Advertisement
Mayo Clinic Proceedings Home

The Search Continues for Optimal Intensive Care Unit Glucose Management and Measurement

      The optimal management of critical illness–induced dysglycemia, defined as hyperglycemia, hypoglycemia, and glucose variability during and after critical illness, continues to be sought, as does the most acceptable method to measure glucose in a timely manner in the intensive care unit (ICU) and acute care environment.
      • Smith F.G.
      • Sheehy A.M.
      • Vincent J.L.
      • Coursin D.B.
      Critical illness-induced dysglycaemia: diabetes and beyond.
      • Fahy B.G.
      • Sheehy A.M.
      • Coursin D.B.
      Glucose control in the intensive care unit.
      A plethora of studies have heightened our awareness of the potential effect of dysglycemia in adult ICU patients. Strict or “tight” glycemic control (80-110 mg/dL [to convert to mmol/L, multiply by 0.0555]) in critically ill adults was advocated from 2001 through 2009.
      • Fahy B.G.
      • Sheehy A.M.
      • Coursin D.B.
      Glucose control in the intensive care unit.
      • Van den Berghe G.
      • Wouters P.
      • Weekers F.
      • et al.
      Intensive insulin therapy in critically ill patients.
      This strategy, however, ultimately fell by the wayside for a more cautionary goal of 180 mg/dL or less after the NICE-SUGAR (Normoglycemia in Intensive Care Evaluation–Survival Using Glucose Algorithm Regulation) trial reported that strict glycemic control (81-108 mg/dL) was associated with more deaths compared with conventional goals of less than 180 mg/dL.
      • Finfer S.
      • Chittock D.R.
      • Su S.Y.
      • et al.
      Intensive versus conventional glucose control in critically ill patients.
      Questions regarding optimal inpatient glucose management remain. First, it seems that specific patient populations have different risks from the same degree of dysglycemia. For example, critically ill patients with known diabetes mellitus (DM) reportedly fare better than those without DM who have a critical illness or stress-induced hyperglycemia and dysglycemia when mean blood glucose (BG) levels are comparably elevated and variable in these patient groups.
      • Plummer M.P.
      • Bellomo R.
      • Cousins C.E.
      • et al.
      Dysglycaemia in the critically ill and the interaction of chronic and acute glycaemia with mortality.
      Such relevant information regarding the presence or absence of DM may not be known in a substantial number of patients at the time of ICU admission because a large and growing percentage of the population has undiagnosed DM, pre-DM, or impaired fasting glucose levels.
      Centers for Disease Control and Prevention
      National diabetes statistics report, 2014.
      Screening measures (eg, glycated hemoglobin) to diagnose patients on admission to the ICU are valuable yet have important confounders in the acutely ill patient, such as the presence of concurrent hemolysis, recent transfusion, aplastic anemia, or hemoglobinopathies. Earlier and more common acquisition of these markers is likely indicated to guide glycemic control decisions and referral for patients with previously undiagnosed DM. This would also guide follow-up because hospital-associated dysglycemia in the ICU patient without DM portends a high risk of future DM.
      • Jivanji C.J.
      • Asrani V.M.
      • Windsor J.A.
      • Petrov M.S.
      New-onset diabetes after acute and critical illness: a systematic review.
      In addition, technical and patient-specific characteristics may limit accurate and timely glucose measurements in adult ICU patients. For example, edema, peripheral temperature, vasoconstriction, and use of some concurrent medications can potentially alter capillary point-of-care glucose measurements.
      • Critchell C.D.
      • Savarese V.
      • Callahan A.
      • et al.
      Accuracy of bedside capillary blood glucose measurements in critically ill patients.
      • Inoue S.
      • Egi M.
      • Kotani J.
      • Morita K.
      Accuracy of blood-glucose measurements using glucose meters and arterial blood gas analyzers in critically ill adult patients: systematic review.
      • Le H.T.
      • Harris N.S.
      • Estilong A.J.
      • Olson A.
      • Rice M.J.
      Blood glucose measurement in the intensive care unit: what is the best method?.
      • Klonoff D.C.
      • Draznin B.
      • Drincic A.
      • et al.
      PRIDE statement on the need for a moratorium on the CMS plan to cite hospitals for performing point-of-care capillary blood glucose monitoring on critically ill patients.
      Two retrospective studies in this issue of the Proceedings provide additional information on glycemic control, addressing, specifically, glucose management and glucose measurement. First, in a study of 6387 ICU patients and 4462 ICU survivors on general care wards at 2 different medical centers, Krinsley et al
      • Krinsley J.S.
      • Maurer P.
      • Holewinski S.
      • et al.
      Glucose control, diabetes status, and mortality in critically ill patients: the continuum from intensive care unit admission to hospital discharge.
      compared mortality risk in patients with and without DM based on mean BG level (stratified as 80 to <110, 110 to <140, 140 to <180, and ≥180 mg/dL), hypoglycemic events (BG level <70 mg/dL), and glycemic variability (coefficient of variation ≥20%). In patients without DM, increased mean BG levels, glycemic variability, and hypoglycemia were all associated with higher mortality for ICU patients and ICU survivor general care patients. Notably, at mean glucose levels of 80 to less than 110, 110 to less than 140, 140 to less than 180, and at least 180 mg/dL, the relationship was more pronounced in ICU patients (4.50%, 7.30%, 12.16%, and 32.82% mortality, respectively) than in ICU survivors on general care wards (2.74%, 2.64%, 7.88%, and 5.66% mortality, respectively). In patients with DM, increased risk of death was seen in ICU and general care patients with hypoglycemic events and in ICU patients with glucose variability. However, neither glucose variability in ICU survivors nor increased mean BG levels in either ICU patients or ICU survivors conferred increased risk of mortality. Given the increased risk of mortality in patients without DM with elevated glucose levels, the authors suggest that liberal glycemic control is not appropriate for these patients.
      This study contributes to the literature by associating mortality with the trajectory of glycemic control from the ICU to the hospital wards, yet several things should be considered before adjusting current ICU glycemic goals. First, DM status was determined prospectively at the time of admission by “all available clinical information obtained from patients, surrogates, and the electronic medical record.”
      • Krinsley J.S.
      • Maurer P.
      • Holewinski S.
      • et al.
      Glucose control, diabetes status, and mortality in critically ill patients: the continuum from intensive care unit admission to hospital discharge.
      Unfortunately, more than 25% of people with DM are undiagnosed,
      Centers for Disease Control and Prevention
      National diabetes statistics report, 2014.
      and hemoglobin A1c level was measured only in a small (<10%) number of patients in this study while hospitalized. Patients in whom underlying DM was unrecognized, if included in the non-DM group in this study, may have altered outcomes, and further delineation of pre-DM status was not evaluated. It is unknown whether this latter group—patients without DM but with impaired fasting BG levels or pre-DM—is better served with the suggested strict control or with conventional management. Second, management of the ICU and floor patients (general care ward) was not standardized. In the ICU, nurses determined the frequency of BG measurement, the method to assess glucose was not standardized, and capillary samples were allowed. Notably, there is only 1 glucose meter that is Food and Drug Administration (FDA) approved for use in the critically ill, and it is not approved for BG measurements using capillary samples in the ICU population due to potential inaccuracies.
      • Critchell C.D.
      • Savarese V.
      • Callahan A.
      • et al.
      Accuracy of bedside capillary blood glucose measurements in critically ill patients.
      • Inoue S.
      • Egi M.
      • Kotani J.
      • Morita K.
      Accuracy of blood-glucose measurements using glucose meters and arterial blood gas analyzers in critically ill adult patients: systematic review.
      • Le H.T.
      • Harris N.S.
      • Estilong A.J.
      • Olson A.
      • Rice M.J.
      Blood glucose measurement in the intensive care unit: what is the best method?.
      • Klonoff D.C.
      • Draznin B.
      • Drincic A.
      • et al.
      PRIDE statement on the need for a moratorium on the CMS plan to cite hospitals for performing point-of-care capillary blood glucose monitoring on critically ill patients.
      • Rice M.J.
      • Smith J.L.
      • Coursin D.B.
      Glucose measurement in the ICU: regulatory intersects reality.
      In addition, on the general care ward, the frequency of BG measurement was determined by the physicians and was not standardized. Third, patients were not differentiated based on other medical history or risk factors where glycemic control has been reported to impact outcomes. Fourth, the risk of hypoglycemia and its attendant adverse outcomes, as observed in recent prospective studies and reconfirmed in this article, may increase with tighter glycemic standards. Fifth, association is not causation, and BG level per se may reflect illness severity such that glycemic control may have limited direct clinical effect. Finally, data regarding readmission to the ICU were not included, which may have altered statistical outcomes.
      The pertinent second study in this issue is by Liang et al
      • Liang Y.
      • Wanderer J.
      • Nichols J.
      • et al.
      Blood gas analyzer glucose measurement accuracy.
      and addresses the accuracy of ICU BG measurements obtained using blood gas analyzers (BGAs) compared with the gold standard central laboratory value as the reference value. This is a critical question because prolonged turnaround time for central laboratory glucose testing can delay glycemic management. Liang et al
      • Liang Y.
      • Wanderer J.
      • Nichols J.
      • et al.
      Blood gas analyzer glucose measurement accuracy.
      examined 2671 such paired glucose measurements, including a critical subset of 50 hypoglycemic checks, and found that 98.1% of paired values were within a 95% limit of agreement, with a general bias of −3.1 mg/dL. This number was even more accurate in the hypoglycemic range (100% with a −0.8 mg/dL mean bias). Although these data met the ISO 2013 standard, they did not meet the FDA 2016 final guidelines, which require device accuracy within 12% of central laboratory reference values, with a 5% maximum percentage of noncompliant values. Thus, based on the Centers for Medicare and Medicaid Services proposed restrictions on BG measurement in the ICU, the use of BGAs may not meet acceptable standards. To determine whether BGAs met the goal of limiting patient risk appropriately, the investigators used consensus (CEG) and surveillance (SEG) error grid analyses. They evaluated the degree of risk posed by inaccurate measurements (CEG) and divided the degree of discordance of values into different risk categories (SEG). The CEG analysis showed that 99.9% of BGA values were within the “no-risk” zone, with zero measurements in the significant risk or danger zones. The SEG analysis supported these findings, with no values in the significant or dangerous risk zones.
      The study provides clinically relevant insight into the use of a BGA to measure BG concentration, as previous studies were smaller in size and limited in their analysis of hypoglycemic values, a critically important measure. Their elegant analysis suggests that although measurements do not meet the 2016 FDA guidance criteria, the errors likely do not imply increased patient risk, even for hypoglycemia.
      These 2 studies lead to important conclusions. Notably, dysglycemia of all types is associated with worse outcomes in different patient populations and warrants close study and management. However, optimal management is still being delineated because association does not equate with causation, specific patient groups may benefit from different strategies, and hypoglycemia presents clear risk across all patient types. Standardization of care is challenging, and it is difficult to apply strict control with varied resources without accounting for specific protocols and requirements for glycemic monitoring and intervention. It does seem, however, that BG measurements via point-of-care BGAs are clinically acceptable without added risk.
      Early identification of patients with DM is critical so as to resolve the appropriate degree of glycemic control and to address other possible risk factors. Diabetes, unrecognized DM, and pre-DM are ubiquitous in our population, and each group likely has altered insulin sensitivities and outcomes with different glycemic control regimens. Separating out the overall population risk of mortality based on mean BG levels needs to be balanced by the individual patient risk and outcomes of sentinel hypoglycemic events. Simply put, our screening regimens need to be improved to match our populations.
      Glycemic management relies heavily on local resources, patient history and conditions, provider practice patterns, and measurement techniques. Krinsley et al
      • Krinsley J.S.
      • Maurer P.
      • Holewinski S.
      • et al.
      Glucose control, diabetes status, and mortality in critically ill patients: the continuum from intensive care unit admission to hospital discharge.
      suggested alteration of control regimens for patients with DM vs without DM, encouraging stricter control for patients without DM. We question this recommendation because previous prospective data demonstrated an association of strict glycemic management with hypoglycemic events and adverse consequences.
      • Finfer S.
      • Chittock D.R.
      • Su S.Y.
      • et al.
      Intensive versus conventional glucose control in critically ill patients.
      • Krinsley J.
      • Grover A.
      Severe hypoglycemia in critically ill patients: risk factors and outcomes.
      • Brunkhorst F.M.
      • Engel C.
      • Bloos F.
      • et al.
      Intensive insulin therapy and pentastarch resuscitation in severe sepsis.
      In addition, future studies may uncover other patient- and diagnosis-specific factors that determine outcomes when different glycemic control regimens are imposed. Thus, before changing glycemic goals, we need more complete prospective data on the glycemic control of such groups using standardized measurement and management strategies. Although the use of BGAs seems safe and may improve our ability to quickly and accurately manage dysglycemia, until these more complete studies are undertaken, we support current hospital glycemic control guidelines.

      References

        • Smith F.G.
        • Sheehy A.M.
        • Vincent J.L.
        • Coursin D.B.
        Critical illness-induced dysglycaemia: diabetes and beyond.
        Crit Care. 2010; 14: 327
        • Fahy B.G.
        • Sheehy A.M.
        • Coursin D.B.
        Glucose control in the intensive care unit.
        Crit Care Med. 2009; 37: 1769-1776
        • Van den Berghe G.
        • Wouters P.
        • Weekers F.
        • et al.
        Intensive insulin therapy in critically ill patients.
        N Engl J Med. 2001; 345: 1359-1367
        • Finfer S.
        • Chittock D.R.
        • Su S.Y.
        • et al.
        Intensive versus conventional glucose control in critically ill patients.
        N Engl J Med. 2009; 360: 1283-1297
        • Plummer M.P.
        • Bellomo R.
        • Cousins C.E.
        • et al.
        Dysglycaemia in the critically ill and the interaction of chronic and acute glycaemia with mortality.
        Intensive Care Med. 2014; 40: 973-980
        • Centers for Disease Control and Prevention
        National diabetes statistics report, 2014.
        (Accessed April 16, 2017)
        • Jivanji C.J.
        • Asrani V.M.
        • Windsor J.A.
        • Petrov M.S.
        New-onset diabetes after acute and critical illness: a systematic review.
        Mayo Clin Proc. 2017; 92: 762-773
        • Critchell C.D.
        • Savarese V.
        • Callahan A.
        • et al.
        Accuracy of bedside capillary blood glucose measurements in critically ill patients.
        Intensive Care Med. 2007; 33: 2079-2084
        • Inoue S.
        • Egi M.
        • Kotani J.
        • Morita K.
        Accuracy of blood-glucose measurements using glucose meters and arterial blood gas analyzers in critically ill adult patients: systematic review.
        Crit Care. 2013; 17: R48
        • Le H.T.
        • Harris N.S.
        • Estilong A.J.
        • Olson A.
        • Rice M.J.
        Blood glucose measurement in the intensive care unit: what is the best method?.
        J Diabetes Sci Technol. 2013; 7: 489-499
        • Klonoff D.C.
        • Draznin B.
        • Drincic A.
        • et al.
        PRIDE statement on the need for a moratorium on the CMS plan to cite hospitals for performing point-of-care capillary blood glucose monitoring on critically ill patients.
        J Clin Endocrinol Metab. 2015; 100: 3607-3612
        • Krinsley J.S.
        • Maurer P.
        • Holewinski S.
        • et al.
        Glucose control, diabetes status, and mortality in critically ill patients: the continuum from intensive care unit admission to hospital discharge.
        Mayo Clin Proc. 2017; 92: 1019-1029
        • Rice M.J.
        • Smith J.L.
        • Coursin D.B.
        Glucose measurement in the ICU: regulatory intersects reality.
        Crit Care Med. 2017; 45: 741-743
        • Liang Y.
        • Wanderer J.
        • Nichols J.
        • et al.
        Blood gas analyzer glucose measurement accuracy.
        Mayo Clin Proc. 2017; 92: 1030-1041
        • Krinsley J.
        • Grover A.
        Severe hypoglycemia in critically ill patients: risk factors and outcomes.
        Crit Care Med. 2007; 35: 2262-2267
        • Brunkhorst F.M.
        • Engel C.
        • Bloos F.
        • et al.
        Intensive insulin therapy and pentastarch resuscitation in severe sepsis.
        N Engl J Med. 2008; 358: 125-139

      Linked Article