J. Große, H. Hornstein, U. Manuwald, J. Kugler, I. Glauche and U. Rothe (2018). Incidence of Diabetic Ketoacidosis of New-Onset Type 1 Diabetes in Children and Adolescents in Different Countries Correlates with Human Development Index (HDI): An Updated Systematic Review, Meta-Analysis, and Meta-Regression. Hormone and metabolic research = Hormon- und Stoffwechselforschung = Hormones et metabolisme, 50(3), 209-222. https://doi.org/10.1055/s-0044-102090

PubMed 29523007

[Abstract]
Type 1 diabetes mellitus (T1DM) is usually diagnosed by insulin deficiency at a young age. Diabetic ketoacidosis (DKA) represents a severe complication occurring before the first diagnosis of T1DM. Actually, the data situation is still unsettled in assessing the current state of diagnosis. This study summarizes the latest rates of DKA of new-onset T1DM in children and adolescents in different countries available over the last five years. Different T1DM-related, geographical and socioeconomic moderators are suitable to explain the heterogeneity of observed DKA rates. A systematic literature search using PubMed, EMBASE*, and MedLine* (*via OVID) was conducted to extract worldwide DKA rates covering publications from April 2011 to May 2016. We define DKA consistently by pH<7.3 or bicarbonate<15 mmol/l. We identified 34 suitable studies covering DKA rates in 25 countries. Overall DKA rates were compared to earlier studies to identify a temporal trend. We further applied a random effects meta-analysis and used meta-regression to reveal moderators of DKA rate heterogeneity. This review evaluating 34 studies includes 47 000 children and adolescents in total. DKA rates varied from 14.7% (Denmark) to 79.8% (Saudi Arabia). DKA rates are still high but a decline can also be recognized. The meta-regression shows that latitude (p<0.000) and human development index (HDI) (p<0.000) are moderators of DKA rates. The frequency of DKA rates occurrence varies widely for different countries. Both latitude and HDI partially explain the observed heterogeneity, while other moderators such as density of physicians showed no obvious correlation.