ABSTRACT
BACKGROUND: Risk stratification and outcome prediction are crucial for intensive care resource planning. In addressing the large data sets of intensive care unit (ICU) patients, we employed the Explainable Boosting Machine (EBM), a novel machine learning model, to identify determinants of acute kidney injury (AKI) in these patients. AKI significantly impacts outcomes in the critically ill. METHODS: An analysis of 3572 ICU patients was conducted. Variables such as average central venous pressure (CVP), mean arterial pressure (MAP), age, gender, and comorbidities were examined. This analysis combined traditional statistical methods with the EBM to gain a detailed understanding of AKI risk factors. RESULTS: Our analysis revealed chronic kidney disease, heart failure, arrhythmias, liver disease, and anemia as significant comorbidities influencing AKI risk, with liver disease and anemia being particularly impactful. Surgical factors were also key; lower GI surgery heightened AKI risk, while neurosurgery was associated with a reduced risk. EBM identified four crucial variables affecting AKI prediction: anemia, liver disease, and average CVP increased AKI risk, whereas neurosurgery decreased it. Age was a progressive risk factor, with risk escalating after the age of 50 years. Hemodynamic instability, marked by a MAP below 65 mmHg, was strongly linked to AKI, showcasing a threshold effect at 60 mmHg. Intriguingly, average CVP was a significant predictor, with a critical threshold at 10.7 mmHg. CONCLUSION: Using an Explainable Boosting Machine enhance the precision in AKI risk factors in ICU patients, providing a more nuanced understanding of known AKI risks. This approach allows for refined predictive modeling of AKI, effectively overcoming the limitations of traditional statistical models.
ABSTRACT
BACKGROUND: Veno-arterial extracorporeal membrane oxygenation (vaECMO) removal reflects a critical moment and factors of adverse outcomes are incompletely understood. Thus, we studied various patient-related factors during vaECMO removal to determine their association with outcomes. METHODS: A total of 58 patients from a university hospital were included retrospectively. Demographic, clinical, and echocardiographic parameters were recorded while under vaECMO support, as well as the need for inotropic and vasoactive-inotropic scores (VIS). Successful weaning was defined as 28-day survival without reinitiation of vaECMO. RESULTS: Patient age differed significantly between patients with a successful and a failed vaECMO weaning (54 ± 14 vs. 62 ± 12 years, p = 0.029). In univariable logistic regression, age (OR 0.952 (0.909-0.997), p = 0.038), the necessities for inotropic agents at the time of echocardiography (OR 0.333 (0.113-0.981), p = 0.046), and vaECMO removal (OR 0.266 (0.081-0.877), p = 0.030) as well as the dobutamine dose during removal (OR 0.649 (0.473-0.890), p = 0.007), were significantly associated with a successful weaning from vaECMO. Age (HR 1.048 (1.006-1.091), p = 0.024) and the VIS (HR 1.030 (1.004-1.056), p = 0.025) at the time of vaECMO removal were independently associated with survival in bivariable Cox regression. In Kaplan-Meier analysis, a VIS of >5.1 at vaECMO removal was associated with impaired survival (log-rank p = 0.025). CONCLUSIONS: In this cohort, age and the extent of vasoactive-inotropic agents were associated with adverse outcomes following vaECMO, whereas echocardiographic biventricular function during vaECMO support was not.
ABSTRACT
[This corrects the article DOI: 10.3389/fmed.2022.1014276.].
ABSTRACT
Background: Shock increases mortality in the critically ill and the mainstay of therapy is the administration of vasopressor agents to achieve hemodynamic targets. In the past, studies have found that the NO-pathway antagonist methylene blue improves hemodynamics. However, the optimal dosing strategy remains elusive. Therefore, we investigated the hemodynamic and ICU outcome parameters of three different dosing strategies for methylene blue. Methods: We performed a retrospective cohort study of patients in shock treated with methylene blue. Shock was defined as norepinephrine dose >0.1 µg/kg/min and serum lactate level >2 mmol/l at the start of methylene blue administration. Different demographic variables, ICU treatment, and outcome parameters were evaluated. To compare the differences in the administration of vasopressors or inotropes, the vasoactive inotropic score (VIS) was calculated at different time points after starting the administration of methylene blue. Response to methylene blue or mortality at 28 days were assessed. Results: 262 patients from July 2014 to October 2019 received methylene blue. 209 patients met the inclusion criteria. Three different dosing strategies were identified: bolus injection followed by continuous infusion (n = 111), bolus injection only (no continuous infusion; n = 59) or continuous infusion only (no bolus prior; n = 39). The groups did not differ in demographics, ICU scoring system, or comorbidities. In all groups, VIS decreased over time, indicating improved hemodynamics. Cardiogenic shock and higher doses of norepinephrine increased the chance of responding to methylene blue, while bolus only decreased the chance of responding to methylene blue treatment. 28-day mortality increased with higher SAPSII scores and higher serum lactate levels, while bolus injection followed by continuous infusion decreased 28-day mortality. No severe side effects were noted. Conclusion: In this cohort, methylene blue as a bolus injection followed by continuous infusion was associated with a reduced 28-day mortality in patients with shock. Prospective studies are needed to systematically evaluate the role of methylene blue in the treatment of shock.
