Implementing simple algorithms to improve glucose and lipid management in people with diabetes and acute coronary syndrome

Aim: Diabetes mellitus is associated with increased risk of adverse outcomes following acute coronary syndrome. Translating evidence-based recommendations into practice is necessary to improve outcomes. We evaluated whether implementing algorithms to guide inpatient care improved glycaemic control, and increased use of sodium–glucose co-transporter 2 (SGLT2) inhibitors and lipid-lowering medication in a tertiary cardiac unit. Method: A 3-month audit (phase 1) was conducted to evaluate hyperglycaemia and dyslipidaemia management, and medication prescriptions. Consecutive people with diabetes admitted for acute coronary syndrome were prospectively identified. Target blood glucose level was defined as 5–10 mmol/l. A multidisciplinary committee designed and implemented decision-support algorithms plus education. A 3-month post-implementation audit (phase 2) was conducted. Results: There were 104 people in phase 1 and 101 in phase 2, with similar characteristics [HbA_1c 64 ± 20 mmol/mol vs. 61 ± 21 mmol/mol (8.0 ± 1.8% vs. 7.8 ± 1.9%]. Post implementation, the incidence of blood glucose levels > 10 mmol/l was lower [phase 1: 46.4% vs. phase 2: 31.8%, rate ratio (RR) = 0.77, 95% confidence intervals (CI) 0.60–0.98; P = 0.031], without a difference in blood glucose levels < 5mmol/l (phase 1: 4.9% vs. phase 2: 4.5%, RR = 1.20, 95% CI 0.70–2.08; P = 0.506). SGLT2 inhibitor prescriptions increased significantly (baseline to discharge: 12.5% to 15.4% vs. 7.9% to 24.8%; P = 0.007) but high-intensity statin prescriptions did not (baseline to discharge: 35.6% to 72.1% vs. 40.6% to 85.1%; P = 0.074). Prescription rates of non-statin lipid-lowering medications were not significantly increased. Conclusions: Implementing decision-support algorithms was associated with improved inpatient glycaemic control and increased use of cardioprotective therapies at discharge in people with diabetes and acute coronary syndrome.