Outcomes of a co-designed clinical decision support tool for inpatient glycaemia management

Hypo- and hyperglycaemia are major issues for hospital inpatients with diabetes, associated with risk of increased adverse outcomes, including higher mortality and prolonged length of stay.

We investigated whether implementation of a multidisciplinary co-designed, non-interruptive electronic medical record (EMR)-based alert for glycaemia was associated with changes in diabetes medication management in hospital.

This was an observational study from September 2022 to January 2023 comparing 7-week pre-intervention and intervention periods separated by a 4-week non-analysis transition. Non-interruptive alerts for hypo- and hyperglycaemia were deployed through the EMR, which provided management recommendations. The primary outcome was electronic alert acknowledgement. Secondary outcomes included medication changes and adverse glycaemic event recurrence.

During the intervention period, rates of alert acknowledgement were very low (hypoglycaemia 1.0% and hyperglycaemia 0.1%). Following alert implementation there was no significant change in medication modification in response to hypoglycaemia (relative risk (RR) 1.42, P = 0.500) or hyperglycaemia (RR 0.64, P = 0.141) events compared with pre-intervention. There was no significant difference in risk for hypoglycaemia recurrence (RR 1.89, P = 0.443), but the risk for hyperglycaemia recurrence was lower (RR 0.79, P = 0.014) during intervention compared to pre-intervention.

Despite multidisciplinary co-design, education and implementation, non-interruptive alerts for glycaemia were not electronically acknowledged in most cases. The alert did not change medication management and thus was unlikely to be the cause of reduction in recurrence of adverse glycaemic events. Further research is required to determine barriers, to develop strategies to improve user engagement and to optimise glycaemic-based electronic alerts, aiming to assist health professionals working in acute care.