Serum lactate monitoring may help to predict neurological function impairment caused by acute metabolism crisis

Interpretable machine learning model for predicting delirium in patients with sepsis: a study based on the MIMIC data

OBJECTIVE

The aim of this study was to construct interpretable machine learning models to predict the risk of developing delirium in patients with sepsis and to explore the impact of delirium on the 28-day survival rate of patients.

METHODS

We enrolled 10,321 patients with sepsis older than eighteen years from the MIMIC-IV (Medical Information Mart for Intensive Care) database. Sepsis is defined as the presence or suspected presence of infection, along with a SOFA (Sequential Organ Failure Assessment) score of ≥ 2. Four machine learning models, namely XGBoost (extreme gradient Boost), SVM (support vector machine), Logistic (logistic regression) and RF (random forest), were established for prediction, and the prediction model was constructed.

RESULTS

A total of 10,321 sepsis patients were included, among whom 4,691 (45.45%) developed delirium. The 28-day mortality rate was markedly elevated in the delirium group (log-rank P < 0.001). The XGBoost model has the best performance. Finally, 5 variables were selected to draw a nomogram: hypertension, SOFA score, chlorine, Hb (hemoglobin), creatinine. The receiver operating characteristic (ROC) curve of the predictive delirium model showed better predictive efficiency, with an AUC of 0.767 (95%CI (confidence interval): 0.726-0.798).

CONCLUSION

The nomogram built on the XGBoost model provides clinicians with an easy tool to quickly assess the risk of developing delirium in patients with sepsis. It provides a new idea and direction for the best model to predict delirium in patients with sepsis, so as to promote the development of delirium related research.

Association Between the Highest Lactate Level on the First Postoperative Day and Postoperative Delirium in Cardiac Surgery Patients

AIMS

The study aimed to determine the correlation between the maximum lactate on the first postoperative day and the incidence of postoperative delirium (POD) in patients after cardiac surgery.

METHODS

The data of cardiac surgery patients were extracted from the Medical Information Mart for Intensive Care IV database. The cut-off value for the first postoperative day maximum lactate was determined, and all patients were categorized into two groups according to the cut-off value. Propensity score matching (PSM) was applied between the two groups, and the difference in the incidence of POD was analyzed. Then, we employed univariate logistic regression, multivariate logistic regression, PSM, and inverse probability of treatment weighting (IPTW) models to examine the relationship between the first postoperative day lactate levels and POD.

RESULTS

Among the 4856 patients enrolled, there was a significant difference in lactate-max on the first postoperative day between patients without POD and patients with POD (median 2.5 vs. 3.1, p < 0.001). The cut-off value of lactate-max was 2.85 mmol/L. For the two groups after PSM, the incidence of POD in the lactate-max ≥ 2.85 mmol/L group was significantly elevated (19.2% vs. 15.9%, p = 0.029). The elevated lactate-max on the first postoperative day was substantially associated with an increased risk of POD in univariate and multivariate logistic regression analyses, PSM, and IPTW models.

CONCLUSIONS

The results demonstrated that the first postoperative day lactate-max was correlated with the risk of POD in patients undergoing cardiac surgery, with the POD risk increasing significantly in patients with a lactate-max ≥ 2.85 mmol/L on the first postoperative day.

Associations of serum lactate and lactate clearance with delirium in the early stage of ICU: a retrospective cohort study of the MIMIC-IV database

Aim

This study aimed to investigate the associations of serum lactate level [within and after 24 h of the intensive care unit (ICU) admission] and lactate clearance rate with delirium and assess associations of lactate and lactate clearance rate with 30-day mortality in delirium patients.

Methods

Data in this retrospective cohort study were extracted from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database in 2012-2019. The associations of lactate and lactate clearance rate with delirium were explored through univariable and multivariable logistic regression analyses, whereas the associations of lactate and lactate clearance rate with 30-day mortality in delirium patients were investigated using univariable and multivariable Cox regression analyses. Subgroup analysis was performed for age, gender, sepsis, hypertension, sedative drug, ventilation, antibiotic drug, vasopressors, and the Sequential Organ Failure Assessment (SOFA) score. The evaluation indexes were odds ratios (ORs), hazard ratios (HRs), and 95% confidence intervals (CIs).

Results

Among 7,812 (14.58%) eligible participants, 4,338 (8.19%) had delirium and 1,903 (24.36%) died within 30 days. After adjusting for covariates, patients with lactic acidosis (lactate level > 5 mmol/L and PH < 7.35) at T0 (within 24 h of the ICU admission) had higher odds of delirium (OR = 1.235, 95%CI: 1.105-1.382). Hyperlactatemia (lactate level 2-5 mmol/L and PH > 7.35) at T1 (after 24 h of the ICU admission) was also associated with higher odds of delirium (OR = 1.277, 95%CI: 1.126-1.447). Lactate clearance rate > 50% was linked to lower odds of delirium (OR = 0.705, 95%CI: 0.613-0.811), and this relationship was also observed in ≥65 years old, female, male, non-sepsis, sepsis, non-hypertension, non-sedative drug use, sedative drug use, ventilation, antibiotic drug use, use of vasopressors, and different SOFA score subgroups (all p < 0.05). Additionally, hyperlactatemia and lactic acidosis (whether at T0 or T1) may be potential risk factors for 30-day mortality in delirium patients, whereas lactate clearance rate ≥ 0 had a potential protective effect on 30-day mortality (all p < 0.05).

Conclusion

Higher serum lactate levels in the early stage of the ICU were associated with a higher risk of delirium and subsequent mortality. Measures taken to increase the lactate clearance rate are necessary to reduce potential delirium or mortality risk in clinical settings. However, more evidence from prospective studies is needed to verify these findings.

Serum Lactate Is an Indicator for Short-Term and Long-Term Mortality in Patients with Acute Pancreatitis

BACKGROUND

Serum lactate, as a single and an easily available biomarker, has been applied in various diseases.

AIMS

In this study, we aimed to explore the predictive value of serum lactate for short-term and long-term prognosis in acute pancreatitis (AP) admitted in intensive care unit (ICU) based on a large-scale database.

METHODS

AP patients admitted in ICU in the MIMIC-IV database were included. We constructed three different models to investigate the relationships between serum lactate and clinical outcomes, including 30-day, 180-day and 1-year mortality in AP. Smooth fitting curves were performed for intuitively demonstrating the relationship between serum lactate and different outcomes in AP by the generalized additive model.

RESULTS

A total of 895 AP patients admitted in ICU were included. The mortalities of 30 days, 180 days, and 1 year were 12.63% (n = 113), 16.87% (n = 151), and 17.54% (n = 157). In model B, with 1-mmol/L increment in serum lactate, the values of OR in 30-day, 180-day and 1-year mortality were 1.20 (95%CI 1.04-1.37, P = 0.0094), 1.21 (95%CI 1.06-1.37, P = 0.0039), and 1.21 (95%CI 1.07-1.38, P = 0.0035). The AUCs of serum lactate for predicting 30-day, 180-day, and 1-year mortality in AP were 0.688 (95%CI 0.633-0.743), 0.655 (95%CI 0.605-0.705), and 0.653 (95%CI 0.603-0.701), respectively. The cut-off value of serum lactate predicting 30-day, 180-day and 1-year mortality in AP was 2.4 mmol/L.

CONCLUSION

Serum lactate could be an indicator for short-term and long-term mortality in patients with AP admitted in ICU.