To our knowledge, this study is the first prospective study to examine the use of real time CGMS in patients with GDM. We found that a single application of RT-CGMS for pregnant women shortly after GDM diagnosis is useful as an educational and motivational tool. It gives the pregnant woman insights on the effect of food, exercise, and insulin on blood glucose, which helps in modifying the patient’s diet and exercise practices. In line with this finding, previous studies have shown the usefulness of RT-CGMS as an educational and a motivational tool for poorly controlled type 1 [3, 4] and type 2 diabetes..
Blood glucose variability has a significant impact on the quality of life and appears to be associated with the development of diabetes complications . A reduction in glycemic variability alone was suggested to improve diabetes outcomes even with no improvement in HbA1c . In the present study, there was significant improvement in blood glucose variability with a single application of RT-CGMS to GDM patients. This is in line with previous studies that used RT-CGMS in type 1 and type 2 diabetic patients [4–12] and other studies that used retrospective CGMS in patients with GDM . In the current study, mean glycaemia, SD of mean glycaemia, and the area under the curve for both hyper and hypoglycaemia were used to assess the glycemic variability, and we found significant improvement in the mean glycaemia and the SD of mean glycaemia by the last day of RT-CGMS application. The area under the curve for both hyper and hypoglycaemia were also improved; however, the results were not statistically significant.
Despite the improvements in glycemic variability during RT-CGMS application, there were no differences in glycemic control between the RT-CGMS and the standard antenatal care group at the end of pregnancy. This is in agreement with a study by Yu et al. in which the researchers found significant improvements in glycemic variability; however, the mean blood glucose levels were similar between the CGM and the routine care group and the researchers study did not evaluate HbA1c at the end of the pregnancy . In addition Secher et al.  did not find improvement in glycemic control when applying RT-CGMS intermittently to pregnant women with well-controlled type 1 and type 2 diabetes. This is also in line with a study on well-controlled non pregnant diabetic patients in whom there was no additional improvement in HbA1c with RT-CGMS application .
On the other hand, some studies have demonstrated the effectiveness of RT-CGMS on glycemic control in non-pregnant patients with type 1 or type 2 diabetes [4–12] as well as the uses of intermittent retrospective CGM in patients with pregestational diabetes . In those studies, participants had higher HbA1c at baseline compared to the participants in our study and other studies with negative results. This might explains the disagreement between the findings. As the level of HbA1c tends to decrease in pregnancy due to the rise in red cell mass and red blood cell turnover, the use of HbA1c to assess glycemic control in GDM women with low HbA1c levels at the initial visit may not be useful . The mean baseline HbA1c in our cohort of women was 40 mmol/mol [range 31–49], (5.8 %) [range 5–6.7], which might explain our negative results on HbA1c.
Our finding of the lack of effectiveness of RT-CGMS on pregnancy outcomes was in agreement with a study by Secher et al. in which RT-CGMS was used in pregnant women with well-controlled type 1 and type 2 diabetes , and is in contrast to the findings from the other studies that confirmed the improvement in pregnancy outcomes when using retrospective CGM in patients with pregestational diabetes  or GDM . Murphy et al.  found lower birth weight and lower risk of macrosomia when using retrospective CGM intermittently in patients with pregestational diabetes, and Yu et al.  found less risk of preeclampsia and caesarean deliveries, lower birth weight, and less neonatal complications when using retrospective CGM intermittently on GDM patients.
The reason for the disagreement between the studies on the effect of CGMS on glycemic control and pregnancy outcomes can be attributed to several factors, including baseline glycemic control, size of the study sample, duration of CGM application, and patient selection. Well-controlled diabetic patients at baseline may have no further benefits from using CGM. The smaller the study sample size the lower the power for finding significant results. Indeed, we found lower levels of HbA1c and mean fasting and postprandial glucose along with lower doses of insulin in the RT-CGMS group than in the standard group; however, the results did not reach statistical significance. In addition, longer application of RT-CGMS may give better results than shorter usage, as there is a learning curve for women using the system. Finally, patient selection is an extremely important factor. Without a doubt, the key issue for RT-CGMS success is selecting appropriate patients. Highly motivated patients who have an interest in using this technology and are enthusiastic to react accordingly will benefit most from using the system. If the patient is reluctant to respond to the data delivered by RT-CGMS, the system will be useless. This was illustrated by a study that applied RT-CGMS for one month to T1DM adolescent patients. Although they found significant improvement in glycemic control, the use of this technology was not efficient for those with a very high HbA1c [HbA1c >86 mmol/mol (10 %)] at baseline, in whom compliance and self-motivation are the main concerns . Therefore, we still believe in the favourable effects of RT-CGMS in patients with diabetes if applied to the right person.
RT-CGMS was accepted by most of the participants, and the reasons for non- acceptance were technical challenges, such as calibration and frustration with sensor alarms, anxiety from continuous awareness of blood glucose level, skin reaction, and disparities between RT-CGMS and SMBG readings. Admittedly, a significant number of our women had readings in the hypoglycaemic range during the first day of RT-CGMS application, which conflicted with the SMBG reading, a finding that has been reported previously by Secher et al. . This indicates that there is still a need for improvement in the RT-CGMS accuracy in the hypoglycaemic range.
Limitations to the current study include single use of RT-CGMS, which might be the reason for non-improvement in glycemic control or pregnancy outcomes. However, our aim was to test the efficacy of a single application of RT-CGMS in GDM because it is more convenient and acceptable to the patients and less costly. Another limitation of our study is the relatively small sample size, which may have limited the power to detect differences in glycemic control or in pregnancy outcomes. A larger clinical trial with a selection of highly motivated patients is recommended to evaluate the effectiveness of RT-CGMS in pregnant women with GDM. In addition, cost effectiveness studies assessing the application of RT-CGMS in these women versus standard care is also needed.