The role of machine learning applications in diagnosing and assessing critical and non-critical CHD: a scoping review.

Machine learning uses historical data to make predictions about new data. It has been frequently applied in healthcare to optimise diagnostic classification through discovery of hidden patterns in data that may not be obvious to clinicians. Congenital Heart Defect (CHD) machine learning research entails one of the most promising clinical applications, in which timely and accurate diagnosis is essential. The objective of this scoping review is to summarise the application and clinical utility of machine learning techniques used in paediatric cardiology research, specifically focusing on approaches aiming to optimise diagnosis and assessment of underlying CHD. Out of 50 full-text articles identified between 2015 and 2021, 40% focused on optimising the diagnosis and assessment of CHD. Deep learning and support vector machine were the most commonly used algorithms, accounting for an overall diagnostic accuracy > 0.80. Clinical applications primarily focused on the classification of auscultatory heart sounds, transthoracic echocardiograms, and cardiac MRIs. The range of these applications and directions of future research are discussed in this scoping review.

Hospital-Acquired Infection in Pediatric Subjects With Congenital Heart Disease Postcardiotomy Supported on Extracorporeal Membrane Oxygenation.

OBJECTIVE: To determine prevalence of and risk factors for infection in pediatric subjects with congenital heart disease status postcardiotomy supported on extracorporeal membrane oxygenation, as well as outcomes of these subjects. DESIGN: Retrospective cohort from the Extracorporeal Life Support Organization. SETTING: U.S. and international medical centers providing care to children with congenital heart disease status postcardiotomy. PATIENTS: Critically ill pediatric subjects less than 8 years old admitted to medical centers between January 1, 2013, and December 31, 2015, who underwent cardiac surgery for congenital heart disease and required extracorporeal membrane oxygenation support within the first 14 postoperative days. Subjects were excluded if they underwent orthotopic heart transplantation, required preoperative extracorporeal membrane oxygenation, and had more than one postoperative extracorporeal membrane oxygenation run. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A total of 1,314 extracorporeal membrane oxygenation subject encounters in the Extracorporeal Life Support Organization registry met inclusion criteria. Neonates comprised 53% (n = 696) of the cohort, whereas infants made up 33% (n = 435). Of the 994 subjects with Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery categorizable surgery, 33% (n = 325) were in Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery category 4 and 23% (n = 231) in Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery category 5. While on extracorporeal membrane oxygenation, 229 subjects (17%) acquired one or more extracorporeal membrane oxygenation-related infections, which represents an occurrence rate of 67 infections per 1,000 extracorporeal membrane oxygenation days. Gram-negative (62%) and Gram-positive (42%) infections occurred most commonly. Forty percent had positive blood cultures. Infants and children were at higher infection risk compared with neonatal subjects; subjects undergoing less complex surgery had higher infection rates. Unadjusted survival to hospital discharge was lower in infected subjects compared with noninfected subjects (43% vs 51%; p = 0.01). After adjusting for confounders via propensity matching, we identified no significant mortality difference between infected and noninfected subjects. CONCLUSIONS: Neonatal and pediatric subjects in this study have a high rate of acquired infection. Infants and children were at higher infection risk compared with neonatal subjects. There was not, however, a significant association between extracorporeal membrane oxygenation-related infection and survival to hospital discharge after propensity matching.

A National Survey on Physician Trainee Participation in Pediatric Interfacility Transport.

OBJECTIVES: To ascertain the national experience regarding which physician trainees are allowed to participate in pediatric interfacility transports and what is considered adequate education and training for physician trainees prior to participating in the transport of children. DESIGN: Self-administered electronic survey. SETTING: Pediatric transport teams listed with the American Academy of Pediatrics Section on Transport Medicine. SUBJECTS: Leaders of U.S. pediatric transport teams. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Forty-four of the 90 U.S. teams surveyed (49%) responded. Thirty-nine (89%) were university hospital-affiliated. Most programs (26/43, 60%) allowed trainees to participate in pediatric transport in some capacity. Mandatory transport rotations were reported for pediatric critical care (PICU) fellows (9/42, 21%), neonatology (neonatal ICU) fellows (6/42, 14%), pediatric emergency medicine fellows (4/41, 10%), emergency medicine residents (3/43, 7%), and pediatric residents (2/43, 5%). Fellow participation was reported by 19 of 28 programs (68%) with PICU fellowships, 12 of 25 programs (48%) with pediatric emergency medicine fellowships, and 10 of 34 programs (29%) with neonatal ICU fellowships. Transport programs with greater than or equal to 1,000 annual incoming transports were more likely to include PICU and pediatric emergency medicine fellows as providers (p = 0.04; 95% CI, 1.04-25.71 and p = 0.02; 95% CI, 1.31-53.75). Most commonly, trainees functioned as medical control physicians (86%), provided minute-to-minute medical direction for critically ill patients (62%), performed intubations (52%), and were code leaders for patients undergoing cardiopulmonary resuscitation during transport (52%). Most transport programs required pediatric residents, PICU, and pediatric emergency medicine fellows to complete a PICU rotation prior to participating in pediatric transports. The majority of transport programs did not use any metrics to determine airway proficiency of physician trainees. CONCLUSIONS: There is heterogeneity with regard to the types of physician trainees allowed to participate in pediatric interfacility transports, the roles played by physician trainees during pediatric transport, and the training (or lack thereof) provided to physician trainees prior to their participating in pediatric transports.

Feasibility and Perceptions of PICU Diaries.

OBJECTIVES: To determine the feasibility of implementing an ICU diary in the pediatric critical care setting and to understand the perceptions held by family members who receive the diaries after PICU discharge. DESIGN: Observational pilot study. SETTING: PICU in a tertiary academic hospital in the United States. PARTICIPANTS: Critically ill pediatric patients admitted to the PICU and their families. INTERVENTIONS: The addition of a PICU diary to a patient's routine care. MEASUREMENTS AND MAIN RESULTS: Twenty families of critically ill children admitted to the PICU were enrolled in the PICU diary pilot study between May 2017 and March 2018. Patients who had an anticipated length of stay of at least 3 days and whose families were English-speaking were included. The median age of patients was 6 years, ranging from newborns to 18 years old, and the median length of stay was 11.5 days (interquartile range, 8.5-41 d). A total of 453 diary entries were written in 19 diaries over 433 PICU days, the majority of which were composed by bedsides nurses (63%). Follow-up surveys sent to parents 2 weeks after PICU discharge revealed that of the parents who had contributed to the diary, most enjoyed doing so (7/8). Nine of 12 parents had reviewed the diary at least once since discharge, and all parent respondents found the diary to be a beneficial aspect of their experience after PICU discharge. CONCLUSIONS: The use of ICU diaries in the PICU setting is feasible and perceived as beneficial by families of critically ill children. Future studies are needed to better understand if PICU diaries may objectively improve psychologic outcomes of patients and family members after PICU admission.

Characteristics of postintensive care syndrome in survivors of pediatric critical illness: A systematic review.

AIM: To synthesize the available evidence focusing on morbidities in pediatric survivors of critical illness that fall within the defined construct of postintensive care syndrome (PICS) in adults, including physical, neurocognitive and psychological morbidities. METHODS: A comprehensive search was conducted in MEDLINE, EMBASE, the Cochrane Library, PsycINFO, and CINAHL using controlled vocabulary and key word terms to identify studies reporting characteristics of PICS in pediatric intensive care unit (PICU) patients. Two reviewers independently screened all titles and abstracts and performed data extraction. From the 3176 articles identified in the search, 252 abstracts were identified for full text review and nineteen were identified for inclusion in the review. All studies reporting characteristics of PICS in PICU patients were included in the final synthesis. RESULTS: Nineteen studies meeting inclusion criteria published between 1995 and 2016 were identified and categorized into studies reporting morbidities in each of three categories-physical, neurocognitive and psychological. The majority of included articles reported prospective cohort studies, and there was significant variability in the outcome measures utilized. A synthesis of the studies indicate that morbidities encompassing PICS are well-described in children who have survived critical illness, often resolving over time. Risk factors for development of these morbidities include younger age, lower socioeconomic status, increased number of invasive procedures or interventions, type of illness, and increased benzodiazepine and narcotic administration. CONCLUSION: PICS-related morbidities impact a significant proportion of children discharged from PICUs. In order to further define PICS in children, more research is needed using standardized tools to better understand the scope and natural history of morbidities after hospital discharge. Improving our understanding of physical, neurocognitive, and psychological morbidities after critical illness in the pediatric population is imperative for designing interventions to improve long-term outcomes in PICU patients.

MyD88-dependent and MyD88-independent pathways in synergy, priming, and tolerance between TLR agonists.

TLRs sense components of microorganisms and are critical host mediators of inflammation during infection. Different TLR agonists can profoundly alter inflammatory effects of one another, and studies suggest that the sequence of exposure to TLR agonists may importantly impact on responses during infection. We tested the hypothesis that synergy, priming, and tolerance between TLR agonists follow a pattern that can be predicted based on differential engagement of the MyD88-dependent (D) and the MyD88-independent (I) intracellular signaling pathways. Inflammatory effects of combinations of D and I pathway agonists were quantified in vivo and in vitro. Experiments used several D-specific agonists, an I-specific agonist (poly(I:C)), and LPS, which acts through both the D and I pathways. D-specific agonists included: peptidoglycan-associated lipoprotein, Pam3Cys, flagellin, and CpG DNA, which act through TLR2 (peptidoglycan-associated lipoprotein and Pam3Cys), TLR5, and TLR9, respectively. D and I agonists were markedly synergistic in inducing cytokine production in vivo in mice. All of the D-specific agonists were synergistic with poly(I:C) in vitro in inducing TNF and IL-6 production by mouse bone marrow-derived macrophages. Pretreatment of bone marrow-derived macrophages with poly(I:C) led to a primed response to subsequent D-specific agonists and vice versa, as indicated by increased cytokine production, and increased NF-kappaB translocation. Pretreatment with a D-specific agonist augmented LPS-induced IFN-beta production. All D-specific agonists induced tolerance to one another. Thus, under the conditions studied here, simultaneous and sequential activation of both the D and I pathways causes synergy and priming, respectively, and tolerance is induced by agonists that act through the same pathway.