1 Hyperglycaemia in T1DM and pancreatitis related diabetes

Introduction With the introduction of blood ketone testing, it is now possible to quantify the level of ketosis prior to the patient developing life threatening diabetic keto-acidosis (DKA). Administering correction doses of short acting insulin based on ketone level, similar to that for diabetic sick day rules, may avoid critical care admissions, use of variable rate insulin infusions and reduce hospital admissions. Testing is not without its challenges in particular cost, risk of inappropriate testing in type 2 diabetics (T2DM) and misdiagnosis alternative causes of ketosis. We describe here the experience of introducing ketone testing to A+E and acute wards at a district general hospital. Methods Blood ketone testing was introduced to A+E, 2 medical wards and 2 surgical wards together with training for ward staff and a written guideline. Ketone testing was triggered if a type 1 diabetic (T1DM) or pancreatitis related diabetic had a BM >14 mmol/L. Based on the ketone level the guideline prompted either no action, correction dose of short acting insulin followed by repeat testing or arterial blood gas analysis. A retrospective analysis was performed after 8 months to investigate compliance with the protocol and effect of the protocol on the incidence of inpatient DKA. Patients were identified through search of the laboratory information management system followed by a review of electronically scanned medical records. Results 910 ketone tests were carried out representing 116 patients. 395 tests had no patient identifier and were therefore untraceable. Over half of patients (51%) had type 2 diabetes. Of the patients with T1DM/pancreatitis related diabetes 9 patients presented with DKA on admission. Nearly three quarters (73%) resolved with additional subcutaneous insulin as per protocol and 23% were started on a Variable Rate Insulin Infusion. 1 patient developed DKA during an inpatient admission and the protocol had not been followed in this case. The main reasons for deviation from the protocol was lack of documentation and repeat BM testing. Conclusions The protocol itself is very effective when used correctly. It clearly helps to identify patients at risk of deteriorating, and clearly outlines what steps should be taken. While the guideline is intended for patients with T1DM and Pancreatic Related Diabetes, there were many occasions where the ketone testing was used inappropriately on patients with T2DM. In response to this the protocol was updated to include additional guidance on management of hyperglycaemia in T2DM. Documentation and lack of patient identification information was another key issue.