Time in blood glucose range 70 to 140 mg/dl >80% is strongly associated with increased survival in non-diabetic critically ill adults

James S Krinsley, Jean-Charles Preiser

Hyperglycemia, hypoglycemia and increased glucose variability are independently associated with increased risk of death in critically ill adults. The relationship between time in targeted blood glucose range (TIR) and mortality is not well described, and may have been a factor that confounded the results of the major interventional trials of intensive insulin therapy. This is a retrospective analysis of prospectively collected data, involving 3,297 patients with intensive care unit (ICU) length of stay (LOS) at least 1.0 day admitted between January 1, 2009 and December 31, 2013 to a single mixed medical surgical ICU. We investigated the relationship between TIR 70-140 mg/dL with mortality, and compared outcomes of non-diabetics (NON) and diabetics (DM), including stratifying by TIR above (TIR hi) and below (TIR lo) the median value for NON and DM. There were 85,799 BG values for NON and 32,651 for DM; 75.5% and 54.8% were between 70-140 (p < 0.0001). The median (IQR) TIR (%) for NON and DM were 80.6 (61.4-94.0) and 55.0 (35.5-71.1) (p < 0.0001). For NON, mortality was 8.47% and 15.71% for TIR hi and TIR lo (p < 0.0001). For DM moratlity was 16.09% and 14.44% for TIR Hi and TIR lo (p = NS). We observed similar relationships for NON when we stratified by ICU length of stay (LOS) or severity of illness, especially in the most severely ill. There was a cumulative interaction of indices of hypoglycemia, hyperglycemia or glucose variability with TIR. Multivariable analysis demonstrated, for NON, that TIR hi was independently associated with increased survival (p = 0.0009). For NON, the observed:expected mortality ratios for TIR hi and TIR lo, using Acute Physiology and Chronic Health Evaluation IV methodology, were 0.53 and 0.78. In contrast, among DM there was no clear relationship between TIR 70-140 mg/dL and survival. Time in BG range 70-140 mg/dL > 80% is strongly associated with survival in critically ill non-diabetic patients, independently of ICU LOS and severity of illness. These findings have implications for the design of clinical protocols for glycemic control in the critically ill as well for the design of future interventional trials of intensive insulin therapy.