Diabetes UK; European Foundation for the Study of Diabetes; JDRF; National Institute for Health Research Cambridge Biomedical Research Centre; Wellcome Trust.

Aim To assess the health‐related outcomes of hypoglycaemia for people with diabetes admitted to hospital; specifically, hospital length of stay and mortality. Methods We conducted a systematic review and meta‐analysis of studies relating to hypoglycaemia (< 4 mmol/l) for hospitalized adults (≥ 16 years) with diabetes reporting the primary outcomes of interest, hospital length of stay or mortality. Final papers for inclusion were reviewed in duplicate and the adjusted results of each were pooled, using a random effects model then undergoing further prespecified subgroup analysis. Results In total, 15 studies were included in the meta‐analysis. The pooled mean difference in length of stay for ward‐based inpatients exposed to hypoglycaemia was 4.1 days longer [95% confidence interval (CI) 2.36 to 5.79; I² = 99%] compared with those without hypoglycaemia. This association remained robust across the pre‐specified subgroup analyses. The pooled relative risk (RR) of in‐hospital mortality was greater for those exposed to hypoglycaemia (RR 2.09, 95% CI 1.64 to 2.67; I² = 94%, n = 7 studies) but not in intensive care unit mortality (RR 0.75, 95% CI 0.49 to 1.16; I² =0%, n = 2 studies). Conclusion There is an association between inpatient hypoglycaemia and longer length of stay and greater in‐hospital mortality. Studies examining this association were heterogenous in terms of both clinical populations and effect size, but the overall direction of the association was consistent. Therefore, glucose concentration should be considered a potential tool to aid the identification of inpatients at risk of poor health‐related outcomes.