Carbohydrate counting implementation on pediatric type 1 diabetes mellitus: systematic review and meta-analysis.

Purpose: Type 1 diabetes mellitus, which is the most common type of diabetes among children, is not curable but can be managed well without a negative effect on quality of life. One of the treatments of type 1 diabetes mellitus is carbohydrate counting. This systematic review and meta-analysis sought to evaluate the efficacy of carbohydrate counting with regard to hemoglobin A1c (HbA1c) reduction in children with type 1 diabetes mellitus.

Methods: Nine studies were assessed, with the primary outcome being glycemic control (HbA1c changes). We searched the following electronic databases: ProQuest, PubMed, Scopus, and ScienceDirect. The quality of studies included was assessed using the risk of bias for randomized control trials and the JBI Critical Appraisal Checklist for observational and cross-sectional studies. Quantitative analyses were made and extrapolated into a forest plot.

Results: A total of 1,693 articles were identified. Four reviewers independently screened titles and abstracts. Of the 36 articles screened, 34 articles were found to be eligible. Of these, 25 studies were excluded because of unsuitable outcomes and study designs. Nine articles were included in the final analysis. Meta-analysis showed that there was a reduction in HbA1c in the carbohydrate counting group as compared to the control group. The cumulative effect of carbohydrate counting on HbA1c was a mean difference of -0.55 (95% confidence interval, -0.81 to -0.28, P<0.001). All of the studies exhibited similar results with the mean difference reduction favoring the interventional group. However, the heterogeneity analysis revealed an I2 value of 88%, implying high heterogeneity in the meta-analysis.

Conclusion: The meta-analysis showed evidence favoring the use of carbohydrate counting in the management of type 1 diabetes mellitus.