Critical care explorations最新文献

Cefepime, a cornerstone antibiotic in critical care, is associated with underrecognized cefepime-induced neurotoxicity (CIN), particularly in patients 65 years old and older. The true incidence is unknown due to inconsistent monitoring and a lack of diagnostic criteria. The recent Antibiotic Choice on Renal Outcomes (ACORN) trial underscored CIN's clinical significance, finding that cefepime recipients experienced 21% fewer delirium- and coma-free days than those on piperacillin-tazobactam. Current guidelines lack active surveillance recommendations, leading to delayed diagnosis and intervention. We propose three informatics-based strategies to address these challenges: 1) electronic health record (EHR)-integrated datasets utilizing machine learning and natural language processing to identify CIN at scale, 2) automated electroencephalogram tools to provide real-time alerts to clinicians, and 3) dynamic risk scores that continuously update from EHR data to guide prescribing. Implementing these safeguards to optimize CIN prevention, which may be relevant for other antibiotics with neurotoxicity risk, can improve neurologic outcomes and patient safety in critically ill populations.

Background: Adverse events (AEs) within the cardiac ICU (CICU) are associated with high mortality and comorbidity. Patients with single ventricle (SV) physiology are particularly vulnerable to experiencing such events before second stage surgery (bidirectional Glenn). Timely identification and management of AEs are critical for improving patient outcomes and enabling earlier, more targeted medical interventions in this vulnerable population.

Objectives: To develop and evaluate machine learning (ML) models using continuous physiologic data to predict and identify AEs including cardiac arrest (CA), extracorporeal membrane oxygenation (ECMO) cannulation, and endotracheal intubation, up to 8 hours before occurrence in SV infants.

Derivation cohort: Retrospective cohort of 158 SV patients (324 admissions) admitted to the tertiary care CICU at Children's National Hospital.

Validation cohort: Internal validation occurred in the held-out testing group of the data (10% of the total dataset).

Prediction model: Supervised ML classifiers were (e.g., decision tree, logistic regression, support vector machine, extreme gradient boosting, random forest [RF]) trained and compared across four observation windows (8, 4, 2, and 1 hr) preceding each event. Model Performance was evaluated using the area under the receiver operating characteristic curve (AUROC). Fifteen physiologic and laboratory variables were extracted, of which six high-quality variables were included in the ML model. AEs were categorized into four categories (e.g., intubation, CA, ECMO, no event) for multiclass classification and ECMO-CA were also combined into a single class (ECMO-CA) to improve stability for rare events.

Results: A total of 256 AEs were analyzed: 157 intubations (61.328%), 42 ECMO events (16.406%), 44 CAs (17.187%), and 13 extracorporeal cardiopulmonary resuscitation events (5.078%). Across all time windows, the RF model achieved the best performance on the held-out test set for AE detection (AUROCs 0.998 at 8 hr, 0.996 at 4 hr, 0.996 at 2 hr, and 0.997 at 1 hr). For the combined-class multiclass classification RF model at 1-hour observation window, the held-out test set results showed AUROCs of 0.819 for intubation, 0.804 for ECMO-CA, and 0.840 for no-event prediction.

Conclusions: A ML model can predict and discriminate between several types of AEs in SV infants before bidirectional Glenn. Accurate predictions may help perform timely interventions, potentially reducing morbidity, mortality, and healthcare costs.

Context: Management of ventilator and venovenous extracorporeal membrane oxygenation (ECMO) settings in critically ill adults requires individualized decisions to balance oxygenation, ventilation, and complication risks. Existing approaches rely heavily on clinician experience, with limited decision support tools available.

Hypothesis: An offline reinforcement learning agent trained on real-world venovenous ECMO clinical data can generate safe, interpretable, and clinically aligned recommendations for ECMO and ventilator management, including support for earlier and more efficient weaning.

Methods and models: We conducted a retrospective study using electronic health record data from 184 adult patients who underwent venovenous ECMO at a tertiary care center. rECMOmender was developed using conservative Q-learning with a physiologically informed reward structure. Multiple model variants were compared across discrete and continuous action spaces and two reward formulations. Performance was assessed using fitted Q evaluation, comparison of action distributions, and alignment with clinician practice.

Results: rECMOmender generated stable, interpretable recommendations across five key parameters: Fio2, positive end-expiratory pressure (PEEP), respiratory rate, sweep gas flow, and blood flow rate. It selected Fio2 values in the 40-50% range most frequently (46.75% vs. 45.65% for clinicians) and favored PEEP of 9-11 cm H2O (43.94% vs. 34.28%), while using high PEEP settings (13-20 cm H2O) 73.43% less often. Compared with clinicians, rECMOmender increased large parameter shifts (> 1 bin) by 72.53% for Fio2, 348.15% for PEEP, 299.21% for respiratory rate, 96.68% for sweep gas, and 34.16% for blood flow, resulting in an overall 120.37% increase in major adjustments (3105 vs. 1409).

Interpretations and conclusions: rECMOmender demonstrated dynamic but safety conscious adjustments that aligned with clinical patterns, indicating potential as a decision support tool that complements clinician judgment.

Importance: Delirium is a neuropsychiatric syndrome characterized by fluctuating disturbances in attention and awareness, associated with worse clinical outcomes and higher mortality. Previous research studies have noted an association between geographic disadvantage and delirium, but it is unknown if this association extends to critically ill adults.

Objectives: This study aimed to explore the relationship between geographic disadvantage and ICU delirium characteristics.

Design, setting, and participants: We performed a secondary analysis of data collected from an National Institutes of Health-funded clinical trial, the Pharmacologic Management of Delirium study. Adults 18 years old or older admitted to the ICU who experienced delirium based on the Confusion Assessment Method for the ICU (CAM-ICU) were included.

Interventions: None.

Measurements and main results: The study population included 326 participants: 54.5% were female and 48% Black, with a mean age of 60.3 years, mean Acute Physiology and Chronic Health Evaluation II score of 20, and in-hospital mortality rate of 12.3%. The area deprivation index (ADI), a composite measure of geographic disadvantage derived from census data that yields a national percentile score ranging from 1 to 100 (with higher scores representing greater disadvantage), was obtained for each participant's address. Main outcome variables included delirium duration, which was assessed by the number of delirium- and coma-free days (DCFDs), and delirium severity, which was assessed by mean CAM-ICU-7 scores. Analysis of covariance models were used to examine differences in DCFDs and mean CAM-ICU-7 scores between ADI quintiles while controlling for demographic and clinical variables. Other clinical outcomes of interest included discharge home rates and in-hospital mortality. The sample was heavily skewed toward higher national ADI percentile scores (indicating greater disadvantage); only 11.7% of patients had an ADI score lower than 50. Our regression analyses did not reveal any associations between ADI quintile and DCFDs or mean CAM-ICU-7 scores, or between ADI quintile and discharge home rates or in-hospital mortality. However, the Black race was associated with longer delirium duration and greater delirium severity in the first week of ICU hospitalization.

Conclusions: Our study did not find an association between geographic disadvantage and delirium duration or severity in the ICU. However, an association with race was observed, highlighting the need for further research into how socioeconomic determinants of health relate to delirium.

Importance: The optimal level of positive end-expiratory pressure (PEEP) during venovenous extracorporeal membrane oxygenation (ECMO) for acute respiratory distress syndrome (ARDS) remains uncertain.

Objectives: This study aimed to evaluate the association between initial PEEP settings at ECMO initiation and the rate of successful ECMO liberation in patients with severe ARDS.

Design, setting, and participants: We conducted a post hoc analysis of the multicenter Japan Chest CT for ARDS Requiring Venovenous ECMO (J-CARVE) registry. Adult patients with severe ARDS treated with venovenous ECMO between 2012 and 2022 at 24 institutions were included. Participants were categorized into three groups according to PEEP at ECMO initiation: low (< 8 cm H2O), middle (8-10 cm H2O), and high (> 10 cm H2O).

Main outcomes and measures: The primary outcome was successful liberation from ECMO within 30 days. Multivariable Cox proportional hazards models were used to evaluate associations. Secondary outcomes included 60-day mortality, duration of ECMO support, and duration of mechanical ventilation.

Results: Among 683 patients analyzed, the overall ECMO liberation rate at 30 days was 69.2%. Liberation rates were 57.8% (103/178), 73.5% (259/352), and 72.5% (111/153) in the low, middle, and high PEEP groups, respectively. After adjustment, the low group had a significantly lower likelihood of successful ECMO liberation (hazard ratio [HR], 0.56; 95% CI, 0.39-0.81) compared with the middle group. No significant difference was observed between the high and middle groups (HR, 0.80; 95% CI, 0.58-1.10). The low group had longer ECMO duration; however, 60-day mortality and hospital length of stay did not differ significantly among groups.

Conclusions and relevance: Lower PEEP levels at ECMO initiation were associated with reduced likelihood of successful ECMO liberation compared with moderate PEEP, whereas estimates for high vs. moderate PEEP were not statistically significant. These findings support avoiding insufficiently low PEEP and underscore the need for prospective studies to refine optimal PEEP strategies in patients with severe ARDS.

Background: Previously healthy children are at risk of developing exertional heat stroke when experiencing extreme heat. Pediatric clinicians in primary care, emergency department, and critical care settings should be versed in the management of complications of exertional heat stroke. Pediatric acute liver failure (PALF) in the setting of heat stroke is rarely reported in published literature.

Case summary: A 9-year-old male presented with heat stroke-induced PALF. He initially presented to an emergency department for altered mental status. During his clinical course, despite appropriate identification and initial treatment of exertional heat stroke, his symptoms progressed, including ongoing agitation, hepatic encephalopathy, coagulopathy, and severe transaminase elevation meeting clinical criteria for PALF.

Conclusions: He was treated with N-acetylcysteine (NAC) with resolution of his PALF without complications. In this article, we review the patient's clinical course, the rationale for treatment with NAC, and the management of heat stroke-induced PALF.

Importance: Delirium is prevalent and associated with poorer clinical outcomes in critically ill children.

Objectives: We sought to determine whether presence of baseline developmental delay (DD) or a primary admission diagnosis of an acute neurologic condition (ANC) was associated with longer delirium duration and/or time to ICU discharge, and whether delirium acts as a mediator among observed outcome associations.

Design setting and participants: Post hoc analysis of a prospective, observational study in patients 6 months to 5 years old admitted to a tertiary PICU regardless of admission diagnosis.

Main outcomes and measures: Patients assessed daily for delirium using the Pediatric Confusion Assessment Method for the ICU series (PEDs CAM-ICU). Baseline demographics and in-hospital outcomes obtained.

Results: Of 282 patients, 79 had baseline DD and 54 were admitted with an ANC. Delirium prevalence among patients with DD was 53% and 43% in those with an ANC. DD was associated with significantly longer delirium duration (p = 0.008), with a meaningful association between ANC and delirium duration (p = 0.056). DD was significantly associated with a lower likelihood of ICU discharge (hazard ratio, HR, 0.76 [95% CI, 0.54-0.95]), with delirium partially mediating this relationship. Patients with ANC diagnoses trended toward a relevant association with lower likelihood of ICU discharge (HR 0.73 [0.53-1.00]) with partial delirium mediation.

Conclusions and relevance: Baseline DD among critically ill infants and children is linked to longer delirium duration and lower likelihood of ICU discharge. An innovative finding is that delirium mediates this relationship. Although ANCs were loosely correlated with both prolonged delirium duration and decreased likelihood of ICU discharge, the true impact of delirium on these patients warrants further investigation. Finally, a focus on how to mitigate the impact of DD (predisposing risk factor) on ICU delirium and outcomes in critically ill infants and children is paramount.

Importance: Viral bronchiolitis is the leading cause of nonelective admission to the PICU. Guidelines recommend management that focuses on supportive care. Evidence suggests that inhaled epinephrine may reduce lower airways resistance and work of breathing in critical bronchiolitis. It has been demonstrated at one institution that it is feasible to administer continuous inhaled epinephrine (CIE) via heated high-flow nasal cannula (HFNC).

Objectives: To describe the demographic and clinical characteristics of patients with bronchiolitis who received CIE and assess their physiologic response to treatment compared with patients who received conventional therapy.

Design setting and participants: We performed a single-center retrospective cohort study of previously healthy children 0-2 years old admitted to the PICU with diagnosis of bronchiolitis from 2017 to 2023.

Main outcomes and measures: Physiologic response to treatment was assessed via analysis of vital signs and use of respiratory rate, oxygenation, heart rate (ROX-HR) index.

Results: One hundred ninety-three patients were included in the study. Patients who received CIE were younger (median age, 4 vs. 7 mo old; p = 0.01) and smaller (7.0 vs. 8.4 kg; p = 0.06), presented with higher initial peak respiratory rates (70 vs. 50 beats/min; p < 0.0001), and had longer ICU stays (3.4 vs. 1.8 d; p < 0.001). Patients treated with CIE exhibited overall higher respiratory rates and lower ROX-HR indices, indicating more severe illness. Within the first 48 hours of treatment, patients receiving CIE demonstrated a more rapid improvement in both respiratory rate and ROX-HR index compared with those receiving conventional therapy.

Conclusions and relevance: In this cohort, CIE was used as novel therapy for younger, smaller patients with more severe bronchiolitis on presentation to the PICU. The physiologic changes suggest potential benefit of this therapy for patients who are not improving with conventional HFNC therapy.

Objectives: Given significant harms of awake paralysis, guidelines recommend combination sedation and analgesia during neuromuscular blockade. In this study, we sought to describe guideline-concordant use of parenteral sedation and analgesia during sustained neuromuscular blockade among adults with critical illness.

Design: Multicenter, retrospective cohort study.

Setting: ICUs in 871 hospitals across the United States.

Patients: Adult patients admitted to an ICU between 2016 and 2022 who received invasive mechanical ventilation and at least one neuromuscular blocking agent (NMBA) on the same calendar day.

Interventions: None.

Measurements and main results: We used charge codes to identify use of parenteral sedatives (propofol, fospropofol, lorazepam, midazolam, diazepam, etomidate, phenobarbital, pentobarbital, and ketamine), and analgesics (fentanyl, hydromorphone, morphine, and ketamine) for each patient-day with neuromuscular blockade, and then categorized patient-days as guideline concordant or not. A total of 363,382 patient-days (among 104,984 hospitalizations) were included in the final cohort. Guideline-concordant sedation and analgesia were used concurrently in 345,660 patient-days (95.1%); only sedation was used in 15,618 patient-days (4.3%), only analgesia was used in 1,348 patient-days (0.4%), and neither sedation nor analgesia was used in 756 patient-days (0.2%). Most included hospitals (856 [98.3%]) used both sedation and analgesia on greater than or equal to 50% of patient-days; however, ten hospitals (1.1%) used only sedation on greater than or equal to 50% of patient-days, and 4 (0.6%) had no predominant sedation and analgesia strategy; however, 42.3% of the variation in guideline-concordant practice was attributable to residual unexplained clustering by hospital after accounting for demographics, severity of illness, and hospital characteristics.

Conclusions: Our findings suggest sedation and analgesia practices with NMBA use in adult ICUs are generally guideline-concordant but require corroboration using more precise, quantitative medication data. Practice variation between hospitals is potentially concerning and warrants further investigation targeting adequacy of sedation and analgesia during NMBA use and assessing the clinical impact of guideline-discordant practices.

Importance: Black and Hispanic/Latino patients are underrepresented in pediatric mode of death (MOD) studies. Although significant disparities have been reported, the associations of MOD with patient-level social determinants of health (SDOH) and the Child Opportunity Index (COI) are unknown.

Objectives: To investigate associations between PICU MOD, race and ethnicity, SDOH, and COI.

Design, setting, and participants: Retrospective, single-center cohort study at a twenty-four-bed PICU within a large safety-net, public hospital, including all PICU deaths between January 2010 and December 2019.

Main outcomes and measures: We examined MOD by race and ethnicity, COI, and SDOH, including preferred language, health insurance, single-parent household status, parental occupation, and healthcare barriers abstracted from medical records. MOD was categorized as limitation of artificial life-sustaining therapies/technology (LOT) or withdrawal of artificial life-sustaining therapies/technology (WOT), failed resuscitation (FR), and death by neurologic criteria (DNC).

Results: Of 238 deaths, Black non-Hispanic/Latino patients comprised 42% (n = 100), Hispanic/Latino patients of all races 35% (n = 83), and White non-Hispanic/Latino patients 19% (n = 46). LOT/WOT was the predominant MOD (75%, n = 174; LOT = 109, WOT = 65). Racial and ethnic groups showed significant differences in COI, SDOH, and healthcare barriers. Despite this, there were no significant differences in MOD by Race and Ethnicity, SDOH, or healthcare barriers. Median COI was lower for DNC compared to LOT and WOT, and for FR compared with WOT. However, when examined within individual racial and ethnic groups, there was no difference in median COI between FR, LOT, and WOT.

Conclusion and relevance: We found no differences in MOD by Race and Ethnicity, SDOH, or barriers. Median COI was lower for FR compared with WOT. This suggests that COI, as opposed to race and ethnicity, may play a role in pursuing or forgoing resuscitation at end-of-life. This study adds to the examination of pediatric healthcare disparities at end-of-life by including SDOH and COI data in MOD analysis.