Glucose Monitoring Devices: Measuring Blood Glucose to Manage and Control Diabetes

Glucose Monitoring Devices: Measuring Blood Glucose to Manage and Control Diabetes
Copyright
Contributors
About the Authors
1. Introduction to SMBG
	Historical perspective and principles of blood glucose control
	The evidence base for SMBG in type 1 diabetes
	The evidence base for SMBG in type 2 diabetes
	Guidelines for SMBG
	The shortcomings of SMBG and future perspective
	References
2. Analytical performance of SMBG systems
	Introduction
	The process for premarket approval of SMBG devices
		Analytical performance according to ISO 15197
			Precision
			Interference evaluation
			Accuracy
		Analytical performance according to FDA
	Postmarket analytical performance
	Advances in analytical performance of SMBG devices
	Conclusion
	List of authors
	References
3. Clinical evaluation of SMBG systems
	References
4. Consequences of SMBG systems inaccuracy
	Introduction
	Quantifying the effect of inaccurate BGM systems
		A complex system
		Patient behavior is the main driver
		Effects spread over time and space
			Clinical outcomes
				Short term
				Short term
				Long term
				Long term
			Quality of life
			Financial outcomes
				Short term
				Short term
				Long term
				Long term
		Requirements
			Physiological variability
			Behavioral variability
			Device and lot variability
			Therapy modes
			Time span
	Accuracy and its consequences
		Modeling and simulation
			Metabolic models and simulators (metabolic variability)
			Behavioral modeling and simulation (behavioral variability and therapy modes)
			Integrated metabolic/behavioral simulation
			Modeling glucose monitoring devices (device and lot variability)
			In silico accuracy studies
		From in silico results to short-term clinical outcomes
		Long-term health and complications
			From in silico results to long-term complications
			From clinical performance to financial outcomes
	An extended illustration
		The in silico study
			Meter models
			Behavioral models
		Clinical and financial outcomes
		Results: clinical outcomes
		Accuracy and clinical outcomes: a regression model
		Determining financial impact
			Cost compared to an ideal
			Worst-case costing
		Observations from the illustrative example
			Limitations
	Conclusions and future work
	References
5. Modeling the SMBG measurement error
	SMBG measurement error
	Why modeling the SMBG measurement error?
	Literature models of SMBG measurement error
	The state-of-the-art modeling method by Vettoretti et al.
		Definition of training and test sets
		Constant-SD zones identification
		Maximum-likelihood fitting
		Model validation
	Derivation of a model of SMBG error distribution for two commercial devices
		Case study 1: modeling the One Touch Ultra 2 measurement error
			Dataset
			Preprocessing of YSI and SMBG-YSI matching
			Model development
			Model validation
		Case study 2: modeling the Bayer Contour Next measurement error
			Dataset
			Preprocessing of YSI and SMBG-YSI matching
			Model development
			Model validation
		Remark
	Applications of the SMBG measurement error models
	Conclusion
	References
6. CGM sensor technology
	Introduction
	Glucose transduction technologies
		Current technologies
			Transduction technologies used in commercially approved CGMs
				Enzymatic, electrochemical-based sensors
				Enzymatic, electrochemical-based sensors
				Nonenzymatic, optical-based sensors
				Nonenzymatic, optical-based sensors
		Transduction technologies in development
		Tissue interface for transcutaneous and subcutaneous transduction
		Noninvasive technologies
	Sensor interface and system connectivity
		Sensor front end electronics
		Transmitter software and sensor calibration
		Skin interface for the wearable transmitter
	System user interface and connectivity
		Connected systems
		Artificial pancreas
		Connected pens
	Commercial systems
		Overview
			Abbott Freestyle Libre flash glucose monitor
			Dexcom G6 CGM
			Medtronic Guardian Connect CGM
			Eversense CGM system
		Summary
	References
7. Clinical impact of CGM use
	Introduction
		History and general rationale for glucose monitoring
	Parameters of glucose control and risk association
		HbA1c
		Hypoglycemia
		Time in range and glucose variability
	Glucose monitoring
		Limitations of SMBG
		Benefits of CGM
			Operational advantage
			Direction, pattern and trends, investigative tool
			Alerts
	Clinical application of CGM
	CGM efficacy
		Retrospective CGM studies
			Observational trials
			Randomized controlled trials
		Real-time CGM studies
			Observational
			Randomized controlled trials
		CGM pregnancy data
		CGM quality of life data
	CGM limitations
		User dependent
		Healthcare provider dependent
		Device dependent
	Available CGM systems
		Dexcom
		Medtronic
		Medtrum
		Senseonics
	Flash glucose monitoring
	Further utility of CGM
		Combining technology
		Inpatient CGM
	Summary
	References
8. Accuracy of CGM systems
	Introduction
	Clinical accuracy
	Numerical (statistical) accuracy
	Conclusions
	References
9. Calibration of CGM systems
	Calibration of minimally invasive CGM sensors
		Problem statement
		Critical aspects affecting calibration
	State-of-art calibration algorithms and today's challenges
		Simple heuristic to deal with the BG-IG system
		Kalman filter-based approaches
		Methods relying on autoregressive models
		A calibration method integrating several local dynamics models
		Two approaches to optimize the computational complexity
		Deconvolution-based Bayesian approach
		Recursive approaches exploiting past CGM data
		Today's challenges for CGM calibration algorithms
	The Bayesian approach applied to the calibration problem
		Description of a Bayesian calibration algorithm
			Estimation of model parameters
				Step 0: parameter initialization
				Step 0: parameter initialization
				Step 1: use of calibration model
				Step 1: use of calibration model
				Step 2: compensation of BG-to-IG kinetics
				Step 2: compensation of BG-to-IG kinetics
				Step 3: match between estimated BG and available SMBG
				Step 3: match between estimated BG and available SMBG
				Step 4: parameter update
				Step 4: parameter update
			Calibration of the current signal
		Example of implementation
			Dataset description
			Implementation
				Prior derivation
				Prior derivation
				Calibration scenarios
				Calibration scenarios
				Performance assessment
				Performance assessment
			Results
	Conclusions
	References
10. CGM filtering and denoising techniques
	Introduction
	The denoising problem
	Possible approaches to CGM denoising
	CGM denoising by Kalman filter
		Overview of the Kalman filter
		Formulation as online self-tunable approach
			A priori model for u(t)
			Determination of λ2 and σ2
			Step 1
			Step 2
		In silico assessment
			Accuracy in SNR determination
				Importance of filter parameters accuracy
				Importance of filter parameters accuracy
				Comparison with MA
				Comparison with MA
		Assessment on data
		Dealing with SNR intraindividual variability
	Conclusions
	References
11. Retrofitting CGM traces
	Introduction
		Chapter organization
	The retrofitting algorithm
		Problem formulation
			Notation
		Algorithm description
			Step A: retrospective Bayesian CGM recalibration
			Step B: constrained regularized deconvolution
	Retrofitting outpatient study data
		Original dataset
		Outpatient-like dataset
		Accuracy outcomes metrics and statistical analysis
			Statistical analysis
		Results
	Retrofitting real-life adjunctive data
		Original datasets
		Real-life-like datasets
		Accuracy outcomes metrics and statistical analysis
		Results
	Accuracy of retrofitted CGM versus number of references available
	Conclusions
	Appendix: data preprocessing
	References
12. Modeling the CGM measurement error
	Introduction
	Methods
		Datasets
		A posteriori recalibration
		Reference-sensor density and delay estimation
		Estimation and modeling of the sensor error distribution
		Estimation and modeling of sensor error time dependency
	Results
		Effect of rate of change on sensor error and delay estimation
		Characterization of recalibrated synchronized sensor errors
		Modeling of subcutaneous sensors
	Conclusions
	References
13. Low glucose suspend systems
	Introduction
	Low glucose suspend system
	Clinical studies with LGS system
	Real-life evidence with TS system
	Cost-effectiveness
	The limitations of the low glucose suspend system
	Future direction
	References
14. Predictive low glucose suspend systems
	Introduction
	Algorithm development
	PLGS clinical studies
	Commercial devices
	Keys to clinical use
	Summary and conclusions
	References
15. Automated closed-loop insulin delivery: system components, performance, and limitations
	Introduction
	Closed-loop glycemic control algorithms
		Proportional-integral-derivative control
		Fuzzy logic control
		Model predictive control
		Zone model predictive control
		Adaptive control
			Recursive modeling
			Subspace-based state-space system identification
				Recursive system identification
				Recursive system identification
			Adaptive generalized predictive control
			Run-to-run control
			Iterative learning control
			Adaptive weights through glycemic risk index
	Quantifying plasma insulin concentrations
		Modulating insulin infusion
	Closed-loop glycemic control results
	Future directions
	Conclusions
	References
16. The dawn of automated insulin delivery: from promise to product
	Introduction
	Continuous subcutaneous insulin infusion therapy: the first building block in developing a closed-loop system
	Continuous glucose monitors: the second step in the construction of a closed-loop system
	The way forward: the JDRF roadmap to an artificial pancreas
	Making the dive less deep and shorter: low glucose suspend systems
	Stopping the plunge: suspend before low systems
	The algorithms: the final piece of the puzzle
	Speeding up the process: the creation of an FDA approved simulator
	Early studies aimed at closing the loop
	Control in clinic: the first closed-loop studies in rigorous research environments
	From transitional environments to tests at home
	From prototype to product: the MiniMed 670G system
	Exploring the equipment: components and characteristics of the 670G
		Guardian sensor 3
		The Medtronic 670G insulin pump
		Auto mode
		Auto mode exits: what are they and why do they happen
		The highs and lows of real-world use of the 670G system
			Challenges that remain
			A bright future
		Patient considerations
	Conclusion
	Short biography
	References
Index
	A
	B
	C
	D
	E
	F
	G
	H
	I
	J
	K
	L
	M
	N
	P
	R
	S
	T
	U
	V
	W
	Z