Gold and Clarke questionnaires identify different people with insulin‐treated diabetes and impaired awareness of hypoglycaemia in almost 60% of cases: A post hoc analysis from the Hypo‐METRICS Study

Aims

Gold and Clarke questionnaires are used to identify impaired awareness of hypoglycaemia (IAH) and severe hypoglycaemic event (SHE) risk in people with diabetes (pwD). We explored their overlap, including Clarke's Hypoglycaemia Awareness Status (Clarke-HAS) subfactor, and subsection differences in SHE incidence (Gold, Clarke, overlap).

Materials and Methods

This post hoc analysis of the Hypo-METRICS study recruited pwD on insulin (type 1: T1D; type 2: T2D) with ≥ 1 hypoglycaemic event in the previous 3 months. IAH was defined as Gold ≥ 4, Clarke ≥ 4, Clarke-HAS ≥ 2.

Results

Gold and Clarke were completed by 232 pwT1D and 285 pwT2D (age: 47 vs. 62 years, p < 0.001; HbA1c: 56 mmol/mol (7.3%) vs. 57 mmol/mol (7.4%), p = 0.03). IAH prevalence in T1D vs. T2D was 21% vs. 26% (p = 0.1), 14% vs. 18% (p = 0.2), and 41% vs. 48% (p = 0.2) according to Gold, Clarke and Clarke-HAS. The overlap between Gold ≥ 4, Clarke ≥ 4 was 40% (T1D: 45%; T2D: 37%); between Gold ≥ 4, Clarke-HAS ≥ 2 it was 45% (T1D: 43%; T2D: 47%). The overlap between Gold ≥ 4, Clarke ≥ 4 was 41% vs. 39% in CGM vs. non-CGM users (T1D: 50% vs. 35%; T2D: 32% vs. 40%); between Gold ≥ 4, Clarke-HAS ≥ 2 it was 39% vs. 53% (T1D: 38% vs. 54%; T2D: 40% vs. 53%). In the IAH sample, SHE rate was lower in people found with IAH by Gold vs. Clarke, 9% vs. 53%, and in those assessed by Gold-Clarke overlap vs. Clarke, 20% vs. 53% (all p < 0.001).

Conclusions

Gold and Clarke show moderate consistency (37%–54%) in identifying IAH in insulin-treated diabetes, with/without CGM use, each associated with different SHE rates, indicating the two questionnaires measure different, independent aspects of IAH.