3):712-715. doi:10.1177/193229681200600328

98.

Bent

Henriquez

Dunn

Cgmquantify: Python and R software packages for comprehensive analysis of interstitial glucose and glycemic variability from continuous glucose monitor data. IEEE Open J Eng Med Biol. 2021;2:263-266.

99.

Kitae

Kimura

Munekawa

, et al. Development of application to automatically calculate mean amplitude of glycaemic excursions using intermittently scanned continuous glucose monitoring data. Diabetes Obes Metab. 2021;23(9):2155-2160. doi:10.1111/dom.14457

100.

Hangelbroek

CGM shiny statistical software package. 2021. https://github.com/rolandhangelbroek/cgmshiny/blob/main/DESCRIPTION. Accessed October 12, 2023.

101.

Shao

Liu

, et al. Continuous glucose monitoring time series data analysis: a time series analysis package for continuous glucose monitoring data. J Comput Biol. 2023;30(1):112-116. doi:10.1089/cmb.2022.0100

102.

Pagacz

Stawiski

Szadkowska

Mlynarski

Fendler

GlyCulator2: an update on a web application for calculation of glycemic variability indices. Acta Diabetol. 2018;55(8):877-880. doi:10.1007/s00592-018-1140-0

103.

Czerwoniuk

Fendler

Walenciak

Mlynarski

GlyCulator: a glycemic variability calculation tool for continuous glucose monitoring data. J Diabetes Sci Technol. 2011;5(2):447-451. doi:10.1177/193229681100500236

104.

Kovatchev

Otto

Cox

Gonder-Frederick

Clarke

Evaluation of a new measure of blood glucose variability in diabetes. Diabetes Care. 2006;29(11):2433-2438. doi:10.2337/dc06-1085

105.

Leelarathna

Thabit

Wilinska

, et al. Evaluating glucose control with a novel composite continuous glucose monitoring index. J Diabetes Sci Technol. 2020;14(2):277-283. doi:10.1177/1932296819838525

106.

Leelarathna

Hood

Bally

, et al. A novel composite glucose index (COGI) for evaluating closed-loop performance in type 1 diabetes. Diabetes. 2023;67(suppl 1):926-P. doi:10.2337/db18-926-P

107.

McDonnell

Donath

Vidmar

Werther

Cameron

FJ.

A novel approach to continuous glucose analysis utilizing glycemic variation. Diabetes Technol Ther. 2005;7(2):253-263.

108.

Magni

Raimondo

Man

, et al. Evaluating the efficacy of closed-loop glucose regulation via control-variability grid analysis. J Diabetes Sci Technol. 2008;2(4):630-635. doi:10.1177/193229680800200414

109.

Guerra

Sparacino

Facchinetti

Schiavon

Man

Cobelli

A dynamic risk measure from continuous glucose monitoring data. Diabetes Technol Ther. 2011;13(8):843-852. doi:10.1089/dia.2011.0006

110.

Nathan

Kuenen

Borg

Zheng

Schoenfeld

Heine

; A1c-Derived Average Glucose Study Group. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31(8):1473-1478. doi:10.2337/dc08-0545

111.

Peyser

Balo

Buckingham

Hirsch

Garcia

Glycemic variability percentage: a novel method for assessing glycemic variability from continuous glucose monitor data. Diabetes Technol Ther. 2018;20(1):6-16. doi:10.1089/dia.2017.0187

112.

Ryan

Shandro

Green

, et al. Assessment of the severity of hypoglycemia and glycemic lability in type 1 diabetic subjects undergoing islet transplantation. Diabetes. 2004;53(4):955-962.

113.

Hermanides

Vriesendorp

Bosman

Zandstra

Hoekstra

Devries

JH.

Glucose variability is associated with intensive care unit mortality. Crit Care Med. 2010;38(3):838-842. doi:10.1097/CCM.0B013E3181CC4BE9

114.

Service

FJ.

Glucose variability. Diabetes. 2013;62(5):1398-1404. doi:10.2337/db12-1396

115.

Schlichtkrull

Munck

Jersild

The M-value, an index of blood-sugar control in diabetics. Acta Med Scand. 1965;177:95-102.

116.

Hirsch

Balo

Sayer

Garcia

Buckingham

Peyser

TA.

A simple composite metric for the assessment of glycemic status from continuous glucose monitoring data: implications for clinical practice and the artificial pancreas. Diabetes Technol Ther. 2017;19(S3):S38-S48. doi:10.1089/dia.2017.0080

117.

Batista

GEAPA

Keogh

Tataw

de Souza

VMA.

CID: an efficient complexity-invariant distance for time series. Data Min Knowl Disc. 2014;28:634-669. doi:10.1007/s10618-013-0312-3

118.

Kovachev

Gonder-Frederick

Cox

Clarke

WL.

Evaluating the accuracy of continuous mathematical model. Diabetes Care. 2004;27(8):1922-1928. doi:10.2337/diacare.27.8.1922

119.

Clarke

Cox

Gonder-frederick

Carter

Pohl

SL.

Evaluating clinical accuracy of systems for self-monitoring of blood glucose. Diabetes Care. 1987;10(5):622-628. doi:10.2337/diacare.10.5.622

120.

Howorka

Pumprla

Gabriel

Thoma

Schabmann

Computerized generation of circadian sensor modal days with continuous glucose monitoring for comparison of various insulin regimens based on insulin glargine in type 1 diabetes. Int J Artif Organs. 2003;26(8):728-734. doi:10.1177/039139880302600805

121.

Miller

Strange

Use of fourier models for analysis and interpretation of continuous glucose monitoring glucose profiles. J Diabetes Sci Technol. 2007;1(5):630-638. doi:10.1177/1932296807001005

122.

Biagi

Bertachi

Giménez

, et al. Individual categorisation of glucose profiles using compositional data analysis. Stat Methods Med Res. 2019;28(12):3550-3567. doi:10.1177/0962280218808819

123.

Herrero

Reddy

Georgiou

Oliver

NS.

Identifying continuous glucose monitoring data using machine learning. Diabetes Technol Ther. 2022;24(6):403-408. doi:10.1089/dia.2021.0498

124.

Thomas

Shin

Jiang

McMahon

Kolassa

Vigersky

RA.

The development of new composite metrics for the comprehensive analytic and visual assessment of hypoglycemia using the hypo-triad. J Diabetes Sci Technol. 2018;12(1):69-75. doi:10.1177/1932296817721242

125.

Vigersky

Shin

Jiang

Siegmund

McMahon

Thomas

The comprehensive glucose pentagon: a glucose-centric composite metric for assessing glycemic control in persons with diabetes. J Diabetes Sci Technol. 2018;12(1):114-123. doi:10.1177/1932296817718561

126.

Zou

Wang

Zheng

Hou

Glycemic deviation index: a novel method of integrating glycemic numerical value and variability. BMC Endocr Disord. 2021;21(1):52. doi:10.1186/s12902-021-00691-z

127.

Henriques

Munshi

Segal

Costa

Goldberger

AL.

“Glucose-at-a-glance”: new method to visualize the dynamics of continuous glucose monitoring data. J Diabetes Sci Technol. 2014;8(2):299-306. doi:10.1177/1932296814524095

128.

Lai

Zhang

, et al. Investigation of glucose fluctuations by approaches of multi-scale analysis. Med Biol Eng Comput. 2018;56(3):505-514. doi:10.1007/s11517-017-1692-0

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB