| Trial | Study Design | Patients | Comparator Groups | Findings | Comments |
| NICE-SUGAR study investigators. [25] | Randomized post-hoc analysis | Critically-ill patients in intensive care units (N = 6026) | Intensive glucose control: BG range 81 - 108 mg/dL Conventional glucose control: Target 180 mg/dL or less | Mortality: No difference between groups Morbidity: Moderate/severe hypoglycemia correlated with longer ICU stays | Mortality increased in moderate/severe hypoglycemia, regardless of treatment group |
| Griesdale, DEG; et al. [26] | Meta-analysis of 26 trials | Critically-ill patients in intensive care units | -- | Intensive glucose control had no effect on the overall mortality rate | Intensive insulin therapy may be beneficial to patients admitted to a surgical ICU |
| NICE-SUGAR study investigators. [27] | Randomized post-hoc analysis | Critically-ill patients in intensive care units (N = 6026) | Moderate hypoglycemia: BG range 41 - 70 mg/dL Severe hypoglycemia: BG range 40 mg/dL or less | Length of Stay: Increased for those experiencing moderate or severe hypoglycemia Mortality: No difference between groups | No causal relationship could be established, but an association was found between hypoglycemic events and death. |
| NICE-SUGAR study investigators. [28] | International randomized- controlled trial subgroup follow-up analysis | Traumatic brain injury (TBI) patients in intensive care units (N = 391) | Intensive glucose control: BG range 81 - 108 mg/dL Conventional glucose control: Target 180 mg/dL or less | Mortality: No difference between groups Morbidity: Glasgow outcomes score unchanged between groups | Follow-up of TBI patients who participated in the NICE-SUGAR study. |
| Krinsley, JS. [29] | Literature review of observational cohort studies | Critically-ill patients in intensive care units | -- | Hypoglycemia, hyperglycemia, and more glucose variability are each associated with increased risk of death. Time in range >80% for nondiabetics is strongly and independently associated with survival. | Limiting glucose variability and improving the time within the goal range may be the key target to improving outcomes. |
| Perez, A; et al. [18] | Meta-analysis | Hospitalized patients, both critically and non-critically ill | -- | Extremes of blood glucose led to poor outcomes. A target glucose range of 110 - 180 mg/dL is most appropriate for critically ill and noncritically ill patients. | Continuous intravenous insulin infusion is preferred for critically ill patients and scheduled basal-bolus-correction insulin is preferred in noncritically ill patients. |