Study/Year/Country	Sample size/Duration	Adults or Children	Diabetes Status	Treatment & Control/Primary Endpoint	Frequency of CGM use	Frequency of SMBG use	Results	Limitations/Comments
Real Time CGM - Pump and CGM Versus Pump and SMBG
RCT Raccah et al. (4) 2009 France	N=132 6 months	Adults (n=81) Children (n=51)	Type 1 diabetes HbA1c ≥8%	Pump and CGM Vs. Pump and SMBG Primary outcome to determine change in HbA1c between CGM and SMBG.	Patients agreed to wear CGM ≥70% of study period.	At least 3 readings daily	No significant difference in ΔHbA1c CGM (-0.81%) and SMBG (-0.57%) (P=0.087). No significant difference in Δ hyperglycemia CGM -0.2±0.7 vs. SMBG -0.2±0.7 or Δ hypoglycemia CGM 0.1±0.9 vs. SMBG 0.1±0.7 episodes per day between the study groups. Adverse Events - CGM 2 ketoacidosis 1 severe hypoglycemia Adverse Events - SMBG 3 ketoacidosis	All patients previously treated with MDI. 23 patients in the CGM group failed to wear the sensors at least 70% of the time. High attrition rate - a total of 20 patients abandoned the study: 14 from the CGM group and 6 from the SMBG group. It is unclear whether there were lifestyle differences/modifications between study groups.
RCT Hirsch et al. (5) 2008 United States	N=138 6 months	Adults (n=98) Children (n=40)	Type 1 diabetes HbA1c ≥7.5%	Pump and CGM Vs. Pump and SMBG Primary outcome to determine average change in HbA1c between CGM vs. SMBG.	2 sensors per week (2 3-day periods)	Not reported.	No significant difference in ΔHbA1c between study groups. (P=0.37). CGM Mean (SD) ΔHbA1c -0.56 (0.72) SMBG Mean (SD) ΔHbA1c -0.71(0.71)	The primary endpoint did not specify whether the change in HbA1c was for within or between study groups.
							Target 7% HbA1c No significant difference (P value not reported) CGM: 12(19.4%) patients SMBG: 16 (24.2%) patients	No sample size calculation or justification was reported in the study Results were not calculated using the intent to treat principle.
							Hyperglycemia Incidence No significant difference in # events between CGM and SMBG (P value not reported)	Lack of confirmation of severe hypoglycaemic episodes (based on patients workbooks).
							Hypoglycemia Incidence No significant difference between CGM and SMBG (P = 0.07).	Hypoglycemic range was measured over a 6 day period at the beginning and the end of the study.
							Severe hypoglycemia 14 events of which 11 occurred in CGM group within 8 patients. Comparison between the groups was statistically significant (P=0.04). 6 events deemed not related or unlikely to be related to the device (i.e., not wearing the device). In remaining 5 instances, it was determined that patients ignored alerts, injected multiple boluses of insulin without using the “Bolus Wizard”, or based treatment decisions on sensor reading only, without confirming with a blood glucose test. 17 serious adverse events. 1 person in SMBG group experienced 2 skin abcesses at the insulin infusion site and 1 person in CGM group had diabetic ketoacidosis. Remaining adverse events were severe hypoglycemic events.

CSII refers to continuous subcutaneous intensive insulin infusion; MDI, multiple daily injections; SMBG, self-monitoring of blood glucose