Early Electronic Specialist-led Diabetes Care Improves Postdischarge Glycemia: The STOIC-D Surgery Follow-up Cohort.

Abstract

To assess the effect of an early, electronic specialist-led model of inpatient diabetes care on glycemic outcomes following discharge and rates of treatment intensification. The STOIC-D trial has demonstrated early, electronic specialist-led care by an inpatient diabetes service reduces inpatient hyperglycemia. This follow-up study assessed glycemic outcomes following discharge in a subgroup of 360 STOIC-D trial patients with admission glycated hemoglobin (HbA1c) ≥ 7% and alive 1 year following hospitalization. First available HbA1c between 3 and 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. The early intervention arm experienced a greater HbA1c reduction following hospitalization (0.4% [3.9 mmol/mol] vs -0.04% [-0.4 mmol/mol]) compared with preadmission (P = .02). Clinically significant reduction in HbA1c occurred more frequently in the intervention arm (45% vs 31% [P = .01], adjusted odds ratio [aOR] 1.8; 95% CI, 1.1-2.9). Adjusted multivariable linear regression demonstrated inclusion in the intervention arm correlated with proportional reduction in posthospitalization HbA1c. Treatment intensification was more common in the intervention compared with control (36% vs 19% [P = .002], aOR 2.2 [95% CI: 1.3-3.6]). In a subanalysis of participants with admission HbA1c of 7% to 8.5%, age <75 years, the aOR for treatment intensification with early intervention was 4.0 (95% CI, 1.6-11.1). Treatment intensification and posthospitalization HbA1c improved following intervention with an early, electronic specialist-led consultation in hospital.