Pragmatic review of interventions to prevent inpatient hypoglycaemia.

Abstract

This pragmatic review aimed to map and summarize the literature on model of care interventions to prevent inpatient hypoglycaemia. Model of care interventions were broadly defined as interventions that either directly target the workforce or where implementation had a strong workforce effect. The review intended to provide information for decision-makers in local health care settings regarding potential interventions to prevent inpatient hypoglycaemia in their local context. PubMed, Embase, CINAHL Plus and Scopus were systematically searched from 2009 to 2019 using key search terms for hypoglycaemia and hospital and evaluation. Included articles had to report an inpatient hypoglycaemia-related outcome. Interventions were categorized by intervention type and setting. Dysglycaemia outcomes were extracted (severe-hypoglycaemia, hypoglycaemia, hyperglycaemia and severe-hyperglycaemia). Forty-nine articles were included in the review. Interventions were categorized as: services (n=8), role expansion (n=6), education (n=9), audit and feedback (n=1), alerts and reminders (n=3), protocol implementation methods (n=1), order sets (n=6), insulin charts (n=1) and electronic glycaemic management systems (n=14). Twenty-one articles reported on ICU-specific interventions, and 28 on interventions in non-ICU-specific settings. Study designs were predominantly non-randomized (n=40). The review found positive evidence for a diverse range of evaluated interventions to prevent inpatient hypoglycaemia. Local decision-makers can use this review to identify interventions relevant to their local context. We suggest they evaluate those interventions using a decision analytic framework that combines the published evidence on effectiveness with local prevalence data to estimate the expected cost-effectiveness of the intervention options when implemented in their local context.