Aim. Evaluate considerations used when thresholds for diagnosis of gestational diabetes (GDM) were defined and the possible impact on rates of diagnosis of GDM in Australia.
Methods. A systematic search using a combination of synonyms for guidelines and GDM was performed in PubMed and National Guideline Clearinghouse to identify relevant guidelines published between January 1st, 2005 and January 1st, 2018. We evaluated whether (1) a clear description of the new definition (2) estimates on increased prevalence of disease, (3) triggers for modification of disease definition, (4) prognostic ability of the new definition to predict clinically important outcomes, (5&6) potential harms and benefits in the newly diagnosed, and (7) the balance of harms and benefits was included or considered. Rates of GDM were obtained from the NDSS Diabetes in Pregnancy Program and publications.
Results. Definitions used different conceptual bases: (i) a percentile (or Gaussian) definition, (ii) a risk-based assessment of different maternal and fetal outcomes, and (iii) harmonisation with type 2 diabetes in non-pregnant adults.A few influential studies were repeatedly cited, and no systematic reviews. The definition process did not include explicit consideration or quantification of benefits versus harms, and appeared to choose somewhat arbitrary cut points, such as the IADPSG consensus choice.
Rates of GDM in Australia appear to have increased from less than 2% in 1990 to 13.7% in 2017. Possible causes are: Aging of the pregnant population, increasing weight, more screening, better recording of cases, changes in the rates of women from high risk ethnic groups and definition change. The sharpest increase was seen after adoption of changes to the ADIPS guidelines in 2013
Discussion.Our analysis of the links and changes in definitions of gestational diabetes reveals a complex history. Differences in definitions may have led to substantial differences in the apparent prevalence of GDM