Critical care explorations最新文献

Objectives: Our aim was to describe trends in tracheostomy utilization in infants requiring congenital heart surgery (CHS) during their index admission with specific focus on clinical and financial outcomes.

Design: A retrospective cohort study.

Setting: Data were obtained from the Pediatric Health Information System database.

Patients: Patients admitted as neonates (≤ 28 d) undergoing CHS with the use of cardiopulmonary bypass (CPB) during admission from 2004 to 2022 were identified. The cohort was divided into patients with vs. without tracheostomy.

Interventions: None.

Measurements and main results: We identified 13,415 neonatal admissions who underwent CHS with use of CPB, of which 391 (3%) underwent tracheostomy. Tracheostomy patients, compared with those without, were more likely to be female (46.8% vs. 40.0%; p = 0.007), of Black race (17.1% vs. 10.6%), preterm (29.2% vs. 14.1%), low birthweight (29.4% vs. 14.1%), had a higher frequency of chromosomal defects (23.5% vs. 8%), congenital airway (24% vs. 3.3%), and pulmonary (19.7% vs. 1.7%) abnormalities (all p < 0.001). Tracheostomy was associated with higher in-hospital mortality (23.8% vs. 8.6%), longer length of stay (183 vs. 26 d), higher cost of hospitalization ($1.2 vs. $0.2 million), and discharge to a location other than home (35.1% vs. 6.3%; all p < 0.001). Tracheostomy rates increased from 1.9% in 2004-2010 to 3% in 2017-2022 (p = 0.002), while the in-hospital mortality in these patients was similar (p = 0.72).

Conclusions: The rate of tracheostomy placement in complex neonates and infants requiring CHS has increased in recent years. Patients with congenital airway or pulmonary abnormalities, cleft lip and/or palate, chromosomal disorders, and those requiring more than one surgery requiring CPB during admission were at greatest risk for tracheostomy placement. Tracheostomy is associated with longer ICU and hospital length of stay, six-fold increase in hospitalization cost, and higher rate of in-hospital mortality in our study population.

Objective: To determine the accuracy of a custom version of the generative pretrained transformer (GPT)-4o large language model (LLM) in identifying PICU admissions with vs. without bacterial pneumonia using clinical notes.

Design: In this retrospective cohort study, the GPT-4o model was provided guidance on our institution's pneumonia diagnosis practices through a custom prompt and instructed to analyze PICU provider notes from the first 2 calendar days of PICU admission to identify bacterial pneumonia diagnoses. Diagnoses from the manually curated Virtual Pediatric Systems (VPS) Registry were used as the gold standard.

Setting: A 48-bed, academic, quaternary care PICU.

Patients: Children 3 months old to 18 years old admitted to the PICU from January 1, 2023, to December 31, 2023.

Interventions: None.

Measurements and main results: GPT-4o analyzed 10,081 notes from 3,317 PICU admissions over 5.0 minutes (mean 0.03 s per note). Of the 3317 study encounters, 481(14.5%) had a VPS admission pneumonia diagnosis. GPT-4o accurately classified 3143 of 3317 (94.8%) encounters. In a post hoc adjudication analysis, a blinded PICU attending reviewed patient charts with VPS-GPT discordant classifications. The GPT-4o classification matched that of the blinded PICU attending in 125 of 174 (71.8%) of such encounters. The most common reason for incorrect classification by GPT-4o was that a pneumonia diagnosis was listed in the initial notes but later rescinded when a different diagnosis was identified.

Conclusions: The GPT-4o LLM was able to accurately and rapidly identify critically ill children with vs. without bacterial pneumonia. This study suggests similar tools could be developed to automate and accelerate processes typically requiring manual chart review.

High-flow nasal cannula (HFNC) for acute hypoxemic respiratory failure is typically restricted to ICUs. We evaluated a ward-based, medical emergency team (MET)-supervised HFNC protocol (flow ≤ 40 L/min, Fio2 ≤ 0.40) with 2-/4-/8-hour nursing and respiratory therapist reassessments. Among 82 ward HFNC initiations (2021-2024), 38 (46%) required immediate ICU transfer (IMT) and 44 (54%) were Ward-Managed After MET (WMAM). Of WMAM patients, 18 transferred to ICU within 48 hours, and 26 remained on ward. WMAM patients accrued a median 1.46 ICU bed-days saved (interquartile range, 0.73-2.67); bootstrapped mean 1.63 (95% CI, 1.32-1.94), equivalent to 163 ICU days-saved per 100 initiations. Intubation (30% vs. 42%; p = 0.24) and 28-day mortality (32% vs. 39%; p = 0.47) were similar between WMAM and IMT; adjusted analyses were directionally consistent. Using an estimated $5,000 per ICU-day, cost avoidance was ≈$815,000 per 100 initiations. This MET-supervised model appears feasible, resource-sparing, and without apparent safety signal.

Objectives: Secondary infections are a common occurrence in critically ill COVID-19 patients. These are difficult to identify, and antibiotic usage is high in this population. Identification of biomarkers for secondary infections would help to ensure antibiotics are being utilized only for patients who require them. This study sought to identify a panel of biomarkers capable of distinguishing critically ill COVID-19 patients with and without secondary infections.

Design: A multicenter retrospective cohort study.

Setting: Three critical care units in Scotland, United Kingdom.

Patients: One hundred five patients admitted to critical care with COVID-19, and 49 healthy volunteer controls.

Interventions: None.

Measurements and main results: Serial blood samples were obtained from critically ill COVID-19 patients with and without confirmed secondary infections, and a single sample was collected from healthy volunteers to provide baseline metabolic profiles. Metabolomic analysis was performed using liquid chromatography-mass spectrometry, and metabolites that were significantly different between patients with and without secondary infections were identified. Additionally, metabolites capable of distinguishing Gram-positive from Gram-negative organisms were also investigated. Forty patients developed a secondary infection during the study period. A significant increase in metabolites creatine and 2-hydroxyisovalerylcarnitine, and a significant reduction in S-methyl-L-cysteine were detected in patients with secondary infections. This metabolite panel could identify patients with secondary infections with an area under the curve (AUC) of 0.83 (95% CI, 0.68-0.97). Metabolites differentiating Gram-positive and Gram-negative infections included betaine, N(6)-methyllysine, and phosphatidylcholines (PCs; 38:6), PC(38:4), PC(40:6), and PC(36:4) with an AUC of 0.88 (95% CI, 0.68-1.0).

Conclusions: Metabolomic profiling of critically ill COVID-19 shows promise for identification of novel biomarkers for secondary infections. Larger validation studies will help to confirm these findings.

Objectives: Survival following severe acute neurologic injury (SANI) is increasing. The complexities of caring for these patients are vast and gaps have been highlighted in post-acute care follow-up. While the development of post-ICU follow-up clinics have been increasing in popularity there is limited literature describing the landscape of post-neuroICU/NeuroRecovery clinics. We sought to describe the current landscape and identify benefits and barriers to clinic development in the United States.

Design: We developed a 19-question cross-sectional survey study.

Setting and subjects: The survey was disseminated to clinicians working in neurocritical care units throughout the United States and open for completion from August 2023 to December 2023. Responses were characterized by descriptive statistics.

Interventions: None.

Measurements and main results: Two hundred eighteen unique individuals responded to our survey. Post-neuroICU and/or NeuroRecovery clinics were uncommon and operational at only 69 of 215 respondents' institutions (32.1%). Forty-two percent reported an interest in engaging with a post-neuroICU clinic and an additional 39% showed interest but had identified other obligations preventing participation. Among the identified potential benefits of a clinic for survivors of SANI mitigating gaps in care, identifying differences between predicted and actual outcome, and reassessment of communication/prognosis ranked highest.

Conclusions: Few post-neuroICU/NeuroRecovery clinics exist in the United States but interest in participating in this aspect of care is common within the neurocritical care community. The identification of gaps in care, obstacles to continued recovery, and potential to adjudicate differences between actual and predicted outcomes ranked among the most important potential benefits for extending the current neurocritical care paradigm to the clinic setting.

Importance: Current approaches for bringing critical care expertise to hospital floors to support deteriorating patients are limited by intensivist availability, costs, and scalability. Using existing tele-critical care clinicians to perform consultations on deteriorating patients at risk for ICU admission may overcome these limitations.

Objectives: To characterize clinician perspectives on tele-critical care consultation during an initial small-scale implementation.

Design, setting, and participants: An on-demand tele-critical care consultation service was implemented on selected medical and surgical floors at a large academic medical center, with options for tele-intensivist recommendations only, time-limited monitoring by tele-nurses ± intensivist co-management, or assistance with immediate ICU transfer. To characterize perspectives of both bedside and telemedicine physicians and nurses who participated in consultations, a survey including the Acceptability of Intervention Measure, Intervention Appropriateness Measure, and Feasibility of Intervention Measure was performed. Some clinicians also participated in semi-structured interviews.

Analysis: Each survey component was described using median and interquartile range. Interviews were transcribed and analyzed using a thematic analysis approach. Open coding was performed independently by two investigators, followed by identification of themes and iterative team discussion until consensus was reached.

Results: Over 1 year, 65 consultations were performed. Across 43 surveys, the median score was 4 of 5 for each of the validated measures. In the 12 interviews, consistently positive themes included excellent quality of care, support for night shift, and reassurance to patients and families. Themes with mixed feedback included effectiveness in time-critical situations, impact on clinician workload, and communication and collaboration between teams. Opportunities for improvement included increased awareness, issues related to mobile carts, and expansion throughout the hospital.

Conclusions: Telemedicine can be used to deliver critical care consultations to hospitalized floor patients who are deteriorating. Bedside and telemedicine clinicians were highly satisfied with the consultation service.

Objectives: To examine the financial burden of critical care (CC), primary cost drivers, and clinical outcomes associated with CC delivery in low- and middle-income countries (LMICs).

Perspective: CC accounts for a significant cost of global health expenditure. LMICs, where the burden of critical illness is high and access remains inequitable, data on CC's financial impact are scarce, undermining efforts to strengthen capacity, optimize delivery, and guide resource allocation.

Setting: ICU at Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife, Tertiary Hospital in Nigeria.

Methods: An observational cross-sectional study was conducted. Patients were recruited via a convenient sampling method. Data were collected within the first 24 hours of admission.

Results: A total of 96 patients were analyzed. The average daily cost of a CC bed was ₦4,780.49 (U.S. dollar [USD] 2.99) without mechanical ventilation (MV) and ₦21,255.56 (USD 13.28) with MV. The mean expenditure within the first 24 hours of admission was ₦144,928.00 (USD 90.58). The mortality rate was 23.96%, with higher ASA scores (III and IV) and age over 40 years associated with increased costs and poorer outcomes. Higher CC costs and lower household income are strongly associated with increased mortality.

Conclusions: The financial burden of CC far exceeds Nigeria's monthly minimum wage of ₦70,000.00 (USD 43.75), highlighting the urgent need for health policy and resource allocation strategies to improve timely and equitable CC outcomes in LMICs.

Objectives: Tracheal intubation in critically ill patients is associated with significant morbidity and mortality. Point-of-care ultrasound (POCUS) may help with hemodynamic optimization and customization of management plans to the patient's tenuous physiology to prevent cardiopulmonary collapse. We report the integration of POCUS in the emergency airway management (EAM) of critically ill patients at a tertiary care academic medical center.

Design: Our study is a retrospective, exploratory research project. We evaluated the feasibility of using Echocardiography Assessment using Subcostal-only-view in Physiologically Difficult Airway (EASy-PDA) protocol to prevent peri-intubation hemodynamic compromise during EAM.

Setting: This study took place at a tertiary academic medical center where requests for EAM were answered by anesthesiologists.

Subjects: The EASy-PDA protocol was performed on 30 patients with PDA outside of the operating room in need of EAM.

Interventions: The EASy-PDA protocol included the acquisition of subcostal four-chamber (SC4C) and inferior vena cava (IVC) images, supplemented by focused lung and gastric ultrasonography. Trained anesthesiology residents performed EASy-PDA examinations before airway management, and subsequently assigned hemodynamic phenotypes based on qualitative assessment of biventricular chamber size, myocardial wall thickness and function, and IVC size and collapsibility. Management was then tailored based on hemodynamic phenotyping.

Measurements and main results: The mean time to complete the EASy-PDA examination was 2.40 minutes. SC4C image could not be obtained in one patient due to severe abdominal pain. Images obtained solely via the EASy-PDA examination were sufficient to inform further patient management in 26 patients (86.7%), with one patient requiring emergent pericardial window creation and two patients requiring gastric decompression before intubation based on examination findings.

Conclusions: We were able to show the feasibility of integrating the EASy-PDA protocol into the management of emergent airways. In our case series, we observed that the EASy-PDA examination findings guided hemodynamic optimization before EAM in critically ill patients. This approach may help reduce intubation-associated morbidity and mortality. Further studies are needed to assess the impact of integration of EASy protocol during EAM on patient outcomes.

Objectives: To determine if common antimicrobials (n = 11) are sequestered or degraded during a pediatric extracorporeal membrane oxygenation (ECMO) simulation.

Design: An ex vivo model of a closed ECMO circuit was established to simulate the treatment of a 3 kg infant. A control was used to quantify spontaneous antimicrobial degradation.

Setting: University research laboratory.

Participants: None.

Main outcomes and measures: The ECMO circuit was primed and maintained at physiologic pH and temperature for 7 hours. After baseline sampling, the antimicrobials were administered as a single bolus into the circuit. Eight plasma samples were taken from the controls and ECMO circuits over 7 hours. Antimicrobial concentrations were measured using validated high-performance liquid chromatography-tandem mass spectrometry methodology. The antimicrobial recovery was compared with baseline. Each simulation was performed in triplicate to assess simulation variability.

Results: The recovery mean (%) in ECMO at 7 hours for ampicillin 78%, cefotaxime 92%, flucloxacillin 72%, meropenem 81%, micafungin 72%, piperacillin 84%, and voriconazole 42% was significantly different from the baseline (p < 0.05). The recovery in the control at 7 hours for ampicillin 83%, cefotaxime 76%, flucloxacillin 90%, gentamicin 85%, meropenem 76%, piperacillin 92%, and tazobactam 93% was also significantly different from the baseline (p < 0.05). A significant relationship was identified in the ECMO model between the antimicrobial recovery (%) and the log partition coefficient (log p) of the studied antimicrobials (R2 = 0.52; p = 0.01). No significant relationship was identified between the protein binding and antimicrobial recovery (R2 = 0.23; p = 0.13).

Conclusions and relevance: The lipophilicity of an antimicrobial is a predictor of antimicrobial recovery in ECMO. Concentrations were significantly reduced at 7 hours for greater than 60% of the study antimicrobials in the ECMO models. Clinical studies are required for children receiving ECMO to determine if the current dosing regimens for antimicrobials provide therapeutic concentrations.

ICU readmissions remain a critical concern, carrying increased morbidity, mortality, and cost. We examined the epidemiology of unplanned ICU readmissions across 19 hospitals in the Common Longitudinal ICU data Format (CLIF) Consortium from January 2020 to December 2021 and the MIMIC-IV database. The cohort included 185,241 adult ICU admissions, excluding postoperative or post-interventional procedure readmissions. Overall, 8.6% of ICU discharges were readmitted during the same hospitalization. Unplanned readmissions occurred in 1.9% of cases within 24 hours, 3.4% within 48 hours, and 4.5% within 72 hours of discharge. Readmitted patients experienced markedly higher in-hospital mortality (20.6% vs. 2.1%; p < 0.001). Compared with initial ICU stays, readmissions were associated with greater use of respiratory support (42.3% vs. 35.3%) and vasopressors (26.1% vs. 23.1%). Hospitals with stepdown units demonstrated comparable unplanned ICU readmission rates. These findings demonstrate that ICU readmissions remain common, are associated with poor outcomes, and require greater organ support. Improved characterization of high-risk subphenotypes is needed to inform safer discharge processes.