Your trust reports 42 severe inpatient hypoglycaemic episodes per month requiring rescue treatment. You have been asked to lead a project to reduce them. Outline your approach.

Question

You are the endocrinology registrar. The diabetes inpatient team has flagged that most episodes occur overnight and that a high proportion involve patients who are nil by mouth or eating poorly.

Accepted Answer

I would begin by characterising the 42 episodes rather than accepting the headline number, because the intervention depends entirely on what is actually driving them. For each episode I would capture the time of day, the insulin or sulfonylurea involved and when it was given, whether the patient was nil by mouth or eating poorly, their renal function and whether it had changed, the ward and specialty, and whether a diabetes specialist nurse had been involved. The stem already suggests two themes — overnight timing and disrupted eating — and I would test those against the data rather than assume them. I would also talk to the ward nurses, who will usually tell me the cause faster than the dataset does. My aim would be SMART: to reduce severe inpatient hypoglycaemic episodes requiring rescue treatment from 42 to fewer than 20 per month within nine months. My outcome measure would be episodes per 1,000 bed days plotted on a run chart, process measures would include the proportion of nil-by-mouth diabetic patients on a correct insulin plan and the proportion of patients with a falling eGFR who had a dose review within 24 hours, and my balancing measures would be hyperglycaemia and DKA rates, because a naive project that simply reduces insulin will trade one harm for a worse one. My interventions would be designed into the prescribing system rather than delivered as teaching: a clear nil-by-mouth and variable-rate insulin protocol embedded in the electronic record, an automatic prompt for insulin dose review when eGFR falls, sensible prescribing defaults for overnight long-acting insulin, and an alert that triggers a diabetes specialist nurse review after any severe episode. I would involve pharmacy, ward nursing, catering and the surgical teams whose patients are most often nil by mouth, test changes in small PDSA cycles on two wards first, and embed the protocol in induction and the SOP with a named owner before I rotate.

Reducing Inpatient Hypoglycaemia

Framework checklists