DISTq: An Iterative Analysis of Glucose Data for Low-Cost, Real-Time and Accurate Estimation of Insulin Sensitivity

Table 1: Overview of the Most Frequently Used Insulin Sensitivity Tests

	Accuracy		Time Demand (Hours)		Notes	References
	Intra- Subject CV	Correlation to Gold Standard	Clinician	Participant	Notes	References
EIC	4 to 10%	Gold standard	10	4	Results is not guaranteed with an inexperienced clinician	ADDIN EN.CITE ADDIN EN.CITE.DATA [⁵, ⁶, ¹¹]
IVGTT	21%	0.44 to 0.89	3-5	3-5	Result generally requires a complex computerised solver	ADDIN EN.CITE ADDIN EN.CITE.DATA [⁸, ¹², ¹³]
OGTT	7 to 15%	0.6 to 0.8	3-4	3-4	Rate of appearance of glucose in plasma is relatively unknown	ADDIN EN.CITE ADDIN EN.CITE.DATA [⁵, ¹⁴, ¹⁵]
2hr-OGTT	15 to 40%	0.45 to 0.74	0.5	2.5	Poor repeatability in terms of re-classification rate	ADDIN EN.CITE ADDIN EN.CITE.DATA [⁷, ¹⁴, ¹⁶]
HOMA	10 to 40%	-0.6 to -0.9	0.25	0.25	Low resolution in region of clinical significance	ADDIN EN.CITE ADDIN EN.CITE.DATA [⁵, ¹⁷, ¹⁸]

Shown are the assessed accuracies in coefficient of variation (CV=SD/mean) and correlation to the gold standard EIC; and the time demand in hours for the clinician and the participant.