OBJECTIVES: To perform an updated systematic review and meta-analysis of the efficacy and safety of intensive (INT) vs. conventional (CONV) blood glucose (BG) targets for critically ill adults on insulin infusions. DATA SOURCES: We conducted a comprehensive search of Embase and OVID Medline databases from inception to October 16, 2023. We manually excluded studies published before 2000 due to potential lack of relevance as glycemic control in the ICU was not routinely practiced before 2000. STUDY SELECTION: We included randomized controlled trials (RCTs) evaluating adult, critically ill patients on insulin infusions comparing INT vs. CONV targets for efficacy and safety outcomes. DATA EXTRACTION: Data were screened and extracted with accuracy confirmed by a second reviewer. Study methodological characteristics, patient population, interventions, and outcome data were recorded. Studies without numerical outcomes were summarized as text statements. DATA SYNTHESIS: Forty-five RCTs were included involving 32,215 patients. No differences were seen between INT and CONV targets for hospital mortality or ICU mortality. INT targets were associated with lower ICU length of stay (LOS), infections, and critical illness polyneuropathy (CIP); however, INT targets demonstrated a 3.6-fold higher risk of severe hypoglycemia. Most of the studies with significant differences contained serious inconsistencies or risk of bias. In the subgroup analyses, INT targets demonstrated favorable neurologic outcomes in neurologic ICU patients, lower ICU LOS in mixed ICU patients, and lower ICU mortality in the cardiac surgery subgroup. CONCLUSIONS: INT BG targets demonstrated mild to moderate improvements in several important morbidity secondary outcomes, including LOS, infections, and CIP, but were associated with a 3.6-fold higher risk of severe hypoglycemia. No differences were seen in ICU or hospital mortality. INT targets should not be routinely used over CONV targets when trying to minimize hypoglycemia as a marker of patient safety. However, as stated in the Society of Critical Care Medicine guidelines, a lower target within the INT range (110-140 mg/dL; 6.1-7.8 mmol/L) may be considered acceptable in select centers where the risk of hypoglycemia is documented to be negligible based on routine assessment and with the use of optimized glycemic management protocols.
Sirimaturos et al. (Tue,) studied this question.