Five RCTs (n=131 children, range 11 to 36 per trial) were included; three were parallel group RCTs and two were cross-over studies. Only two RCTs reported adequate generation of the allocation sequence. Two trials were double blinded and one single blinded. One trial reported that an intention-to-treat analysis had been conducted. All trials provided an adequate description of drop-outs. Only one trial was judged to be at low risk of bias.
Glycated haemoglobin levels (five trials): There was no difference between children whose insulin doses were adjusted on the basis of Continuous Glucose Monitoring System (CGMS) plus self-monitoring data compared with self-monitoring data alone (p=0.87). Sensitivity analysis did not alter the findings.
Secondary outcomes: One RCT found that children using the CGMS had an increased number of insulin dose changes per patient per month compared with those using self-monitoring alone (WMD 6.3 dose changes, 95% CI 2.88 to 9.72). There were no significant differences between treatment groups for any of the other outcomes assessed: improvement in fructosamine (one trial); major hypoglycaemic episodes (no events reported in any group in any trial); minor hypoglycaemic episodes (one trial); mean daily time and daily area under the CGMS curve for glucose below 3.89 mmol/L (one trial); mean daily area over the CGMS curve for glucose above 9.99 mmol/L (one trial); and ketoacidosis (one trial).
Local adverse effects were reported in two RCTs, including redness at the application site (23% children), redness and itching (16% children), and painful redness (one child). One child withdrew after 12 hours of CGMS due to skin irritation at the sensor site.