Regional variations in definitions and rates of hypoglycaemia: findings from the global HAT observational study of 27 585 people with Type 1 and insulin-treated Type 2 diabetes mellitus.

AIM: To determine participant knowledge and reporting of hypoglycaemia in the non-interventional Hypoglycaemia Assessment Tool (HAT) study. METHODS: HAT was conducted in 24 countries over a 6-month retrospective/4-week prospective period in 27 585 adults with Type 1 or insulin-treated Type 2 diabetes mellitus. Participants recorded whether hypoglycaemia was based on blood glucose levels, symptoms or both. RESULTS: Hypoglycaemia rates were consistently higher in the prospective compared with the retrospective period. Most respondents (96.8% Type 1 diabetes; 85.6% Type 2 diabetes) knew the American Diabetes Association/European Association for the Study of Diabetes hypoglycaemia definition, but there were regional differences in the use of blood glucose measurements and/or symptoms to define events. Confirmed symptomatic hypoglycaemia rates were highest in Northern Europe/Canada for Type 1 diabetes (63.9 events/year) and in Eastern Europe for Type 2 diabetes (19.4 events/year), and lowest in South East Asia (Type 1 diabetes: 6.0 events/year; Type 2 diabetes: 3.2 events/year). Unconfirmed symptomatic hypoglycaemia rates were highest in Eastern Europe for Type 1 diabetes (5.6 events/year) and South East Asia for Type 2 diabetes (4.7 events/year), and lowest for both in Russia (Type 1 diabetes: 2.1 events/year; Type 2 diabetes: 0.4 events/year). Participants in Latin America reported the highest rates of severe hypoglycaemia (Type 1 diabetes: 10.8 events/year; Type 2 diabetes 3.7 events/year) and severe hypoglycaemia requiring hospitalization (Type 1 diabetes: 0.56 events/year; Type 2 diabetes: 0.44 events/year). The lowest rates of severe hypoglycaemia were reported in South East Asia (Type 1 diabetes: 2.0 events/year) and Northern Europe/Canada (Type 2 diabetes: 1.3 events/year), and the lowest rates of severe hypoglycaemia requiring hospitalization were in Russia (Type 1 diabetes: 0.15 events/year; Type 2 diabetes: 0.09 events/year). The blood glucose cut-off used to define hypoglycaemia varied between regions (Type 1 diabetes: 3.1-3.6 mmol/l; Type 2 diabetes: 3.5-3.8 mmol/l). CONCLUSIONS: Under-reporting of hypoglycaemia rates in retrospective recall and regional variations in participant definitions of hypoglycaemia may contribute to the global differences in reported rates. Discrepancies between participant definitions and guidelines may highlight a need to redefine hypoglycaemia criteria. (Clinical Trials Registry No: NCT01696266).