Blood glucose and ketone testing practices in acute admissions of patients with diabetes mellitus: A retrospective audit.

BACKGROUND: Diabetes mellitus (DM) accounts for approximately 20% of hospital admissions. Admission glucose and ketone testing is essential to detect acute metabolic complications. The Joint British Diabetes Societies for Inpatient Care (JBDS-IP) and the American Diabetes Association (ADA) recommend glucose and ketone testing for all hospitalised diabetic patients. We evaluated testing practices, assessed adherence to recommendations, and explored clinical factors associated with testing. METHODS: We retrospectively reviewed 100 consecutive diabetic patients admitted via the emergency department over three months. Adults aged ≥ 16 years with length of stay (LOS) > 24 h were included. Exclusions were advanced renal failure, pregnancy, severe anaemia, significant blood loss, or haemoglobinopathies. Data were extracted from electronic medical records. Analyses used SPSS v22. Group comparisons used Mann-Whitney U, chi-square, or t-tests. Multiple linear regression assessed factors associated with log-transformed LOS; logistic regression evaluated factors associated with glucose testing. RESULTS: Median age was 59.5 years (IQR 13.0); 46% were female. 28% had type 1 and 72% type 2 DM. Glucose was measured on admission in 83% and ketones in 58%. Four patients (4%) with neither developed metabolic acidosis. Patients with type 2 DM were older than type 1 (median 67 vs 42 years, p < 0.001). Median LOS was 8 days (IQR 12.8), with no differences by DM type or sex. Regression models for LOS and glucose testing were not significant. CONCLUSION: Admission glucose and ketone testing was suboptimal and may contribute to preventable metabolic complications. Education and workflow changes are planned, with a re-audit to evaluate impact.