American Journal of Critical Care最新文献

Artificial intelligence technology has emerged rapidly and is being incorporated into the health care delivery system as a new bedside tool. Artificial intelligence has the potential to be paradigm-changing; however, critical care clinicians have a long history of adopting new technologies that transform care delivery in remarkable ways. Just as introduction of the mechanical ventilator revolutionized life support, artificial intelligence promises to transform modern critical care by improving prediction of disease course and patient outcomes, supporting clinical decision-making, and optimizing clinician workflow, with future potential for autonomous therapeutic adjustments and personalized care. However, as with mechanical ventilation, adopting artificial intelligence as a new technology requires caution and rigorous validation. Key challenges include ensuring accountability, preventing algorithmic bias, maintaining patient safety, and preserving the human element in care. It is vital to maintain a human-centered approach, in which artificial intelligence serves as a tool to augment, not replace, the nuanced judgment of health care professionals. To navigate this transition, critical care clinicians must become literate in artificial intelligence, understanding the capabilities and limitations of these new tools. It is crucial to cultivate a critical mindset, continuously validating artificial intelligence outputs against clinical judgment. Integrating artificial intelligence into team workflows, developing clear ethical guidelines, and fostering collaboration between clinicians and data scientists are essential for successful implementation. By proactively preparing for the transition, the critical care community can harness artificial intelligence's power to improve patient recovery and survival while ensuring that technology remains guided by human expertise and compassion.

Background: Delirium affects many critically ill patients receiving mechanical ventilation and is an independent predictor of death, length of stay, cost of care, and acquired dementia. More evidence is needed for nonpharmacological interventions that reduce delirium in patients receiving mechanical ventilation in intensive care units (ICUs).

Objectives: A structured intervention, Family Automated Voice Recording (FAVoR), used recorded voices of family members to provide patients receiving mechanical ventilation with hourly reorientation to the ICU environment during daytime hours. The primary aim was to compare the effect of the FAVoR intervention vs usual care on delirium in adults receiving mechanical ventilation in the ICU.

Methods: This prospective, 2-arm, blinded randomized controlled trial included 178 adults receiving mechanical ventilation in 9 ICUs at 2 large hospitals in south Florida. Delirium was measured with the Confusion Assessment Method for the ICU, administered by study personnel twice daily for 7 days or until ICU discharge. Data analyses included descriptive statistics, χ2 tests, and multivariable modeling analysis following the intent-to-treat principle.

Results: Clinical characteristics and demographics were similar between groups. Patients in the FAVoR group (n = 89) had more delirium-free days than did those in the usual-care group (n = 89) (P < .001). Response to the intervention was dose dependent; more doses of intervention were associated with less delirium (P < .001).

Conclusions: The FAVoR intervention is a nonpharmacological, low-resource-using intervention to reorient ICU patients receiving mechanical ventilation. In this trial, FAVoR was effective in preventing delirium among these patients. ClinicalTrials.gov identifier: NCT03128671.

Background: The ABCDEF bundle improves outcomes of patients in the intensive care unit, but its adoption remains low worldwide due to multiple implementation barriers. Interprofessional simulation training may help facilitate ABCDEF bundle implementation.

Objectives: To evaluate the feasibility of an interprofessional simulation-based training to facilitate ABCDEF bundle implementation.

Methods: A 90-minute interprofessional training program was designed and implemented. Each simulation started with a deeply sedated patient receiving mechanical ventilation, and interprofessional learners assessed sedation and delirium, conducted spontaneous awakening trials, managed delirium and agitation using multicomponent nonpharmacologic interventions, and mobilized the patient out of bed. Feasibility of the training was evaluated by using a participant survey on simulation experiences and measuring participants' attitudes toward ABCDEF bundle implementation before and after training.

Results: Sixteen clinicians from different disciplines and with various years of intensive care unit experience participated. Most participants agreed that the simulation was realistic, relevant, important to practice, and stress provoking and would improve their clinical performance. After the training, participants reported increased comfort and willingness to perform spontaneous awakening trials, comfort in managing patients with delirium and agitation, and confidence in collaborating with other team members to mobilize patients receiving mechanical ventilation out of bed.

Conclusions: The interprofessional simulation training was feasible and well accepted and increased positive attitudes toward ABCDEF bundle implementation.

This article is the second in a 2-part series examining the role of commercial and food-based enteral formulas in critical illness. Part I (published in the September 2025 issue of AJCC) reviewed the historical evolution, administration, and nutritional content of formulas; part II focuses on clinical outcomes and future directions. Enteral feeding is essential in the intensive care unit to support patients who cannot meet nutritional needs orally. Although commercial formulas remain the standard of care, the use of food-based formulas is expanding. In this part of the review, we examine the evidence comparing commercial versus food-based formulas across gastrointestinal symptoms (eg, constipation, diarrhea, reflux), renal effects (eg, electrolyte disturbances, uremia), and endocrine effects (eg, hyperglycemia, insulin resistance). We also explore nonclinical outcomes such as patient satisfaction, circadian rhythm effects, environmental considerations, and access disparities. Notably, much of the current evidence arises from pediatric or outpatient settings, underscoring the need for high-quality research in intensive care unit populations. These evolving patterns highlight critical gaps in knowledge that must be addressed to optimize intensive care unit nutrition practices. Together with part I, this article offers a comprehensive overview to guide evidence-based selection and implementation of enteral nutrition in critical care.

Background: Children undergoing extracorporeal membrane oxygenation (ECMO) are at risk of hematologic and coagulation complications, including intracranial hemorrhage. Thromboelastography is used to monitor anticoagulation in patients receiving ECMO, but no data for its use in children are available.

Objective: To evaluate the incidence of complications when managing anticoagulation with thromboelastography or with conventional coagulation tests (activated clotting time, plasma fibrinogen level, and platelet count) among children receiving venoarterial or venovenous ECMO.

Methods: This single-center, prospective, retrospective, observational study included patients less than 18 years old who required ECMO. Outcomes were compared between a conventional test group and a thromboelastography group. The primary outcome was critical hemorrhage or pump failure leading to treatment termination.

Results: Each group included 17 patients. Twenty-one patients (17 in the conventional test group, 4 in the thromboelastography group) were enrolled retrospectively. Thirteen patients (all in the thromboelastography group) were enrolled prospectively beginning December 26, 2022. Patient age did not differ between groups (median [IQR] age, 2 [0-6] months). Significantly fewer critical complications occurred in the thromboelastography group (11%) than in the conventional test group (59%) (odds ratio, 10.71; 95% CI, 1.84-62.5; P = .01). The 30-day survival rate after ECMO was higher in the thromboelastography group than in the conventional test group, but the difference was not significant (P = .08).

Conclusions: Managing anticoagulation with thromboelastography, as compared with conventional tests, decreased the frequency of critical complications among children undergoing ECMO.