There are three distinct forms of malglycemia in ICU independently associated with hospital mortality.
•Each of these displays unique potencies and temporal expressions.
•Hypoglycemia (absolute and relative) was dominant in the first 24 h.
•This may have implications for interpretation of interventional studies.
AbstractIntroductionThe relationship between critical care mortality and combined impact of malglycemia remains undefined.
MethodsWe assessed the risk-adjusted relationship (n = 4790) between hospital mortality with malglycemia, defined as hypergycemia (hours Glycemic Ratio ≥ 1.1, where GR is quotient of mean ICU blood glucose (BG) and estimated average BG), absolute hypoglycemia (hours BG < 70 mg/dL) and relative hypoglycemia (excursions GR < 0.7 in those with HbA1c ≥ 8%).
ResultsEach malglycemia was independently associated with mortality - hyperglycemia (OR 1.0020/h, 95%CI 1.0009–1.0031, p = 0.0004), absolute hypoglycemia (OR 1.0616/h, 95%CI 1.0190–1.1061, p = 0.0043), and relative hypoglycemia (OR 1.2813/excursion, 95%CI 1.0704–1.5338, p = 0.0069). Absolute (7.4%) and relative hypoglycemia (6.7%) exposure dominated the first 24 h, decreasing thereafter. While hyperglycemia had lower risk association with mortality, it was persistently present across the length-of-stay (68–76% incidence daily), making it the dominant form of malglycemia. Relative contributions in the first five days from hyperglycemia, absolute hypoglycemia and relative hypoglycemia were 60%, 21% and 19% respectively.
ConclusionsAbsolute and relative hypoglycemia occurred largely in the first 24 h. Relative to all hypoglycemia, the associated mortality from the seemingly less potent but consistently more prevalent hyperglycemia steadily accumulated with increasing length-of-stay. This has important implications for interpretation of study results.
Section snippetsBackgroundIn contrast to the expectations raised by the initial Leuven I study [1], subsequent prospective controlled randomized trials did not result in an improved outcome for patients randomized to tight glycemic control [[2], [3], [4], [5], [6], [7], [8]].
Stress associated with acute physiology (surgery, trauma, acute medical disease) can vary widely in both the degree of stress experienced and the associated temporal patterns [[9], [10], [11]]. This is a result of a time-varying magnitude of insulin
MethodsThis observational cohort study comprised patients admitted to the ICU between 10/11/11 and 11/30/19 and was approved by the institutional review board of Stamford Hospital (Western IRB #1–1,167,662-1) with no need for informed consent.
Study setting details, background BG targets, management approaches, patient selection processes and patient characteristics for this cohort have been described previously [[13], [14], [15]] in related papers and are identical to those in this study. These are
ResultsThere were 431,633 h of ICU exposure for the entire cohort (n = 4790) consisting of 2056 h of exposure to BG < 70 mg/dL (0.47% of total ICU time) and 166,038 h of exposure to GR ≥ 1.1 (38.5% of total ICU time), while 342 (7.1%) were exposed to relative hypoglycemia.
The three forms of malglycemia were independently associated with in-hospital mortality (Table 1). The hours at GR ≥ 1.1 (p = 0.0004), hours at BG < 70 mg/dL (p = 0.0043), and excursions below GR 0.7 in those with HbA1c ≥8.0% (p
DiscussionIn this single centre large cohort study three markers of malglycemia (hours at BG < 70 mg/dL, excursions GR < 0.7 for HbA1c ≥ 8%, and hours at GR ≥ 1.1) maintained distinct and independent associations with mortality in modelling adjusted for each other and for risk-adjusted mortality. Exposure to hypoglycemia, both absolute and relative, was most prevalent in the first 24 h of ICU admission, accounting for the majority of malglycemia risk associated with mortality on the first day. Exposure
ConclusionThree distinct elements of malglycemia in the ICU have been defined and are independently associated with increased mortality. Each form of malglycemia has its own distinct potency profile and is associated with unique temporal patterns of exposure over the ICU LOS. These are crucial for understanding the impact on associated mortality risk and have important implications for interpretation of study findings. Malglycemia risk evaluation incorporating individualised quantification of exposure to
Funding sourcesThis research did not include any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Drs. Krinsley is a consultant for Dexcom. Dr. Hirsch's institution received funding from Dexcom and Mannkind; he disclosed that he is a consultant for Abbott, Roche, and GWave. Dr. Umpierrez institution received funding from Dexcom, Abbott, and Baxter. The remaining authors have disclosed that they do not have any potential conflicts of interest.
CRediT authorship contribution statementGregory W. Roberts: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. James S. Krinsley: Writing – review & editing, Writing – original draft, Validation, Resources, Methodology, Investigation, Data curation, Conceptualization. Jean-Charles Preiser: Writing – review & editing, Writing – original draft, Conceptualization. Stephen Quinn: Writing – review & editing, Writing –
AcknowledgementsNone.
References (28)J. Gunst et al.Tight blood-glucose control without early parenteral nutrition in the ICUN Engl J Med
(2023)
G. Umpierrez et al.Randomized controlled trial of intensive versus conservative glucose control in patients undergoing coronary artery bypass graft surgery: GLUCO-CABG trialDiabetes Care
(2015)
J. Bohé et al.Individualised versus conventional glucose control in critically-ill patients: the CONTROLING study-a randomized clinical trialIntensive Care Med
(2021)
The NICE-SUGAR study investigatorsIntensive versus conventional glucose control in critically ill patientsN Engl J Med
(2009)
View full text© 2024 Elsevier Inc. All rights reserved.
留言 (0)