Table 2

Key intervention studies in hypoglycemia unaware patients with T1D

BaselineEnd of study
ReferenceIntervention and duration of follow-upStudy design and durationNo. and duration of diabetes (years)HbA1c (%)SH rate and awareness scoreHbA1c (%)SH rate and awareness scoreComments
Fanelli et al. (42)Change from twice daily to MDI regimen Daily contact Re-educationProspective, single center Before and after 1-year follow-upn = 21 (11.4 ± 1.8)5.8 ± 0.1 (normal 3.8–5.5)9 of 21 had SH in previous year All had IAH6.9 ± 0.1 (normal 3.8–5.5)No SH in 12-month follow-up 10-fold reduction in mild hypoglycemic eventsRestoration of symptom and counterregulatory hormone responses; sustained to 1 year
Cranston et al. (40)Intensive insulin therapy Frequent contact Intensive educationProspective, single center Before and after 4-month follow-upn = 12 men (11–32)Group A: 6.9 ± 0.3 Group B: 8.7 ± 0.3N/A At least three CBG <3 with no symptoms in 2 weeksNo significant deterioration (numbers not reported)SH: N/A Restoration of symptoms and improvement in counterregulatory responses after 3 weeks of no CBG <3.0 mmol/LMean 4.1 ± 1.1 months of frequent contact required to achieve 3 weeks of hypoglycemia avoidance
Dagogo-Jack et al. (41)Scrupulous avoidance of iatrogenic hypoglycemiaProspective, single center Before and after 3-month follow-upn = 6 (15.5 ± 4.4) Age <358.3 ± 1.010.6 ± 3.7 All IAH9.8 ± 1.1SH not reported 60% reduction in CBG <3.3 mmol/LRestored symptom but not hormonal responses at 3 months
Hermanns et al. (44)HyPOS vs. control intervention, with most patients also moving to CSII with more than one SH or IAH with >10 years of T1DRCT 6 months (31-month follow-up reported separately)n = 164, 85.3% with T1D (HyPOS: 20.2 ± 10.8 Control: 22.1 ± 10.9)HyPOS: 7.2 ± 0.9 Control: 7.4 ± 1.1HyPOS: 3.5 ± 3.6/patient-year Clarke score 4.8 Control: 3.6 ± 36/patient-year Clarke score 5.0HyPOS: 7.1 ± 0.9 Control: 7.3 ± 1.1HyPOS: 0.9 ± 1.9/patient-year Clarke score 2.3 Control: 0.6 ± 1.2/patient-year Clarke score 3.030% reduction in mild hypoglycemia with HyPOS; further 31-month follow-up shows sustained benefit in HyPOS over standard education
Cox et al. (43)HAAT vs. control Increased SMBG and monthly visits No previous education or SMBGRCT 18-month follow-upn = 60 (13)HAATT: 8.1 ± 0.7 Control: 8.0 ± 0.7HAATT: 2.0 Control: 1.8HAATT: 8.0 Control: 8.1HAATT: 0.4 Control: 1.7Hypoglycemia-specific education program performed better than frequent contact
Little et al. (48)HypoCOMPaSS MDI vs. CSII and SMBG vs. CGM in 2 × 2 design Both arms: education and monthly visits plus telephone supportRCT 24-month follow-upn = 96 (28–30)8.2 ± 1.28.9 ± 13.4 77% had SH in prior year Gold score 5.1 ± 1.18.1 ± 1.20.8 ± 1.9 20% had SH in prior year Gold score 4.1 ± 1.6No differences among CSII, CGM, and SAP CGM had slightly greater reduction in mild and SH, but higher baseline than SMBG
Pedersen-Bjergaard et al. (56)HypoAna study [Analog insulin (detemir and aspart) vs. human insulin (NPH and regular)]Multicenter, open-label, crossover RCT (1:1) 2-year follow-upn = 1598.0 ± 1.0All had 2 or more episodes of SH in prior yearDecrease of 0.13% with analog treatment compared with human insulin (P = 0.02)Insulin analogs: 105 episodes of SH Human insulin: 136 episodes of SHAbsolute rate reduction of 0.51 episodes of SH per patient-year with insulin analogs
Ly et al. (65)CSII vs. SAP in young patients with IAHRCT 6-month follow-upn = 95 (11.0)7.4 ± 0.2SAP-LGS: 6 SH and 0.22/patient-year Clarke score 5.9 ± 1.5 CSII: 6 SH and 0.25/patient-year Clarke score 6.4 ± 1.57.4 ± 0.2 in both armsSAP-LGS: 0 events = 0/patient-year Clarke score 4.7 (4.0–5.1) CSII: 6 events = 0.26/patient-year Clarke score 5.1 (5.5–6.4)Young patients with short duration SAP-LGS reduced SH compared with CSII alone, although event rates were low Reduced time <70 and <60 mg/dL in LGS arm
de Zoysa et al. (47)DAFNE-HART 6 sessions based on cognitive behavioral therapy and motivational interviewing 1-year follow-upSmall trial; 2 sites Uncontrolled pilot clinical trialn = 24 (31 ± 12) Previous DAFNE 8 CSII7.8 ± 1.1Median 3.0 (0–103) Clarke score 5.4 ± 1.2 Ryan score 9487.8 ± 1.2Median 0 (0–3) 70% reported no further SH Clarke score 3.8 ± 1.8 Ryan score 3729 of 20 restored awareness Some accepted CSII or CGM, which they were resistant to previously
Choudhary et al. (66)Observational clinical audit of CGM in patients with SH despite previous DAFNE and CSII 1-year follow-upRetrospective auditn = 35 (29.6 ± 13.6)8.1 ± 1.2SH 8.1 ± 13.6/patient-year Gold score 5.1 ± 1.57.6 ± 1.00.6 ± 1.2/patient-year Gold score 5.2 ± 1.9 54% reported subjective improvement in awarenessReduction in SH with CGM 23 of 35 used SAP-LGS 3 with SH on SAP had no further episodes on SAP-LGS
Giménez et al. (58)CSII and education program with 2–3- month follow-upSmall, uncontrolled pilot trial 2-year follow-upn = 20 (16.2)6.7 ± 1.11.25 ± 0.4/patient-year Clarke score 5.5 ± 1.26.3 ± 0.9SH 0.05/patient-year at 2 years Clarke score 1.6 ± 2.0316 of 19 restored awareness Reduced mild hypoglycemia Time <70 mg/dL reduced from 13.7 to 9.1%
Sämann et al. (35)Structured education 5-day course (DTTP) 9,583 patients at 96 centers 1-year follow-upSubgroup analysis of patients with 3 or more SH at baselinen = 341 (18.7 ± 11.1)7.4 ± 1.96.1 ± 9.6/patient-year7.2 ± 1.51.4 ± 5.4/patient-year56% had 1, 20% had 2, and only 15% had more than 3 SH events Reduced time in hospital 8.6 ± 15.4 to 3.9 ± 10.7 days/patient-year
Hopkins et al. (16)Structured education program in flexible insulin therapy (U.K. DAFNE) 5-day course 1-year follow-up IAH subgroup reportedRetrospective auditn = 215 (Duration not reported)N/A3.6 ± 13.6/patient-yearN/A1.3 ± 5.9/patient-yearImproved awareness of hypoglycemia at a blood glucose >3 mmol/L
O’Connell et al. (92)Islet treatment (961 ± 445 kIEQ) Antithymocyte globulin followed by tacrolimus + mycophenolate mofetil 2 patients had 1 treatment, 7 had 2 treatments, and 8 had 3 islet grafts3-center observational study 1-year follow-upn = 17 Duration not reported8.3 ± 2.0Hyposcore 2,976 ± 3,4946.5 ± 1.32 of 17 patients with graft failure had SH82% achieved composite end point of HbA1c <7% and absence of SH at 1 year
Brooks et al. (93)Islet treatment (535 kIEQ)Observational multicenter trial 2-year follow-upn = 20 30 (16.5–38.5)8.0 (7.0–9.6)20 (7–50)6.2 (5.7–8.4)0.3 (0–1.6)70% maintained composite end point of HbA1c <7% and no SH at 2 years
  • CBG, capillary blood glucose; DTTP, diabetes teaching and treatment program; kIEQ, kilo islet equivalent; N/A, not available.