Patients who suffer hypoglycaemia in the community requiring the services of an ambulance are known to have a high short term mortality based on previous work. What is not known is the demographics of this group and also what the cause of death was for individuals whom unfortunately passed away in the time following ambulance callout. Importantly, studies looking at interventions to reduce poor outcomes in this group are lacking. The investigators conducted a pilot trial with the main goals to:

1. Characterise this group in more detail and ascertain what was recorded as cause of death in those whom passed away in the months/years following community hypoglycaemia
2. Ascertain if a simple nurse led intervention, focusing on educating participants on avoidance of hypoglycaemia, could improve outcome.

Previous studies, including our own, have shown that in those with diabetes, severe hypoglycaemia (defined as requiring the assistance of emergency ambulance in the community) carries with it a high mortality in the immediate months/short years following the event. What isn't known is what this group of patients dies of and what the demographics of this group may be. This is important as this is a vulnerable patient group in whom few interventions to reduce this excess mortality have been trialled.

Through a collaboration with a local ambulance service, the hospital diabetes research team set out to recruit participants to a small, pilot, randomized controlled trial to try and ascertain more about the demographics of this group, follow them up in a longitudinal fashion and report on their causes of death using information recorded on death certificates (a statutory requirement in the UK). Additionally the investigators sought to provide pilot data on whether a structured nurse led intervention programme could reduce risk.

The local ambulance service contacted the research team, with patient consent, once they had been called to treat a patient with diabetes whom had severe hypoglycaemia. Within 7 days a research nurse approached those whom had given consent to invite them to take part in a clinical trial. Full eligibility criteria are available elsewhere in this document. Following this, groups were split in the following way:

1. Intensive group - participants were randomized to receive a structured nurse led intervention aiming to educate them on the treatment of and how to avoid hypoglycaemia.
2. Standard group - participants were randomized to continue on their current diabetes treatment and were returned to their standard diabetes care provider.
3. Observational group - participants were happy to have a baseline visit, and for their case to reviewed using electronic records moving forward, but they did not wish to be randomized to an intervention.

All three groups had baseline data collected including:

• Diabetes specific information, including HbA1c, type of diabetes, diabetes therapy, duration of diabetes, presence of complications. Blood tests were taken at baseline for all groups if participants consented to have this done.
• Demographic data - including age, gender, smoking status, BMI
• Information of co-morbidity - Any other illness the participant may suffer from, any medications they take.

Information regarding baseline characteristics was confirmed using electronic patient records.

Those whom were randomized to a nurse led intervention received a structured education programme for 12 months, with the bulk of the work coming in the first 12 weeks.

At 12 months, all participants active engagement in the programme was terminated. Data was collected electronically on new hospital admissions and deaths over this period as well as changes in HbA1c. Participants also had their electronic records screened at study termination and thus data could be provided on outcome at 12 months and study end (in this case 42.6 months.)

Inclusion Criteria:

Diabetes mellitus Can provide written informed consent

Exclusion Criteria:

Hypoglycaemia from cause not related to diabetes mellitus

No plan to do so is currently in place. Of course once data is submitted if other researchers contact us we would consider providing data for the purpose of meta analysis

Due to the nature of the intervention 9a nurse led education programme) it was not possible to mask participant or trial team to the intervention