Mortality In ICU Patients Linked to Blood Glucose Levels
Overview
The study investigated the intricate relationship between mean blood glucose (MBG) levels and mortality among critically ill patients, utilizing data from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Through multivariate logistic regression analysis, the study aimed to discern the association between MBG and intensive care unit (ICU) mortality across different subgroups of critically ill patients.
A cohort of 8,973 critically ill patients was examined, among whom 1,244 succumbed within 28 days, encompassing both male and female individuals. Multivariate adjusted restricted cubic spline analyses unveiled a distinctive “J” shaped relationship between MBG and ICU mortality.
Notably, logistic regression analysis unveiled that in group 3, characterized by glucose levels ≥10 mmol/L, the adjusted odds ratio for 28-day mortality stood at 2.06 (with a 95% confidence interval of 1.65–2.57). Subgroup analyses further elucidated that hyperglycemia exerted a more pronounced impact on ICU mortality among patients lacking diabetes, hypoglycemia, and liver disease. Intriguingly, non-diabetic patients consistently faced a higher ICU mortality risk compared to diabetic patients exhibiting similar levels of hyperglycemia.
In conclusion, the study’s findings illuminate a nuanced “J” shaped relationship between MBG levels and mortality among critically ill patients. These insights underscore the complexity of glycemic management in ICU settings and highlight the need for tailored interventions to optimize patient outcomes.
Introduction
This study addresses the prevalence of abnormal blood glucose levels in critically ill patients, encompassing both pre-existing hyperglycemia and stress-induced hyperglycemia. While observational studies have linked elevated blood glucose levels to increased morbidity and mortality among hospitalized patients, there’s also evidence suggesting that correcting hyperglycemia could potentially improve prognosis. However, findings from the NICE-SUGAR Study, a large international randomized trial, contradicted this notion, revealing that intensive glucose control actually led to higher mortality rates among adults in the intensive care unit (ICU).
The rationale for this study stems from earlier research, notably a 2001 randomized controlled trial by van den Berghe et al., which demonstrated a significant reduction in hospital mortality among patients undergoing critical surgery with strict glycemic control. This promising result prompted widespread adoption of glycemic control guidelines. However, subsequent trials, including a large randomized controlled trial in a mixed medical and surgical ICU setting, failed to replicate the mortality benefit observed in van den Berghe’s study. Moreover, a cohort study involving over 18,000 critically ill adults indicated a potential increase in mortality associated with strict glycemic control, even after adjusting for disease severity.
The primary objective of this study is to examine the relationship between total blood glucose levels and all-cause ICU mortality in critically ill patients during their admission. By investigating this relationship, the study aims to contribute valuable insights into the management of blood glucose levels in the ICU setting, potentially informing future guidelines and treatment protocols.
Method
This retrospective cohort study utilized the MIMIC-IV version 2.0 database, a comprehensive repository of ICU stays spanning from 2008 to 2019, encompassing a total of 76,943 cases. The database contains a wealth of patient data, including demographics, vital signs, laboratory results, treatments, and in-hospital mortality. Yuanyuan Li, ID: 11357708, with authorization from the cooperative institution, accessed and analyzed the data.
The study population comprised ICU patients aged 18 years and above at admission for their first hospitalization and ICU stay. Patients with ICU stays shorter than 2 days or missing daily blood glucose (BG) records were excluded. Mean blood glucose (MBG) levels were calculated using all available BG measurements during ICU stays, and patients were categorized into three groups based on MBG levels: ≤7.8, 7.8–10, and ≥10 mmol/L. Hypoglycemia was defined as BG <3.9 mmol/L.
Data extraction included demographic information, laboratory results, comorbidities (e.g., myocardial infarction, diabetes), vasoactive drug usage, mechanical ventilation, and disease severity assessed by the Simplified Acute Physiology Score (SAPS) II. The primary outcome measure was 28-day mortality, extracted directly from the MIMIC-IV database.
This study design allowed for a thorough investigation into the relationship between MBG levels and 28-day mortality in ICU patients. The extensive dataset facilitated a comprehensive analysis of various factors potentially influencing patient outcomes.
Statistical Analysis
This retrospective study utilized observational datasets stratified by mean blood glucose (MBG) levels to investigate its correlation with 28-day mortality in critically ill patients. Descriptive analysis involved categorical variables expressed in numbers and percentages, with comparisons made using appropriate statistical tests. Continuous variables were assessed for normality, with means and standard deviations presented for normally distributed data and medians with interquartile ranges for non-normally distributed data. Multigroup comparisons of continuous variables were conducted using appropriate statistical tests.
Multivariate logistic regression was employed to assess the relationship between MBG and 28-day mortality, with MBG modeled as both continuous and categorical variables using cut-off values of 7.8 and 10 mmol/L. Adjustments were made for potential confounding factors in three progressive models, including age, sex, comorbidities (such as liver disease and diabetes), and disease severity scores.
Subgroup analyses were conducted based on age, diabetes status, sex, hypoglycemia, and liver disease to further explore the correlation between MBG and 28-day all-cause mortality in critically ill patients. Additionally, a dose-response analysis using a restricted cubic spline model was performed to examine the relationship between MBG levels and mortality.
Statistical analyses were conducted using R software and Free Statistics software, with two-tailed tests performed and statistical significance set at p < 0.05. The study’s rigorous methodology and comprehensive analyses aimed to elucidate the relationship between MBG and mortality outcomes in critically ill patients, providing valuable insights for clinical practice and future research in this field.
Result
In the exploration of 76,943 ICU records from the MIMIC-IV database, meticulous screening led to the inclusion of 8,973 patients for analysis. Key clinical characteristics, stratified by mean blood glucose (MBG) categories, revealed notable trends, including higher mortality rates at elevated MBG levels. Specifically, patients in the third MBG level exhibited substantially higher 28-day ICU mortality compared to other groups.
Multivariable logistic regression unveiled a significant association between MBG levels and ICU mortality, with every 1 mmol/L increase in MBG correlating with a 1.16-fold rise in mortality risk. Further stratification by MBG tertiles highlighted escalated ICU mortality risks for patients with MBG levels between 7.8 and 10 mmol/L and ≥10 mmol/L.
Subgroup analyses underscored the heightened impact of hyperglycemia on ICU mortality in patients without diabetes, liver disease, or hypoglycemia, while patients with diabetes or liver disease did not exhibit significant adverse effects from varying hyperglycemic levels. Remarkably, ICU mortality rates were consistently higher among non-diabetic patients, even at similar high blood glucose levels compared to diabetic counterparts.
Age and sex did not exert significant adverse effects on ICU mortality, suggesting that these demographic factors might not substantially influence outcomes in this context.
Visualized through restricted cubic spline analyses, the relationship between MBG levels and 28-day mortality demonstrated a J-shaped association on a continuous scale, further reinforcing the impact of hyperglycemia on mortality outcomes.
Overall, the study highlights the intricate relationship between MBG levels and ICU mortality, particularly emphasizing the heightened risk among non-diabetic patients. These findings underscore the importance of vigilant glycemic management in ICU settings and prompt further investigation into tailored interventions for specific patient subgroups.
Conclusion
The study confirmed that mean blood glucose (MBG) levels serve as an independent predictor for 28-day all-cause mortality among critically ill patients. Particularly, MBG levels equal to or exceeding 10 mmol/L correlated with a significant 2.45-fold increase in ICU mortality risk. This adverse impact of hyperglycemia was more pronounced in patients without diabetes, hypoglycemia, and liver disease.
Hypoglycemia, a condition characterized by low blood glucose levels, poses grave risks in critically ill patients, including neurological complications and even death due to permanent brain damage. Interestingly, among patients without diabetes, those with MBG levels of ≤7.8 mmol/L exhibited the lowest mortality rates.
The study’s findings align with previous research indicating a strong association between hyperglycemia and increased mortality in critically ill patients, especially in those without diabetes. It suggests that while diabetes patients may have adapted to high blood sugar levels over time, non-diabetic individuals experience greater vulnerability to hyperglycemia-induced mortality and stand to benefit significantly from treatment.
The study proposes that the mechanism behind hyperglycemia’s detrimental effects involves oxidative stress, endothelial injury, and impaired immune responses. It suggests that patients without diabetes may lack the adaptive mechanisms seen in diabetes patients, making them more susceptible to the toxic effects of moderate hyperglycemia in acute care settings.
Despite the study’s strengths, such as its large cohort and subgroup analysis, several limitations exist. These include its retrospective nature, single-center design, and lack of information on factors like insulin infusion and corticosteroid usage. Furthermore, the study underscores the importance of evaluating blood glucose management strategies separately for different patient subgroups, considering factors such as diabetes status, liver disease, and hypoglycemia events.
In conclusion, while higher MBG levels were associated with increased mortality risk in critically ill patients, the study highlights the nuanced impact of hyperglycemia across different patient subgroups. It emphasizes the need for tailored management strategies and further research to better understand and address the complexities of blood glucose control in critical care settings.
Oncology Related Tools
Other
Latest Research
