Control of blood glucose levels among critically ill patients continues to evoke intense attention. Van den Berghe et al., in a landmark study, demonstrated that maintaining blood glucose levels within a narrow range, between 80–110 mg/dl may improve clinical outcomes, including ICU and hospital mortality among patients admitted to a surgical ICU.1 In a similar study of medical patients that followed, there was no difference in hospital mortality in patients who received intensive insulin therapy. However, there was a lower incidence of acute kidney injury, more rapid weaning from mechanical ventilation, and earlier discharge from ICU and hospital.2
Subsequent randomized controlled trials failed to reproduce the findings of these early studies, with reports of undesirable levels of hypoglycemia associated with tight glucose control. The NICE-SUGAR study, in a group of mixed ICU patients, revealed increased mortality with tight glucose control, mostly attributable to hypoglycemic episodes.3 Let us examine the current understanding of the importance of control of blood glucose levels in critically ill patients.
Probable reasons for contrasting results
Could it be possible that the conflicting results in major randomized controlled trials were related to the blood glucose targets aimed for? In the Leuven studies, insulin infusion was commenced when the blood glucose level exceeded 215 mg/dl in the control arm, while in the NICE-SUGAR study, insulin therapy was initiated at a blood glucose level higher than 180 mg/dl. In fact, blood glucose levels were higher in the control arm of the Leuven studies compared to the NICE-SUGAR study. One could argue that the lower blood glucose levels in the control arm of the NICE-SUGAR study may be the optimal range and further lowering of glucose levels may, in fact, lead to harm. One could hypothesize a “U” shaped curve of mortality depending on blood glucose levels. The mortality may reduce with lowering blood glucose up to a certain threshold level; a further decrease in levels may increase mortality. Blood glucose levels were measured on arterial blood samples, using blood gas analyzers in the Leuven studies. In the NICE-SUGAR study, arterial, venous, or capillary glucose levels were measured, depending on the circumstances. Could possible inaccuracies with measurement have led to suboptimal insulin dosing and a higher incidence of hypoglycemia? Furthermore, unlike in the NICE-SUGAR study, early parenteral nutrition was administered in the Leuven studies. Early parenteral nutrition may have resulted in higher blood glucose levels and offered protection against hypoglycemia. Bolus doses of insulin were allowed in the NICE-SUGAR study; this may also have contributed to a higher incidence of hypoglycemia.
Glycemic variability and time in the target range
The mean glucose levels over a finite period of time may remain in the target range; however, significant fluctuations may occur over the same period. The frequency (number of times the blood glucose levels are outside the target range) and the magnitude of variability (the degree of change of glucose levels outside the target range) may have an impact. Egi et al. studied the variability of glucose levels in 7,049 critically ill patients. Both higher mean blood glucose levels and a wider standard deviation from the mean were significantly associated with intensive care and hospital mortality. The standard deviation of glucose levels remained an independent predictor of ICU, and hospital mortality.4 A subsequent retrospective cohort review reinforced the importance of glycemic variability. In this study among mixed medical-surgical ICU patients, there was a step-wise increase in mortality with an increase in the standard deviation of blood glucose levels.5
The time in the target range (TiTR) describes the period of time during which the blood glucose levels remain within the acceptable range expressed as a percentage of the total time. In a study of postoperative cardiac surgical patients, TiTR more than 80% was associated with a significantly lower incidence of atrial fibrillation, shorter duration of mechanical ventilation and ICU stay, and a lower incidence of postoperative wound infection.6 In a retrospective analysis, Krinsley et al. examined the relationship between TiTR and mortality in a mixed medical-surgical ICU. In non-diabetic patients, mortality was significantly higher with the TiTR less than 80%. However, the association between TiTR and mortality was not significant in patients with pre-existing diabetes.7
In a multicentric, retrospective analysis of prospectively collected data, maintenance of blood glucose levels between 80–140 mg/dl among non-diabetic patients was independently associated with lower mortality and higher levels with an increased risk of mortality. In contrast, mortality was higher with a blood glucose range between 80–110 mg/dl compared to a higher target range of 110–180 mg/dl among diabetic patients.8 Several other studies also point to a “diabetes paradox” in critically ill patients, with no significant increase in mortality associated with pre-existing, insulin-treated diabetes mellitus.9,10 Chronic hyperglycemia, as seen in diabetic patients, may lead to cellular conditioning that may protect against episodes of acute glycemia.11 Absence of such preconditioning may lead to poor tolerance of hyperglycemic episodes in non-diabetic patients. Hence, it may be reasonable to assume that tighter control of blood glucose levels may be more important in non-diabetic patients.
Is glycosylated hemoglobin (HBA1c) estimation useful in the critically ill?
There may be a marked difference in clinical outcomes with the maintenance of blood glucose levels within a narrow range between diabetic and non-diabetic critically ill patients. Would the estimation of HBA1c be of value to evaluate the degree of blood glucose control prior to the acute illness and titrate control of blood glucose levels? In a retrospective, observational study, among patients with HBA1c levels higher than 7%, a time-weighted glucose level of more than 180 mg/dl resulted in lower mortality. These findings suggest that among patients with poor metabolic control preceding critical illness, there was greater survival when blood glucose levels were maintained at higher levels; in contrast, survival was lower when euglycemic levels were targeted.12 These findings have been corroborated in a recent before-after interventional trial which aimed for target blood glucose levels of 80–140 mg/dl for patients with HBA1c below 7% and 110–160 mg/dl for those with HBA1c above 7%. In diabetic patients with HBA1c above 7%, higher target blood glucose levels between 110–160 mg/dl resulted in lower mortality.13
What blood glucose levels do we aim for among critically ill patients?
Similar to most other areas of medical practice, one size may not fit all with control of blood glucose among the critically ill. Patients exposed to chronic hyperglycemia and those who were normoglycemic in the premorbid period may require different blood glucose targets when they develop an acute illness. Diabetic patients who have adequately controlled blood glucose levels may also differ from those with poor metabolic control. Based on the best evidence available, blood glucose levels between 140–200 mg/dl for non-diabetics and diabetic patients with HbA1c less than 7% may be targeted; a higher level of 160–220 mg/dl for diabetics with HbA1c more than 7% may be appropriate.14
The current practice of measuring blood glucose levels at fixed intervals is less than ideal. Near real-time blood glucose monitoring with closed-loop control has been reported using different algorithms. Clearly, the utilization of such systems will pave the way for maintenance of blood glucose level within a narrow therapeutic range without the risk of wide fluctuations. Interstitial and intravascular sensors have been tested; however, the accuracy of many of these devices do not meet the requisite standards, with no approval from regulators for use in critically ill patients. However, near real-time glucose monitoring with closed-loop control holds the promise to enable maintenance of blood glucose levels within the targeted range, without the risk of hypoglycemia.
The bottom line
- Aiming to maintain blood glucose levels in a narrow therapeutic range may lead to an increased incidence of hypoglycemia and poor outcomes.
- Glycemic variability and time spent in the therapeutic range may impact clinical outcomes.
- Patients with pre-existing high blood glucose levels may tolerate the lower range of blood glucose levels poorly.
- Diabetic patients with adequate glucose control and non-diabetics may require lower blood glucose targets.
- Near real-time measurement of blood glucose levels and closed-loop control holds promise in maintaining levels within a narrow therapeutic range.
1. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med. 2001;345(19):1359-1367. doi:10.1056/NEJMoa011300
2. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med. 2006;354(5):449-461. doi:10.1056/NEJMoa052521
3. NICE-SUGAR Study Investigators, Finfer S, Chittock DR, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283-1297. doi:10.1056/NEJMoa0810625
4. Egi M, Bellomo R, Stachowski E, French CJ, Hart G. Variability of blood glucose concentration and short-term mortality in critically ill patients. Anesthesiology. 2006;105(2):244-252.
5. Krinsley JS. Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med. 2008;36(11):3008-3013. doi:10.1097/CCM.0b013e31818b38d2
6. Omar AS, Salama A, Allam M, et al. Association of time in blood glucose range with outcomes following cardiac surgery. BMC Anesthesiol. 2015;15:14. doi:10.1186/1471-2253-15-14
7. Krinsley JS, Preiser J-C. Time in blood glucose range 70 to 140 mg/dl >80% is strongly associated with increased survival in non-diabetic critically ill adults. Crit Care Lond Engl. 2015;19:179. doi:10.1186/s13054-015-0908-7
8. Krinsley JS, Egi M, Kiss A, et al. Diabetic status and the relation of the three domains of glycemic control to mortality in critically ill patients: an international multicenter cohort study. Crit Care Lond Engl. 2013;17(2):R37. doi:10.1186/cc12547
9. Vincent J-L, Preiser J-C, Sprung CL, Moreno R, Sakr Y. Insulin-treated diabetes is not associated with increased mortality in critically ill patients. Crit Care Lond Engl. 2010;14(1):R12. doi:10.1186/cc8866
10. Graham BB, Keniston A, Gajic O, Trillo Alvarez CA, Medvedev S, Douglas IS. Diabetes mellitus does not adversely affect outcomes from a critical illness. Crit Care Med. 2010;38(1):16-24. doi:10.1097/CCM.0b013e3181b9eaa5
11. Klip A, Tsakiridis T, Marette A, Ortiz PA. Regulation of expression of glucose transporters by glucose: a review of studies in vivo and in cell cultures. FASEB J Off Publ Fed Am Soc Exp Biol. 1994;8(1):43-53. doi:10.1096/fasebj.8.1.8299889
12. Egi M, Bellomo R, Stachowski E, et al. The interaction of chronic and acute glycemia with mortality in critically ill patients with diabetes. Crit Care Med. 2011;39(1):105-111. doi:10.1097/CCM.0b013e3181feb5ea
13. Krinsley JS, Preiser J-C, Hirsch IB. Safety and efficacy of personalized glycemic control in critically ill patients: a 2-year before and after interventional trial. Endocr Pract Off J Am Coll Endocrinol Am Assoc Clin Endocrinol. 2017;23(3):318-330. doi:10.4158/EP161532.OR
14. Marik PE, Egi M. Treatment thresholds for hyperglycemia in critically ill patients with and without diabetes. Intensive Care Med. 2014;40(7):1049-1051. doi:10.1007/s00134-014-3344-2