Thank you for giving me the opportunity to review the manuscript »The Effect of Do-It-Yourself Real-Time Continuous Glucose Monitoring on Psychological and Glycemic Variables in Children with Type 1 Diabetes: A Randomized Crossover Trial”.
Elbalshy et al conducted a crossover RTC including 55 pairs of children with t1d (age 2-13 years) and their parents using either isCGM or DIY isCGM transformed to rtCGM using a third-party device). Their primary aim was to evaluate the fear of hypoglycaemia (parental) and secondarily to evaluate glycaemic control (CGM-based metrics) and other psychosocial measures.

Their main finding was that while there was no difference in fear of hypoglycaemia, glycaemic control (Time in range and time above range) improved with DIY-rtCGM use, in addition to improved parental treatment satisfaction. Although there was no difference in the primary outcome, the results of this study are novel and clinically meaningful.
The manuscript is clearly written, study design is robust, and their conclusions are scientifically sound.

1. I agree with the authors that there was no difference in the primary endpoint due to low baseline FOH and low rates of TBR throughout the study (approx. 1%) – suggesting that hypo-avoidance strategy was predominant. It would be important to emphasize that hyperglycemia and glucose variability are very damaging in this young age group and (eg. doi: 10.2337/dc20-2125, DOI: 10.2337/dc20-0171) and implementing time in range targets would be beneficial (doi: 10.2337/dci19-0028). Please amend intruduction section accordingly.

2. Consequently, the improvement in time above range – especially above 13,9 mmol/l for 6.7% - is both highly significant and relevant. This could be discussed more in the manuscript.

3. did you collect data regarding body weight, height and BMI (SDS) also at the end of the study? It would be interesting to see if hypo-avoidance led to weight gain.

4. Please report other standardised glucose metrics in the study results (doi: 10.2337/dc17-1600), including data on glucose variability (both CV and SD of mean glucose) and AGP (maybe as a Supplemental Figure).

5. Did you analyze separately per insulin delivery modality (maybe as a sensitivity analysis)?

6. Did you adjust your statistical analysis for multiple comparisons?

7. Was there normal distribution for baseline demographics? Otherwise, reporting as median (IQR) would be more appropriate.

8. Was there normal distribution for TIR and TAR (If I understand correctly, TBRs are reported as median due to not normally distributed, correct)? Please clarify.

9. Table 3 – please unify data reporting (% placement).

10. In the discussion, I suggest replacing older studies (eg. JDRF) with newer (eg. doi: 10.2337/dc20-1060, doi: 10.1111/pedi.12729, doi: 10.1016/j.diabres.2018.10.001)

11. I suggest adding available (glucose control) data from participants who dropped out to supplemental material. Is there any explanation why did they leave the study (to my understanding when switching to intervention arm)? This might extremely clinically relevant.

Thank you for your time and effort preparing this revision - greatly improved.
If editors agree, I suggest that information regarding the other standardised glucose metrics in the study results (doi: 10.2337/dc17-1600), including data on glucose variability (both CV and SD of mean glucose) and AGP (maybe as a Supplemental Figure) are added to this manuscript.