Pediatric Critical Care Medicine最新文献

Objectives: To compare the proportion of PICU patients returning to the emergency department (ED) or readmitted within 14 days of hospital discharge, between those discharged directly home from the PICU and those transferred to acute care before discharge home; we hypothesized that rates of return-to-care would be similar.

Design: Propensity-matched multicenter cohort study.

Setting: Forty-five U.S. hospitals participating in Pediatric Health Information Systems.

Patients: Children admitted to a non-neonatal cardiac or PICU from 2016 to 2023.

Interventions: None.

Measurements and main results: Of 560,815 PICU discharges, 150,126 (26.8%) were discharged directly home, although this proportion varied by center (9.8-55.6%). We matched 94,048 children (62.6%) discharged directly home to 153,887 ward-transferred children at admission year, admission type, principal diagnosis, and a propensity score estimating the likelihood of being discharged directly home. Compared with ward-transferred peers, children discharged directly home had similar rates of return-to-ED care (2.9% vs. 3.0%; odds ratio [OR], 0.94 [0.89-0.99]) and hospital readmission (4.8% vs. 4.9%; OR, 0.97 [0.94-1.01]) within 14 days. Once readmitted, however, children discharged directly home were more likely to be readmitted to a PICU (2.4% vs. 1.6%; OR, 1.58 [1.49-1.67]). Costs for the index hospitalization were lower for children discharged directly home compared with ward-transferred peers, leading to lower inpatient healthcare costs over 14 days (median, 15,023 [7,614.5-34,294.6] vs. 30,750 [14,558.3-68,830.6]; p ≤ 0.001).

Conclusions: Discharge directly home from the PICU is common; children discharged directly home have comparable likelihood of return-to-ED or inpatient care as matched, ward-discharged peers. Discharge directly home for appropriate patients may provide increased efficiency for healthcare systems.

Objectives: To use a multidimensional outcomes portfolio to assess neurodevelopmental sequelae among PICU survivors during the first 3 years after hospital discharge.

Design: Prospective study of a cohort recruited 2017-2018, with interval follow-up for 3 years.

Setting: PICU at an urban academic tertiary care center.

Patients: Children 0-17 years admitted to the PICU with anticipated discharge home.

Interventions: None.

Measurements and main results: We evaluated outcomes using a PICU Outcomes Portfolio (POP) survey, which combined a study-specific Healthcare and Neurodevelopmental Profile and the Family Impact Survey and standardized measurement tools, including the Pediatric Quality of Life Inventory, the Strengths and Difficulties Questionnaire, and the National Institute for Children's Health Quality Vanderbilt Assessment Scales, to identify various components of post-PICU challenges. Our POP survey identified a sustained impact of child health on family finances and parental employment. Our multidimensional outcomes assessment flagged more at-risk children than individual measures of neurodevelopmental functioning.

Conclusions: Children and families face diverse challenges during recovery from critical illness. Parent-reported outcomes and a multidimensional outcomes portfolio identify the broad impact of critical illness on family well-being as well as the long-term outcomes among PICU survivors. Future mixed-methods studies incorporating parental input regarding post-discharge needs are needed to enrich the evaluation of post-PICU outcomes using standardized measures and guide the development of post-PICU follow-up programs.

Objectives: To investigate whether the urine output trajectory is associated with dialysis independence in critically ill children with acute kidney injury (AKI).

Design: Retrospective cohort study.

Setting: A PICU in Japan.

Patients: Children younger than 16 years old who received continuous kidney replacement therapy (CKRT) for AKI between July 1, 2014, and June 30, 2023.

Interventions: None.

Measurements and main results: We identified 61 eligible patients, including 16 patients who remained dialysis-dependent 30 days after CKRT initiation. Compared with dialysis-independent patients, dialysis dependence was associated with lower urine output on days 3, 7, 14, and 21 after CKRT initiation. Dialysis independence, when compared with dialysis dependence, was associated with higher median (interquartile range) urine output (mL/kg/hr) at each timepoint (day 3: 0.3 [0.1-1.6] vs. 0.0 [0-0.2]; p = 0.001; day 7: 1.3 [0.4-2.0] vs. 0.0 [0-0.1]; p < 0.001; day 14: 1.8 [1.0-3.5] vs. 0.0 [0-0; p < 0.001]; and day 21: 2.1 [1.1-3.0] vs. 0.0 [0-0]; p < 0.001). The area under the receiver operating characteristic curve (AUROC with 95% CI) for identifying dialysis independence at day 30 after CKRT initiation, based on urine output on day 14, was 0.96 (95% CI, 0.88-1.00). Using the DeLong test, this AUROC was higher than that on day 7 (0.88 [95% CI, 0.77-0.99]; p = 0.009). Also, on day 14, with a pre-test probability of dialysis independence of 71%, the post-test probability increases to 97% when using a test urine output greater than or equal to 0.41 mL/kg/hr. The sensitivity analysis with the exclusion of neonates yielded similar results.

Conclusions: In this 2014-2023 cohort of critically ill children with AKI supported with CKRT, using a urine output greater than or equal to 0.41 mL/kg/hr on day 14, CKRT may be an effective diagnostic test of dialysis independence on day 30. Further validation studies are needed.

Objective: To develop a set of pediatric neurocritical care (PNCC) entrustable professional activities (EPAs) for pediatric critical care medicine (PCCM).

Design: Survey and Delphi methodology in a panel of experts from the Pediatric Neurocritical Care Research Group (PNCRG) and the Education in Pediatric Intensive Care (EPIC) Research Collaborative.

Setting: Interprofessional local focus group, national focus group, and subsequent national multi-institutional, multidisciplinary expert panel in the United States.

Subjects: The interprofessional local group of 23 members carried out work March 2022 to June 2022 and the national group of 19 members October 2022 to November 2022. Subsequently, 38 physicians from the PNCRG and EPIC networks carried out work December 2022 to August 2024.

Interventions: None.

Measurements and main results: First, a preliminary set of 15 PNCC EPAs was developed by two local and national interprofessional groups. The EPAs were based on the American Board of Pediatrics (ABP) practice analysis for PCCM, the ABP PCCM content outline, and stakeholder opinion. Next, a panel of critical care, neurology, and education experts used Delphi methodology to generate consensus, edit, and finalize the EPAs, with content validity. All EPAs were edited; two were deemed non-essential and not included in the final set of 13 EPAs. The EPAs fit three categories: general management and principles; disease-specific management; and neuroprognostication and end-of-life care. Consensus was reached after three Delphi rounds, with response rates of 31 of 38, 29 of 31, and 29 of 31, respectively. The final set of EPAs was approved by 30 respondents (response rate 30/31), with content validity indices 0.81-1.00.

Conclusions: The 2024 set of 13 EPAs are intended to be a valuable framework for competency-based curriculum and assessment to ensure consistent PCCM proficiency in the provision of neurocritical care while also promoting standardization in curriculum development for PCCM fellowship trainees.

Objectives: To identify factors associated with death, requirement for extracorporeal membrane oxygenation (ECMO), or cardiac intervention in neonates referred for higher level neonatal ICU (NICU) due to respiratory failure.

Design: Retrospective cohort study, 2018-2020.

Setting: Referrals for transport to tertiary-level NICUs using the London Neonatal Transfer Service in the United Kingdom.

Patients: Neonates with a diagnosis of severe respiratory failure who were intubated and receiving F io2 greater than 60% at referral. We excluded neonates younger than 34 weeks corrected gestational age, less than 2 kg, or with a known cardiac diagnosis.

Interventions: None.

Measurements and main results: We identified 170 neonates with a median (interquartile range [IQR]) age of 4 hours (2-11 hr) at referral and 9 hours (IQR, 7-16 hr) at time of departure from the referring NICU. Overall, 21 of 170 babies required immediate transfer to a center providing ECMO, of whom two of 21 died and two of 21 received ECMO support. Of the 149 transferred to NICUs that do not provide ECMO, 11 of 149 died (7%) and a further 16 of 149 (11%) required secondary transfers to an ECMO center where one of 16 died and three of 16 required ECMO. In total, there were 23 of 170 neonates with outcome of death, need for ECMO, or cardiac intervention. A composite score of Vasoactive-Inotropic Score (VIS) greater than or equal to 18 and oxygenation index (OI) greater than 24 after patient stabilization was associated with death, a need for ECMO or a previously undiagnosed cardiac lesion, with a sensitivity of 83% and a specificity of 73%.

Conclusions: This 2018-2020 cohort of neonates with severe respiratory failure managed in London, United Kingdom, shows that VIS greater than or equal to 18 and OI greater than 24 after stabilization were associated with death, need for ECMO, or cardiac intervention. These parameters may have the potential to trigger discussion with ECMO centers for early consideration of transfer but needs validation in a wider neonatal population.

Objectives: In children with sepsis, thrombocytopenia is linked to poor outcomes, including longer hospital length of stay, increased bleeding risk, and higher mortality. However, the clinical significance of changes in platelet counts over time remain poorly characterized. We have examined dynamic platelet patterns and their association with mortality and patient-illness factors.

Design: Single-center retrospective cohort study.

Setting: Hospital pediatrics and PICU at Guangdong Women and Children Hospital, China.

Patients: Children with sepsis between January 2015 and December 2023.

Interventions: None.

Measurements and main results: Group-based trajectory analysis was used to examine the trend in platelet count during the first 7 days of hospital admission. A regression model was developed to investigate possible associations between patient characteristics with platelet count trajectory. Additionally, a multivariable Cox proportional hazards model, adjusted for age, sex, comorbidities, and site/source of infection, was constructed to evaluate the association between platelet count trajectories and 28-day mortality. Among 1010 children with sepsis, we identified three platelet count trajectories with distinct characteristics. Age, fibrinogen level, activated partial thromboplastin time, and lactic acid were each associated with platelet count trajectories. The overall 28-day mortality for the cohort was 5.4%, varying across groups: 1.2% in group 3 with persistently high platelet count; 2.2% in group 2 with high-normal platelet count; and 12.6% in group 1 with low platelet count. In the multivariable Cox proportional hazards model, compared with group 1, both groups 2 and 3 were independently associated with reduced hazard of death at 28 days (hazard ratio, 0.26; p < 0.001 for group 2 and hazard ratio, 0.18; p = 0.021 for group 3).

Conclusions: We have identified three distinct and clinically relevant platelet count trajectories in children with sepsis, which serve as robust associations with survival in this patient population.

Objectives: Many PICU admissions start with presentation in the emergency department (ED). However, we do not know whether there are any ED-related factors associated with the subsequent development of pediatric delirium (PD) within 24 hours of PICU admission.

Design: Retrospective cohort study.

Setting: Single-center ED and PICU serving a quaternary referral center in the United States.

Patients: Children younger than 18 years old presenting to the ED between January 2022 and December 2023 who required direct admission to the PICU, and who had at least one positive delirium screening within 24 hours of the admission.

Interventions: None.

Measurements and main results: The presence of delirium in the PICU was defined as a Cornell Assessment of Pediatric Delirium (CAPD) score of 9 and higher. We identified 138 patients for the final analysis. Overall, 51 of 138 patients (37%) developed PD within 24 hours of admission. The majority of delirium cases were classified as hypoactive (53%, 27/51) and mixed (31%, 16/51) subtypes, while only 16% (8/51) were identified as hyperactive delirium. Factors associated with greater odds (95% CI) of delirium in the PICU in multivariable analysis were use of mechanical ventilation (odds ratio [OR], 3.42 [95% CI, 1.09-10.78]; p = 0.04) and intermittent urinary catheterization (OR, 3.7 [95% CI, 1.21-11.30]; p = 0.02). Initial CAPD score positively correlated with PICU length of stay (LOS; r = 0.32; p < 0.01), Pediatric Index of Mortality 3 (PIM 3) score ( r = 0.26; p < 0.01), and negatively correlated with emergency severity index (ESI) in the ED ( r = -0.35; p < 0.01).

Conclusions: Mechanical ventilation and intermittent urinary catheterization in the ED are associated with greater odds of PD within 24 hours of PICU admission. CAPD at admission positively correlates with PICU LOS, PIM 3 scores, and negatively correlates with ESI in the ED.