Early electronic specialist-led diabetes care improves post-discharge glycaemia: the STOIC-D Surgery follow-up cohort.

Abstract Objective To assess the effect of an early, electronic specialist-led model of inpatient diabetes care on glycaemic outcomes following discharge and rates of treatment intensification. Methods The STOIC-D trial has demonstrated early, electronic specialist-led care by an inpatient diabetes service reduces inpatient hyperglycaemia. This follow-up study assessed glycaemic outcomes following discharge in a subgroup of 360 STOIC-D trial patients with admission HbA1c ≥ 7% and alive 1 year following hospitalisation. First available HbA1c between 3-15 months following discharge was collected. Multivariable logistic regression identified predictors of clinically significant reduction in HbA1c, defined as ≥ 0.5%. Multivariable linear regression assessed correlation of baseline characteristics with change in HbA1c. Multivariable logistic regression identified predictors of treatment intensification in hospital. Results The early intervention arm experienced a greater HbA1c reduction following hospitalisation [0.4% (3.9mmol/mol) vs -0.04% (-0.4mmol/mol)] compared with pre-admission (p = 0.02). Clinically significant reduction in HbA1c occurred more frequently in the intervention arm [45% vs 31%, (p = 0.01), adjusted odds ratio (aOR) 1.8 (95% confidence interval (CI):1.1-2.9)]. Adjusted multivariable linear regression demonstrated inclusion in the intervention arm correlated with proportional reduction in post-hospitalisation HbA1c. Treatment intensification was more common in the intervention compared with control [36% vs 19%, (p = 0.002), aOR 2.2 (95% CI: 1.3-3.6)]. In a sub-analysis of participants with admission HbA1c 7-8.5%, age <75 years, the aOR for treatment intensification with early intervention was 4.0 (95% CI: 1.6-11.1). Conclusions Treatment intensification and post hospitalisation HbA1c improved following intervention with an early, electronic specialist-led consultation in hospital.