A Psychoeducational Program to Restore Hypoglycemia Awareness: The DAFNE-HART Pilot Study

Hypoglycemia and fear of hypoglycemia remain major barriers to achieving optimal glucose control and quality of life for people with type 1 diabetes. Structured education in flexible insulin therapy (e.g., the U.K.’s Dose Adjustment for Normal Eating [DAFNE]) and/or use of insulin pump therapy reduces severe hypoglycemia (SH) (1), but some continue to experience impaired awareness of hypoglycemia (IAH) with high rates of SH, their problematic hypoglycemia resistant to intervention. We hypothesized that many such people have motivational and cognitive barriers to hypoglycemia avoidance and resolution of IAH. We designed and piloted an intervention using motivational interviewing and cognitive behavioral theory targeting these barriers.

An intervention, teaching aids, curriculum, and manual (DAFNE-Hypoglycemia Awareness Restoration Training [DAFNE-HART]) were designed by a clinical psychologist, doctors, educators, and patient representatives. It revised relevant sections from DAFNE and interventions targeting problematic hypoglycemia (2,3). Participants were taught to look for hypoglycemia cues and consider their own causes and consequences of IAH and how to reduce hypoglycemia exposure. The educational material was presented within a motivational interviewing framework to support behavior change and minimize resistance. Cognitive behavioral techniques were used to identify and restructure unhelpful thoughts such as needing to “soldier on” through episodes, underestimating the consequences of hypoglycemia, and worrying excessively about intermittent hyperglycemia. In three weekly full-day group sessions, theories about hypoglycemia and awareness were reviewed, the concept of a “body scan” (a structured guide to find subjective cues to blood glucose concentration) was taught, insulin action was revised, and theories linking thoughts with behavior were explored, supporting patients to recognize and challenge their own cognitions around hypoglycemia. “Homework” used home glucose monitoring to test the learning and, during weeks 4 and 5, try newly learned skills/strategies, with scheduled individual face-to-face and telephone support. A final full-day group session focused on relapse prevention.

Twenty-four people (12 male) with type 1 diabetes, using DAFNE principles for insulin self-adjustment, with persistent IAH assessed clinically and scoring ≥4 on the Gold score, in which patients rate their awareness of hypoglycemia from 1 (“I am always aware of my hypoglycemia”) to 7 (“I am never aware of my hypoglycemia”) (4), were recruited. Hypoglycemia experience, including self-report of SH (hypoglycemia [<63 mg/dL/3.5 mmol/L] that could not be self-treated, requiring assistance), over the preceding 12 months and moderate hypoglycemia (<63 mg/dL/3.5 mmol/L, self-treated but disrupting daily activity) over the last 6 weeks, hypoglycemia awareness and burden (Gold, Clarke, and Ryan scores) (4–6), and mood and self-care behaviors around glucose control were documented by questionnaires, including the Hospital Anxiety and Depression Scale (HADS) (7), the Problem Areas In Diabetes (PAID) Questionnaire (8), the Hypoglycemia Fear Survey II (9), and the Hyperglycemia Avoidance Score (courtesy of L. Gonder-Frederick, University of Virginia, Charlottesville, VA), and glycated hemoglobin (HbA_1c) was measured. At baseline and 3 months postcourse, up to 6 days continuous glucose monitoring (CGM) (Medtronic, Inc.) was performed. Participants were reassessed 12 months postcourse.

Five DAFNE educators (specialist nurses and dietitians) were trained in the curriculum and relevant psychological skills in a 2-day workshop. They delivered four courses with weekly supervision (face-to-face, telephone, and e-mail) from a clinical psychologist.

The protocol was approved by the National Research Ethics Committee London. King’s College Hospital NHS Foundation Trust sponsored the protocol. Participants provided written informed consent. Data were compared using Student t, Wilcoxon signed rank (nonparametric variables), and χ² or related samples McNemar (categorical variables or when including episodes of no incidence) tests, using IBM SSPS version 4. Hypoglycemia in valid CGM data was defined as <3 mmol/L for ≥20 min. P < 0.05 was considered significant.

All participants completed courses. One missed follow-up and two returned only SH and HbA_1c data at 12 months. Mean (±SD) age for the 23 completers was 54.4 ± 7.9 years; diabetes duration 30.7 ± 11.9 years; time since DAFNE 7.1 ± 4.1 years. Fifteen were using twice-daily background and premeal quick-acting insulin injections, and eight were using insulin pumps. None had renal impairment, hypoadrenalism, hypothyroidism, or growth hormone deficiency. All described themselves as having “impaired awareness,” not experiencing symptoms with a blood glucose <54 mg/dL (3 mmol/L) when awake. Their SH in the year before, defined above, included 14 participants reporting unconsciousness (0–30/participant, median 1), 3 attending emergency departments (one 24 times), and 3 admissions. Nineteen people answering the question reported an additional six ambulance call-outs.

Twelve months postcourse, rates of severe and moderate hypoglycemia fell (Table 1). Seventeen participants reported no further SH (P < 0.0001). Measures of awareness improved (Table 1), with 9 of 20 regaining awareness with Gold and/or Clarke scores <4 (P < 0.04 vs. baseline for either) and 8 of 20 reporting no episodes of hypoglycemia (<54 mg/dL) without symptoms vs. none precourse (Table 1). Behavior scores around hyperglycemia avoidance improved and worry scores about hyperglycemia reduced significantly. Worries around hypoglycemia tended to reduce. PAID scores improved. HbA_1c did not deteriorate.

At 3 months postcourse, the duration of daytime hypoglycemia on CGM fell and hypoglycemia burden (Ryan score) fell from equivalent to people registering for islet transplantation toward that of routine clinic populations (6).

The main defense against hypoglycemia during insulin treatment of type 1 diabetes is subjective awareness of a falling blood glucose, impaired awareness being associated with a sixfold increase in SH (10). Awareness can be improved by hypoglycemia avoidance (e.g., 11). Retraining patients in insulin self-adjustment (1), training in hypoglycemia recognition and avoidance, as provided by Blood Glucose Awareness Training (BGAT) and Hypoglycemia Anticipation, Awareness and Treatment Training (HAATT) (2,3), and insulin pump therapy (12) all reduce SH. DAFNE restores awareness to 43% of people with IAH (1). Continuous glucose sensing reduces SH without restoring awareness (13). A group of people remain resistant to interventions to restore hypoglycemia awareness and continue to experience SH. We hypothesized that these patients have cognitive and motivational barriers to hypoglycemia avoidance, based on neuroimaging showing different activation in reward pathways during hypoglycemia in people with IAH, qualitative research describing patient beliefs about hypoglycemia awareness, and clinical audit (14–16). Although we cannot attribute outcomes to particular elements of the intervention from this uncontrolled pilot, the preliminary evidence of reduced hypoglycemia and increased awareness suggests that addressing beliefs and motivation is very relevant to this population.

The novelty of our intervention lies in the following: first, in training diabetes educators in motivational interviewing to address resistance around behavior change that might have occurred and did occur; second, in providing educators with basic cognitive behavioral therapy skills to provide patients with an explicit model of how thoughts can influence behavior; and third, in offering accessible visual metaphors (e.g., the ostrich burying its head in the sand for underestimation of the impact of hypoglycemia) by which patients could identify their own “thinking traps” and explore alternatives. Although additional technologies addressing hypoglycemia, such as pumps, sensors, and islet transplantation, were available at our centers, few of our patients had accessed them; some were more open to these after DAFNE-HART. This, together with the evidence of behavior change described above, suggests that people may be more ready to use additional technologies to address problematic hypoglycemia if motivational factors to regain awareness are addressed.

These data are from a pilot: small patient numbers and not controlled for the impact of the educational element alone. The early evidence of success supports a key role for cognitive barriers in apparently intractable IAH. The intervention now requires refinement, based on participant (patients and educators) experiences, and testing in a randomized, controlled trial. It has potential to be delivered at scale by existing diabetes educators, trained and supported by clinical psychologists, rather than requiring referral to a new cadre of diabetes-experienced psychologists. Meanwhile, we conclude that motivational and cognitive barriers to hypoglycemia avoidance in people with significantly impaired hypoglycemia awareness are legitimate therapeutic targets.

Acknowledgments. The authors are very grateful to D. Cox and L. Gonder-Frederick (both of University of Virginia, Charlottesville, VA) for advice and assistance in preparing the intervention and the generous gift of questionnaires to assess fear of hypoglycemia and hyperglycemia. The authors are also grateful to Dr. K. Winkley (King’s College London) and Dr. C. Emery (University of Sheffield) (research managers for the National Institute for Health Research [NIHR] programme grants RP-PG-0606-1142 and RP-PG-0606-1148) for their support in administering the educator training and validation, D. Fairbairn and M. Patterson (members of the DAFNE User Advisory Group [DUAG], supported by the Central DAFNE Office, Northumbria Healthcare Foundation Trust, Northumbria, U.K.) for helping in the course development, and educators V. Francis and P. Marks (King’s College Hospital NHS Foundation Trust) for help in delivering courses. The authors also thank the people with diabetes who contributed to and participated in this pilot study.

Funding. This study contains independent research funded by the NIHR under its Programme Grants for Applied Research scheme (RP-PG-0606-1142 and RP-PG-0606-1148).

Duality of Interest. S.H. has undertaken consultancy on behalf of Eli Lilly and Company, Novo Nordisk, Sanofi, and LifeScan, for which his institution has received payment, and spoken on behalf of Eli Lilly and Company and Novo Nordisk, for which he has received payment. P.C. has undertaken consultancy on behalf of and/or received speaker fees from Novo Nordisk, Eli Lilly and Company, Johnson & Johnson, Medtronic, and Merck Sharp & Dohme. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. N.d.Z. designed the intervention and supervised its implementation and drafted the manuscript. M.S. conducted the analysis of the 12-month data and assisted in its collection. H.R., P.C., and J.E. contributed to the protocol and the construction of the intervention and its delivery and analyzed data. C.G., S.B., and S.H. contributed to the protocol and the construction of the intervention and its delivery. E.B. analyzed data and created the database for the analysis. S.A.A. contributed to the protocol and the construction of the intervention and its delivery, created the database for the analysis, and drafted the manuscript. All the listed authors made substantial contributions to the study design, performance, and/or analysis and contributed to and/or reviewed and revised the final manuscript. S.A.A. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. This study was presented in oral form at the 73rd Scientific Sessions of the American Diabetes Association, Chicago, IL, 21–25 June 2013.