Insulin treatment for type 1 diabetes inevitably carries risk of hypoglycaemia (low blood sugar) which can be severe enough to cause coma, seizure, even death. Being unable to feel when blood glucose is falling, a condition called impaired awareness of hypoglycaemia (IAH), increases risk of severe hypoglycaemia 6-fold. IAH can be reversed and risk of severe hypoglycaemia reduced when people are taught how to adjust their insulin around their life-styles through structured education but problematic hypoglycaemia may persist. Many people with apparently intractable IAH and recurrent severe hypoglycaemia have thoughts about hypoglycaemia that form barriers to their ability to avoid hypoglycaemia. They cannot benefit from conventional treatments to reduce hypoglycaemia. The investigators developed the Hypoglycaemia Awareness Restoration Programme for people with type 1 diabetes and problematic hypoglycaemia despite otherwise optimised self-care (HARPdoc), a novel intervention that combines revision of knowledge about hypoglycaemia avoidance with psychological therapies that directly address unhelpful health beliefs about hypoglycaemia. HARPdoc is delivered over six weeks, by diabetes educators to groups of 6 people. In a pilot study, severe hypoglycaemia was greatly reduced in 23 people with very longstanding IAH and recurrent severe hypoglycaemia.

The investigators propose a group-randomised controlled trial of HARPdoc, comparing it to an established educational intervention (Blood Glucose Awareness Training, BGAT) which has also been shown to reduce severe hypoglycaemia. 96 people with type 1 diabetes and problematic hypoglycaemia persisting despite otherwise optimised insulin self-management will be recruited into groups which will be randomised to receive either HARPdoc or BGAT, in 4 centres. The investigators will measure severe hypoglycaemia over two years following courses; hypoglycaemia risk and experience; overall diabetes control and quality of life.

This will be a group randomised trial of HARPdoc, a novel intervention for adults with type 1 diabetes (T1DM) and treatment-resistant impaired awareness of hypoglycaemia (IAH) and severe hypoglycaemia (SH), against Blood Glucose Awareness Training (BGAT), an existing educational programme that has been shown to reduce severe hypoglycaemia rates.

The hypothesis is that HARPdoc, because of its inclusion of psychological therapies addressing cognitive barriers to hypoglycaemia prevention, will be superior to BGAT training in reducing, and maintaining the reduction in, severe hypoglycaemia and in durably restoring awareness of hypoglycaemia in adults with type 1 diabetes whose problematic hypoglycaemia has persisted despite otherwise optimised diabetes self-management strategies. That optimised self-management is defined as a minimum of having attended a structured education programme in flexible intensive insulin therapy and be using its principles as a minimum.

HARPdoc and BGAT are both "talking therapies" of education plus, in the case of HARPdoc, a specific hypoglycaemia-focussed cognitive behavioural therapy (CBT), delivered to small groups (4 to 8 participants at a time) by established diabetes educators trained and supported to deliver each intervention.

Educators (nurses and dietitians) will be trained by the study clinical psychologists using existing curricula, to deliver EITHER HARPdoc OR BGAT. Educators delivering HARPdoc (2 per course), with its novel elements of hypoglycaemia specific cognitive behavioural therapy, will also receive weekly supervision from the study clinical psychologist, during the delivery of patient courses. BGAT educators (1 per course) will be able to seek support on an ad hoc basis.

Both BGAT and HARPdoc have their own course curricula and teaching aids. Participants will be recruited, will document their hypoglycaemia experience over a period of up to 12 months, and then be randomly allocated to a group in which they will either receive HARPdoc or BGAT . Follow up with be for two years with a planned analysis at 12 months. Documentation of SH will be monitored throughout using statistical process control and control mapping, to allow early termination of the study if one intervention is clearly outperforming the other, as assessed by an independent data monitoring committee.

The study will be conducted in the diabetes outpatient facilities and clinical research facilities of participating centres. Potential participants will be invited by their usual diabetes care teams to consider the study. Those expressing interest will be invited to an initial screening visit (visit 1), where the study will be explained and the person's characteristics assessed against the eligibility criteria. Those meeting the criteria by virtue of having impaired awareness of hypoglycaemia (Gold and Clarke scores of 4 or more); and severe hypoglycaemia in the previous 2 years despite having completed, and using the principles of, structured education in flexible insulin with either multiple daily insulin injections or insulin pump therapy, monitored by self-monitored blood glucose tests a day at least four times a day will be given the opportunity to review the patient information literature and consent to, in writing, participation in the study. They will receive a unique study identification code and commenced on formal documentation of their hypoglycaemia experience, using short forms to document any event of severe hypoglycaemia experienced (form A) and reporting retrospectively all hypoglycaemia in the previous month (form B). These forms will be completed throughout the study and participants will be asked for permission to contact them on a monthly basis if there are any missing data. Blood will be taken for measurement locally of the glycated haemoglobin, a reflection of diabetes control, and for potential contributors to high hypoglycaemia risk (liver, kidney and thyroid function, measures of adrenal and pituitary function, screens for coeliac disease and malabsorption and level of anti-insulin antibodies), if these data are not available in the patient records in the last year. The participant's glucose meter will be downloaded and the results of their over the preceding two weeks recorded. Each participant will then be offered two sets of dates for upcoming courses, but will only undertake one. Once two courses have been filled, participants will be randomised to receive either HARPdoc or BGAT.

A second visit (baseline) will be made not more than 3 months before the participant's course. Blood will be taken for central measurement of HbA1c. Participants will complete a questionnaire booklet that enquires about attitudes to, worry about and behaviours around hypoglycaemia (Attitudes to Awareness (A2A) and Hypoglycemia Fear Survey II (HFS-11) scores); symptoms of anxiety and depression (Hospital Anxiety and Depression score (HADS) and Problem Areas in Diabetes (PAID); quality of life (Diabetes Specific Quality of Life (DSQoL), Short Form 12 (SF12) and the EuroQol five dimensions questionnaire (EQ5-D) and self-management skills (Dose Adjustment for Normal Eating (DAFNE) self-management). This booklet has been reviewed and approved by our user group. This visit may be combined with visit one if consent has already been given.

HARPdoc includes 4 full day attendances and 2 one-to-one contacts between the participant and the educator, via telephone or face-to-face as the patient chooses. In weeks 1-3, participants meet weekly as a group. These are full day sessions, as in current DAFNE and the Bournemouth Type 1 Diabetes Education Programme (BERTIE) courses. They cover the pathophysiology, presentation, detection and treatment of hypoglycaemia and IAH; use motivational interviewing to support behaviour change and encourage small changes, with Cognitive Behavioural Therapy (CBT) to address unhelpful cognitions that may be barriers to hypoglycaemia avoidance. In weeks 4 and 5, participants try out their new skills/strategies, with two scheduled individual face-to-face and telephone educator support. Week 6 is a final group session, focusing on relapse prevention and support for significant others.

The BGAT curriculum will be re-planned to be delivered in 4 group sessions spread over 6 weeks to allow participants to be booked inot both courses at one time prior to randomisation. The curriculum will not however be extended to occupy full day sessions.

Sessions will, with participant permission, be video taped, for assessment of fidelity of course delivery to the curricula.

Attendance of at least the first 3 group sessions plus, for HARPdoc participants and the 1:1 sessions, will be considered adherence to the intervention/ control.

Follow-up visits, at 3, 6, 12, 18 and 24 months, are in line with clinical practice, recognizing that these patients are high service users. They will last about 2 hours and be conducted in the groups. Participants will have contact details (telephones with recorded answering and next-working day reply; e-mail addresses) for their educators through which they can seek advice at will.

Scheduled follow-up in groups for re-inforcement of the programme principles and for data collection (locally measured HbA1c, documentation of hypoglycaemia experience and awareness scores) at 3, 6, 12, 18 and 24 months. The 12 and 24 month visits will include completion of the full questionnaire pack as described for visit 2. Additional non-study visits can be scheduled if required in the opinion of the participant and/or the educator. Likewise, participants will be free to seek advice from their educator at any time by telephone or e mail. Participants will be made aware that the forms on which they document their hypoglycaemia experience each month will only be reviewed anonymously and they need to seek advice if they require it, for example, after a hypoglycaemia event, as is routine clinical practice.

For each visit a one month window on either side of the due date is acceptable

Twenty four participants (12 from each group) will, be invited to participate in a substudy in which they wear a real-time continuous glucose monitor for up to one week at baseline and 12 months.

A planned analysis comparing the outcomes of the two interventions is planned when all participants have completed 12 months of follow-up, with a final analysis at 2 years. Throughout the data on severe hypoglycaemia will be collected, blinded as to course, by the Implementation Science group, and entered into control maps, to allow inspection of rates of severe hypoglycaemia in individual patients and in each intervention in real time. These data will be reviewed at 6 monthly intervals by the Data Monitoring Committee, which has the ability to interrupt the RCT if rates of severe hypoglycaemia are increasing in either group or one intervention is clearly superior to the other.

The described randomisation procedure is to prevent "researcher bias" with researchers preferentially selecting higher risk patients for HARPdoc over BGAT. To minimise contamination, different educators will be trained in and will deliver each intervention. A further safeguard against contamination is that both HARPdoc and BGAT patient courses are curriculum driven and the unique elements of HARPdoc are not included in the BGAT curriculum. The fidelity of delivery of each course will be monitored from video-taped sessions . Recordings will be coded using a check list of behaviour change techniques and items from relevant measures such as the Behavioural Change Counselling Index (BECCI) and the Motivational Interviewing Treatment Integrity (MITI) scales.

Participants who wish to withdraw from the trial will remain under services which routinely collect primary end-point data (severe hypoglycaemia rate, HbA1c, Gold score). Unless refused permission, we will collect case-record data +/- 3 months of due dates. This will allow us to conduct both an intention-to-treat and a per protocol analysis.

The target is to recruit 96 participants, 6 (4 - 8) participants per group course into at least 8 HARPdoc courses and 8 BGAT courses), delivered in 4 centres. The study targets adults (people aged 18 and above) with type 1 diabetes who have IAH and severe hypoglycaemia after optimised medical therapy, which may include, or have included, use of insulin pump therapy and/or continuous glucose monitoring. This will allow detection of superiority of HARPdoc over BGAT at the 5% significance level with 90% power, The power calculation was based on data from our pilot of HARPdoc, and published data on BGAT, looking at rates of severe hypoglycaemia. Conservative estimates were used, allowing for 2 episodes per patient per year in the present study from HARPdoc and 5.7 episodes per patient per year from for BGAT, have allowing 20% drop out.

Participants will be identified by study centres and enhanced by referrals from neighbouring diabetes centres and pathways in place for the follow up of people with type 1 diabetes making use of emergency services for severe hypoglycaemia, Potential participants will also be able to self-refer, and will be offered a review by the clinical team of the participating centre before deciding with the patient whether the trial might be suitable for them. People experiencing severe hypoglycaemia who do not fulfil eligibility criteria for the trial or who chose not to participate will be referred into the centre's usual treatment pathway for problematic hypoglycaemia, with the option to be included in a HARPdoc course outside the trial. The exclusion criteria list conditions that led to exclusion from the qualitative research programme on which the HARPdoc curriculum is based.

Inclusion criteria:

96 people, of whom 24 will be recruited in the US centre under their ethical regulations.

• 18 years or older

• type 1 diabetes(1) for at least four years,

• Experiencing problematic hypoglycaemia(2) for at least one year, despite structured education(3) in flexible insulin therapy and on-going optimal conventional care.

• Current use of an appropriate (in the investigator's estimation) multiple daily insulin injection regimen or CSII (insulin pump) therapy(4)

• Willingness to comply with study design, including willingness and ability to perform SMBG up to 4 times a day routinely

• Ability to communicate in written and spoken English

• Ability to give written informed consent.

  1. Type 1 diabetes will be defined clinically, usually based on starting insulin for diabetes within one year of diagnosis and/or a history of diabetic ketoacidosis
  2. Having Gold and Clarke scores of 4 or more and having had ≥ 1 episode/s of severe hypoglycaemia [events requiring assistance of another person to actively administer carbohydrates, glucagon, or take other corrective action, because of impaired cognitive function, and which may include episodes that were not treated by another but included loss of consciousness or seizure] in the last 2 years and at least one since starting current treatment modality.
  3. Structured education requires a programme with a curriculum, taught by trained educators, which covers insulin dose adjustment around carbohydrate counting and lifestyle issues, and a physiological 24 hr basal insulin replacement separate from meal insulin replacement) or as judged equivalent by the local investigator.
  4. Participant should be using dose adjustment around carbohydrate counting and lifestyle issues, and an appropriate, separate basal replacement)

Exclusion criteria:

• People with type 2 diabetes, or type 1 diabetes and good hypoglycaemia awareness

• People with type 1 diabetes and impaired hypoglycaemia awareness who have not attended structured education in flexible intensive insulin therapy, such as DAFNE, BERTIE, the Joslin course (DO IT) or as judged equivalent by the local investigator.

• People not fluent in spoken English

• Current pregnancy (5)

• People with severe mental disorders (schizophrenia, manic depression, depressive psychosis, active suicidal ideation, learning disability, dementia, alcohol and substance dependence, personality disorders, eating disorder)(6).

• Cognitive impairment independent of hypoglycaemia (e.g. clinical diagnosis of dementia(7), advanced Parkinson's disease, neurodegenerative disease)

• Existence of co-morbid medical disease other than diabetes mellitus contributing to hypoglycaemia (e.g. inadequately treated Addison's disease or growth hormone deficiency or hypothyroidism; untreated coeliac disease; uncontrolled gastroparesis; end stage renal disease), which must have been checked since the onset of problematic hypoglycaemia.

  (5) Participants who continue to experience severe hypoglycaemia episodes 6 months after they have stopped breastfeeding may be included in the trial.

  (6) "Manic depression" covers conditions such as bi-polar disorder. Pre-existing depression that is on-going but, in the opinion of the investigator, stable and not a barrier to potential benefit from BGAT or HAPRdoc is not an exclusion criterion. Exclusion of individuals with a personality disorder includes those with current or previous clinical diagnosis AND current or previous mental health care for that disorder

  (7) Dementia would cover either an existing diagnosis or a Mini Mental State Examination [MMSE]) score of less than 24.

Participants who have expressed interest in the study, have consented, and have impaired awareness of hypoglycaemia but do not otherwise meet the inclusion criteria due to low number of SH episodes may be included in the HARPdoc educational course as 'fillers' if space is available. These patients will not be randomised, nor entered into the study database. However, they may be asked to complete the open baseline, 12 and 24-month data collection if they agree to do so.