165-OR: Early Electronic Consultation Reduces Glucose and Health Care–Associated Infection in Hospital—The STOIC-D Surgery Randomised Controlled Trial

Abstract

Aims: To assess the effect of early intervention with electronic-based proactive specialist diabetes care in surgical inpatients on glycaemia and clinical outcomes. Methods: The Specialist Treatment of Inpatients: Caring for Diabetes (STOIC-D) Surgery randomised controlled trial (RCT) recruited consecutive adults admitted to surgical units of the Royal Melbourne Hospital (Australia) in 2021 with diabetes or blood glucose ≥200 mg/dL and length of stay (LOS) ≥24 hours. Intervention arm patients received remote proactive consultation by the inpatient diabetes service (IDS) in the electronic medical record (Epic®) within 24 hours of admission and, if escalation criteria were met, received a bedside consultation. Patients receiving standard care were reviewed by the IDS at the bedside only following referral. Insulin and non-insulin agents were used to target glucose 90-180 mg/dL. Outcomes included glucometrics, healthcare-associated infection (HAI), and mortality. Registration: ACTRN12620001303932. Results: 1,383 admissions met inclusion criteria; 689 received the intervention. The primary outcome of mean patient-day mean glucose was lower in the active (158.4 mg/dL, standard deviation [SD] 48.6) vs. control arm (162.0 mg/dL, SD 46.8, p<0.001). HAI (most commonly pneumonia) was lower in the active vs. control arm (11% vs. 16%, p=0.02). Mortality (2.4% vs. 4.2%, p=0.08) and LOS (10.7 vs. 10.0 days, p=0.26) were no different. The number needed to treat for HAI prevention was 22. Hypoglycaemia <72 mg/dL was not increased (1.0% active vs. 0.9% control, p=0.23). The IDS performed a bedside consultation in 333 (49%) of the active vs. 93 (14%) of the control arm. Conclusion: The STOIC-D Surgery trial is the largest RCT of a diabetes model-of-care intervention in non-critical care. Early, electronic-based specialist diabetes intervention significantly reduced patient-day mean glucose and HAIs in a surgical population. Disclosure R.Barmanray: None. L.J.Worth: None. S.Fourlanos: Advisory Panel; Viatris Inc., Pfizer Inc., Speaker's Bureau; Novo Nordisk, AstraZeneca, Boehringer Ingelheim and Eli Lilly Alliance. M.Kyi: None. P.G.Colman: None. L.M.Rowan: None. L.Collins: None. L.E.Donaldson: Stock/Shareholder; Medtronic, Novo Nordisk. S.Montalto: None. E.Sun: None. M.V.H.Le: None.