Journal of critical care
-
Journal of critical care · Mar 2024
Multimodal monitoring in patients with acute brain injury - A survey from critical care providers.
Multimodal neuromonitoring (MMM) aims to improve outcome after acute brain injury, and thus admission in specialized Neurocritical Care Units with potential access to MMM is necessary. Various invasive and noninvasive modalities have been developed, however there is no strong evidence to support monitor combinations nor is there a known standardized approach. The goal of this study is to identify the most used invasive and non-invasive neuromonitoring modalities in daily practice as well as ubiquitousness of MMM standardization. ⋯ Our survey revealed overall limited use of MMM with no established institutional protocols among institutions. Ongoing research and further standardization of MMM will clarify its benefit to patients suffering from severe brain injury.
-
Limitations of life sustaining therapies (LLST) are frequent in intensive care units (ICUs), but no previous studies have examined end-of-life (EOL) care and LLST in South Africa (SA). ⋯ Withholding was more common than withdrawing treatment both in SA and worldwide, although both were significantly less frequent in SA compared with the world average.
-
Journal of critical care · Mar 2024
Artificial intelligence and machine learning: Definition of terms and current concepts in critical care research.
With increasing computing power, artificial intelligence (AI) and machine learning (ML) have prospered, which facilitate the analysis of large datasets, especially those found in critical care. It is important to define these terminologies, to inform a standardized approach to critical care research. This manuscript hopes to clarify these terms with examples from medical literature. ⋯ ML, a subset of AI, is typically focused on supervised or unsupervised learning tasks in which the output is based on inputs and derived from iterative pattern recognition algorithms, while AI is the overall ability of a machine to "think" or mimic human behavior; and to analyze data free from human influence. Even with successful implementation, advanced AI and ML algorithms have faced challenges in adoption into practice, mainly due to their lack of interpretability, which hinders trust, buy-in, and engagement from clinicians. Consequently, traditional algorithms, such as linear and logistic regression, that may have reduced predictive power but are highly interpretable, continue to be widely used.
-
Journal of critical care · Mar 2024
What every intensivist should know about..Patient safety huddles in the ICU.
Patient safety huddles are brief, multidisciplinary conversations that focus on a specific topic or event. Huddles have been shown to improve communication among healthcare providers in a variety of settings, including the intensive care unit (ICU). This paper presents key features of patient safety huddles and describes the ways in which huddle techniques may be particularly relevant to the practice of critical care.