Assessment of IAH
IAH was common. Of a total of 131 subjects, data for the Gold score was available for 120 subjects, while data for the Clarke score was available for 122 subjects. The remainder of the subjects had not completed the assessment for these scores.
Thirty-four subjects (28.3%) had a Gold score >= 4, while 29 subjects(24.6%) had a Clarke score >= 4. In the 120 subjects with both GOLD and Clarke scores available, 13.3% of subjects had IAH on both scores (Gold and Clarke scores both >= 4), while 60.8% of subjects had intact awareness of hypoglycaemia on both scores (Gold and Clarke scores both <= 3) (Table 2) . When the Gold and Clarke scores were discordant, a higher proportion of subjects were identified to have IAH by a Gold score >= 4 (15%, 18 subjects) than by a Clarke score >= 4 (10.8%, 13 subjects) (Table 2) . There was moderate correlation between the Clarke and Gold scores (r2 = 0.415).
The Clarke score was associated with an increased frequency of hypoglycaemia during the prospective 4 week period of SMBG. The Clarke score was also associated with an increased frequency of severe hypoglycaemia over the last one year, as well as an increased frequency of ever being previously hospitalized or attending emergency department for hypoglycaemia (Table 3) .
The HYPO score was computed only for the subjects who performed at least 3 blood glucose tests daily. The median HYPO score was 6 (90th percentile 80). Significantly higher HYPO scores were seen in subjects with Clarke score >= 4 compared to those with Clarke score =< 3, although no significant difference was observed based on the GOLD score (Table 3).