Development and validation of a federated learning framework for detection of subphenotypes of multisystem inflammatory syndrome in children (preprint)

Background: Multisystem inflammatory syndrome in children (MIS-C) is a severe post-acute sequela of SARS-CoV-2 infection. The highly diverse clinical features of MIS-C necessities characterizing its features by subphenotypes for improved recognition and treatment. However, jointly identifying subphenotypes in multi-site settings can be challenging. We propose a distributed multi-site latent class analysis (dMLCA) approach to jointly learn MIS-C subphenotypes using data across multiple institutions. Methods We used data from the electronic health records (EHR) systems across nine U.S. childrens hospitals. Among the 3,549,894 patients, we extracted 864 patients < 21 years of age who had received a diagnosis of MIS-C during an inpatient stay or up to one day before admission. Using MIS-C conditions, laboratory results, and procedure information as input features for the patients, we applied our dMLCA algorithm and identified three MIS-C subphenotypes. As validation, we characterized and compared more granular features across subphenotypes. To evaluate the specificity of the identified subphenotypes, we further compared them with the general subphenotypes identified in the COVID-19 infected patients. Findings Subphenotype 1 (46.1%) represents patients with a mild manifestation of MIS-C not requiring intensive care, with minimal cardiac involvement. Subphenotype 2 (25.3%) is associated with a high risk of shock, cardiac and renal involvement, and an intermediate risk of respiratory symptoms. Subphenotype 3 (28.6%) represents patients requiring intensive care, with a high risk of shock and cardiac involvement, accompanied by a high risk of >4 organ system being impacted. Importantly, for hospital-specific clinical decision-making, our algorithm also revealed a substantial heterogeneity in relative proportions of these three subtypes across hospitals. Properly accounting for such heterogeneity can lead to accurate characterization of the subphenotypes at the patient-level. Interpretation Our identified three MIS-C subphenotypes have profound implications for personalized treatment strategies, potentially influencing clinical outcomes. Further, the proposed algorithm facilitates federated subphenotyping while accounting for the heterogeneity across hospitals.

Understanding pediatric long COVID using a tree-based scan statistic approach: An EHR-based cohort study from the RECOVER Program (preprint)

Objectives: Post-acute sequalae of SARS-CoV-2 infection (PASC) is not well defined in pediatrics given its heterogeneity of presentation and severity in this population. The aim of this study is to use novel methods that rely on data mining approaches rather than clinical experience to detect signals associated with PASC. Materials and Methods We used a propensity-matched cohort design comparing children identified using the new PASC ICD10CM diagnosis code (U09.9) (N=1250) to children with (N=6250) and without (N=6250) SARS-CoV-2 infection. We used a tree-based scan statistic to identify potential condition clusters co-occurring more frequently in cases than controls. Results We found significant enrichment among children with PASC in cardiac, respiratory, neurologic, psychological, endocrine, gastrointestinal, and musculoskeletal systems, the most significant related to circulatory and respiratory such as dyspnea, difficulty breathing, and fatigue and malaise. Discussion Our study addresses methodological limitations of prior studies that rely on pre-specified clusters of potential PASC-associated diagnoses driven by clinician experience. Future studies are needed to identify patterns of diagnoses and their associations to derive clinical phenotypes. Conclusion We identified multiple conditions and body systems associated with pediatric PASC. Because we rely on a data-driven approach, several new or under-reported conditions and symptoms were detected that warrant further investigation.

Leveraging serology testing to identify children at risk for post-acute sequelae of SARS-CoV-2 infection: An EHR-based cohort study from the RECOVER program (preprint)

ABSTRACT The impact of post-acute sequelae of SARS-CoV-2 infection (PASC) in children is underrecognized. We developed an EHR-based algorithm across eight pediatric institutions to identify children with COVID-19 based on serology testing from 3/2020 through 4/2022 who had not been identified by PCR. Overall, serology tests were used 100-fold less than PCR. Seroprevalence of IgG anti-nucleocapsid antibodies remained stable, while rates of positive IgG anti-spike antibodies increased in teenagers after COVID-19 vaccine approval. Through data harmonization and after excluding 1,410 serology test results that may have been influenced by vaccines, we identified 2,714 children that were COVID-19 positive exclusively by serology. These patients were frequently tested as inpatients (24% vs. 2%), had chronic conditions more frequently (37% vs 24%), and a MIS-C diagnosis (23% vs. <1%) compared with PCR-positive children. Identification of children that could have been paucisymptomatic, not tested, or missed is critical to define the burden of PASC in children.

Clinical features and burden of post-acute sequelae of SARS-CoV-2 infection in children and adolescents: an exploratory EHR-based cohort study from the RECOVER program (preprint)

Importance The post-acute sequelae of SARS-CoV-2 (PASC) has emerged as a long-term complication in adults, but current understanding of the clinical presentation of PASC in children is limited. Objective To identify diagnosed symptoms, diagnosed health conditions and medications associated with PASC in children. Design, Setting and Participants Retrospective cohort study using electronic health records from 9 US children’s hospitals for individuals <21 years-old who underwent reverse transcriptase polymerase chain reaction (RT-PCR) testing for SARS-CoV-2 between March 1, 2020 – October 31, 2021 and had at least 1 encounter in the 3 years before testing. Exposure SARS-CoV-2 PCR positivity. Main Outcomes and Measures We identified syndromic (symptoms), systemic (conditions), and medication PASC features in the 28-179 days following the initial test date. Adjusted hazard ratios (aHRs) were obtained for 151 clinically predicted PASC features by contrasting PCR-positive with PCR-negative groups using proportional hazards models, adjusting for site, age, sex, testing location, race/ethnicity, and time-period of cohort entrance. We estimated the incidence proportion for any syndromic, systemic or medication PASC feature in the two groups to obtain a burden of PASC estimate. Results Among 659,286 children in the study sample, 59,893 (9.1%) tested positive by PCR for SARS-CoV-2. Most were tested in outpatient testing facility (50.3%) or office (24.6%) settings. The most common syndromic, systemic, and medication features were loss of taste or smell (aHR 1.96 [95% CI 1.16-3.32), myocarditis (aHR 3.10 [95% CI 1.94-4.96]), and cough and cold preparations (aHR 1.52 [95% CI 1.18-1.96]). The incidence of at least one systemic/syndromic/medication feature of PASC was 41.9% among PCR-positive children versus 38.2% among PCR-negative children, with an incidence proportion difference of 3.7% (95% CI 3.2-4.2%). A higher strength of association for PASC was identified in those cared for in the ICU during the acute illness phase, children less than 5 years-old, and individuals with complex chronic conditions. Conclusions and Relevance In this large-scale, exploratory study, the burden of pediatric PASC that presented to health systems was low. Myocarditis was the most commonly diagnosed PASC-associated condition. Acute illness severity, young age, and comorbid complex chronic disease increased the risk of PASC. Key Points Question What are the incidence and clinical features of post-acute sequelae of SARS-CoV-2 infection (PASC) in children? Findings In this retrospective cohort study of 659,286 children tested for SARS-CoV-2 by polymerase chain reaction (PCR), the symptom, condition and medication with the strongest associations with SARS-CoV-2 infection were loss of taste/smell, myocarditis, and cough and cold preparations. The incidence proportion of non-MIS-C related PASC in the PCR-positive group exceeded the PCR-negative group by 3.7% (95% CI 3.2-4.2), with increased rates associated with acute illness severity, young age, and medical complexity. Meaning PASC in children appears to be uncommon, with features that differ from adults.