SCCM Resource Library-Characteristics and Outcomes of 728,047 Children in the National COVID Cohort Collaborative (N3C)

Video Transcription

Good afternoon, everyone. My name is Blake Martin, and I will be presenting for our research group our study entitled  Characteristics and Outcomes of 728,047 Children in the National COVID Cohort Collaborative. I have an asterisk there because since our initial submission, we were able to update  our study data, and it now includes over 1,068,000 patients. Very briefly, I'm an assistant professor of pediatrics and a pediatric intensivist at Children's Hospital Colorado.  My interests include the incorporation of machine learning and predictive modeling into clinical decision support within the PICU, as well as the use of large data sets to inform  our understanding of pediatric SARS-CoV-2. More importantly, though, I think that this project is an excellent example of team science.  We had an amazing group of researchers, physicians, statisticians, programmers. We had fantastic leadership in our research group's head, Dr. Tel Bennett, and really  there is just so many other people that contributed to this project not listed here.  In terms of a brief overview, we'll go through the background associated knowledge gaps motivating our study objectives. I'll discuss the N3C structure briefly.  We'll go through our study's methods and results, and then end with a summary and talk about next steps.  So the knowledge gap that we wanted to fill is that there are really not any large but also very granular pediatric SARS-CoV-2 studies.  And so this made it difficult for us as clinicians to predict which children were going to get really sick, which were going to get hospitalized.  It was difficult to track over time what medications were even being used to treat them.  We didn't have a great sense for what to expect in terms of inpatient vital sign and lab value trajectories.  And lastly, we also wanted to accomplish characterizing the differences of those children with acute  COVID-19 versus those who were diagnosed with MIS-C or multisystem inflammatory syndrome in children.  And so those knowledge gaps motivated our study objectives, which were to determine the case and hospitalization incidence over time of infected children, to determine what  were the risk factors for development of higher severity disease once a child was hospitalized with SARS-CoV-2.  We wanted to illustrate what medication and medication regimens were being used and how those were changing over time.  We wanted to identify the differences in trajectories of inpatient vital sign and laboratory values among patients of different peak clinical severity.  And lastly, we wanted to identify the key differences in the characteristics and clinical outcomes of patients with acute COVID-19 as compared to those with MIS-C or multisystem  inflammatory syndrome in children. The data set that we used for the study is the National COVID Cohort Collaborative, or N3C.  This is one of the largest limited data sets that's been put together in U.S. history and includes over 12 million patients, of which 4.3 million or so were positive.  Overall, there's 13.8 billion rows of data, which includes information on procedures, vital signs, and lab results. And this information is current as of February 25th of 2022.  Patients were included in our study if they were less than 19 years old at the time of their first SARS-CoV-2 test.  They needed to have a positive SARS-CoV-2 PCR antigen or antibody test, and their encounter date needed to be between March 1st, 2020 and September 24th of 2021.  We stratified patients by their peak clinical severity, as defined by an adapted version of the World Health Organization's Clinical Progression Scale, which is shown here.  However, because we did not have access to the exact modality of oxygen delivery nor the FiO2, we adapted this to the following severity definition.  The system we used has four discrete categories. Mild are those managed in an outpatient setting only. Mild ED includes those that at most require an emergency department visit.  Moderate are those patients that were hospitalized but did not require invasive ventilation, vasoactive, or ECMO support, whereas severe includes patients who were hospitalized and  needed any of the support modalities or who experienced mortality. In addition to traditional demographic variables, we also assessed whether one of these three  preselected comorbidities would impart increased odds for severe disease once a child was hospitalized for SARS-CoV-2.  Specifically, given reports of increased risk in adults, we looked at asthma, diabetes, both type 1 and type 2, as well as obesity.  We also looked at whether or not the presence of a pediatric complex chronic condition, or PCCC, was associated with increased risk.  This is an idea that was developed by Chris Feudner et al. and is defined as any medical condition that can be reasonably expected to last at least a year and to involve either  several different organ systems or one severely enough that it requires specialty pediatric care and probably some period of hospitalization in a tertiary care center.  So these are children that are interacting with the health care system frequently for some sort of chronic condition.  This includes many different categories, including cardiovascular, respiratory, renal, prior malignancy, et cetera.  And lastly, we assessed whether any of these demographic or comorbidity variables were associated with increased odds of severe disease via multivariable logistic regression.  Moving on to results, we identified over 1,068,000 children who were tested for SARS-CoV-2.  Among those children, there were 167,262 positive cases, so about a 15.6% positivity rate, and these children came from 56 different health care sites across the country.  There was just over 6% of these cases who end up being hospitalized, and among those  children, around 14%, 1,423, developed severe disease at some point during their hospitalization.  And so in terms of what qualified a patient as having severe disease, this included mechanical ventilation, vasoactive support, ECMO, or deaths, which thankfully were low.  Among the traditional demographic variables of age, sex, race, and ethnicity, we identified male sex, black or African-American race, and age greater than 12 as being risk factors  that were associated with an increased odds for severe disease. The odds ratio for age is 0.88 because this was the risk of a less than 12-year-old developing  severe disease compared to one over 12. Among the preselected comorbidities that we looked at, obesity, the presence of any pediatric  complex chronic condition, as well as the cardiovascular, malignancy, and technology-dependent  PCCCs were all associated with increased odds of a child developing severe disease once they were hospitalized with SARS-CoV-2.  Interestingly, the kidney PCCC was associated with decreased odds for development of severe disease.  Asthma itself did not emerge as a risk factor, and there were not enough patients with diabetes to be able to be included in the multivariable logistic regression model.  This figure demonstrates how the proportion of patients in four pre-specified age groups has changed over time.  So specifically, the black represents patients who are 0 to 1 years old, the dark blue are those 1 to 5, light blue 5 to 12, and the teal color are teenagers 12 to 18 years old.  In general, we see that there has been a subtle but statistically significant decrease in the percentage of patients falling into that 12 to 18 category over time.  Next, when you look at the white line traversing the chart there, that's the total number of cases per month that were PCR antigen positive.  And in general, we see here that the monthly trend overall matched that of adults with a peak in December of 2020.  When we look at hospitalized patients who had a positive antibody test, we see that the peak there lags a little bit, the PCR and antigen tests, with a peak in January of 2021.  Now, because testing somebody for an antibody for SARS-CoV-2 is itself kind of a surrogate  for evaluation for MIS-C, this actually matches up pretty well with the literature that demonstrates peak MIS-C risk in the 2 to 5 weeks following the initial SARS-CoV-2 infection.  These figures demonstrate how the proportion of children and adults in the four different peak clinical severity subgroups has changed over time.  So just to orient you, the light green color corresponds to patients managed in the outpatient setting. Dark green means that they needed at most an emergency department visit.  Yellow indicates that they were hospitalized, and red indicates that they were hospitalized  but also required either invasive ventilation, vasopressor, or ECMO support, or they experienced mortality.  And we have pediatric N3C patients on top and the adult patients on the bottom. In general, you can see that for the pediatric group, there's been an overall decrease in  the number of patients in those moderate and severe subgroups, those that were hospitalized. Additionally, there is overall a very close alignment of adult and pediatric trends with  the most recent peak in June of 2021. And lastly, you can just see here that compared to adults, there is overall lower pediatric  severity in terms of the percentage of patients who were hospitalized. Now these charts demonstrate the change over time of a particular medication or medication  regimen in terms of its use among hospitalized adults or children. So the solid line corresponds to inpatient pediatric patients with SARS-CoV-2, whereas  the dotted line represents the inpatient adults. So you can see here that as the study period has progressed, we've seen a decrease in the  overall percentage of hospitalized children and adults who have received a systemic antibiotic.  Additionally, we've seen that for systemic corticosteroids writ large, as well as dexamethasone  specifically, we've seen an increase in their use among hospitalized children and adults.  Now these red arrows here correspond to two specific studies, the first being the dexamethasone in hospitalized patients with COVID-19 by the Recovery Collaborative Group, and the  second being a landmark study describing MIS-C. And so after the publication of both of these in July of 2020, we did see an uptick in the  use of systemic corticosteroids and dexamethasone specifically, though a little bit more notable in the adults than in the pediatric group.  So now we're getting to what I think is the coolest part of this study, which is how the vital sign and lab values changed over time during a patient's hospitalization.  So just to orient you on the x-axis, we have the day of hospitalization, so day zero is the day they were admitted.  On the y-axis, we have the median value for that particular vital sign across all patients in the different subgroups.  And in terms of the subgroups, the solid line corresponds to pediatric patients, and the dotted lines correspond to adult patients.  Yellow are those that were hospitalized, and red are those that were hospitalized but experienced  the highest clinical severity, again, ventilation, ECMO, vasopressors, or mortality. And what we see here is that for the vast majority of these vital signs, there are subtle  but statistically significantly different median values on that day of hospitalization, day zero.  So the heart rate, the respiratory rate are higher, the oxygen saturation, the systolic blood pressure are lower among those children who went on to develop severe disease compared  to those that stayed in the moderate severity subgroup. Along those same lines, we saw the same phenomenon among children with respect to their lab values by day of hospitalization.  So on day zero, the day of admission, we saw increased markers of organ dysfunction, specifically higher levels of ALT, creatinine, and troponin among those children who went  on to develop the highest clinical severity compared to those who remained in the moderate severity subgroup. The same is true when you look at markers of inflammation.  So for C-reactive protein, ferritin, and procalcitonin, children who went on to develop severe disease  had higher levels on day zero compared to those who remained in the moderate severity subgroup.  And we also see that cell line abnormalities were also more pronounced in the pediatric severe subgroup, specifically hemoglobin and absolute lymphocyte count being lower and  platelet count being lower as well. The last objective of our study was to try to understand the differences in the clinical  characteristics and comorbidities and clinical outcomes of patients that were diagnosed with  MIS-C as compared to those who were treated as acute COVID-19. And by MIS-C, I mean they had to have some sort of positive SARS-CoV-2 test and they  also had to have one of two diagnosis codes present in their chart, either M35.8 used during the early part of the pandemic or M35.81, which became effective January 1st of 2021.  And by acute COVID-19, we mean those patients who had either a positive PCR antigen test  but did not have a positive antibody test and did not have an MIS-C diagnosis code. Overall, we had 707 patients fit into the MIS-C group and just over 8,200 patients in  the acute COVID-19 group. In general, you see here that MIS-C patients are a little bit younger with higher percentages  of children in the one to five and five to 12-year-old age group as compared to those in the acute COVID-19 group. Additionally, you had more patients who were male, 59% versus 48%.  And we also, within the MIS-C group, saw a higher percentage of black or African-American children. Here, you can see the odds ratios for those risk factors.  Again, patients who developed MIS-C compared to acute COVID-19 were more likely to be younger, have male sex, and have a black or African-American race.  When we looked at specific preselected comorbidities, we found that patients with obesity were more  likely to have MIS-C compared to acute COVID-19, whereas patients with asthma were less likely to receive an MIS-C diagnosis.  Additionally, when we looked at pediatrics complex chronic conditions, we saw that the presence of any PCCC and for the vast majority of PCCCs, that patients were more likely to  have PCCC in the acute COVID-19 group, but less likely in the MIS-C group. So in general, apart from obesity, these children were actually more likely to be otherwise  healthy and not have most of these pediatric complex chronic conditions. Lastly, looking at the laboratory profiles and clinical outcomes, we see that patients  in the MIS-C group, there was a higher percentage of children who had at least one abnormal lab value for each of the 17 preselected labs compared to acute COVID-19.  So you see here, 72% of children had an elevated ESR in the MIS-C group compared to just 8% in the acute COVID-19 group.  And that higher proportion of patients with an abnormal lab value was present for each of those 17 labs listed.  In terms of clinical outcomes, we saw that a higher proportion, 37% versus 10% of MIS-C patients developed severe disease and that a higher proportion of patients required invasive  ventilation and vasoactive isotropic support as compared to the acute COVID-19 group. So 27% over one in four children with an MIS-C diagnosis required a vasoactive at  some point in their hospital stay. So in summary, in this study, we have been able to describe the clinical characteristics and outcomes of children infected with SARS-CoV-2.  We have been able to demonstrate several demographic and comorbidity risk factors that were associated  with development of severe disease once a child was hospitalized with SARS-CoV-2. We were also to illustrate the changes in age and severity distribution over time, as  well as the changes in medication regimen utilization over time during the study period. We were able to identify key differences on the day of hospitalization for both vital  signs and lab values that differed between children of peak clinical severity subgroups.  And lastly, we were able to identify the key differences in the characteristics, risk factors,  and the clinical phenotype and clinical outcomes of children with MIS-C compared to those with acute COVID-19.  In terms of next steps, the head of our research group, Dr. Tel Bennett, has a grant from the NICHD to develop a pediatric COVID-19 dashboard.  And this tool will take data from the N3C and present illustrations of changes in pediatric SARS-CoV-2 outcomes over time.  Additionally, I think the main takeaway from this study is that we have several key data elements available on the day of hospital admission that might allow a team of clinicians  to predict which patients would go on to develop severe disease. Specifically, we could put together a clinical decision support system that is able to incorporate  these demographic comorbidity risk factors, as well as the day zero vital sign and lab values, and be able to predict and alert clinicians to which children are at the highest risk  of disease and deterioration. Many thanks to the N3C leads who helped support this project, Melissa Handel, Chris Schutt, and Ken Gersing.  And with that, I'm happy to take any questions.

Video Summary

Blake Martin, an assistant professor of pediatrics and a pediatric intensivist at Children's Hospital Colorado, presented a study entitled "Characteristics and Outcomes of 728,047 Children in the National COVID Cohort Collaborative." The study aimed to fill the knowledge gap regarding pediatric SARS-CoV-2 and aimed to determine the case and hospitalization incidence over time, identify risk factors for severe disease, analyze medication regimens, track vital sign and lab value trajectories, and compare acute COVID-19 with multisystem inflammatory syndrome in children (MIS-C). The study used the National COVID Cohort Collaborative dataset, which includes over 12 million patients. Results showed that male sex, black or African-American race, and age greater than 12 were risk factors for severe disease. Obesity, any pediatric complex chronic condition, and several PCCCs were associated with increased odds of severe disease. MIS-C patients were younger, more likely to be male and black or African-American, and had higher rates of abnormal lab values and severe disease compared to acute COVID-19 patients. The study's findings could help clinicians predict which children are at highest risk of disease and deterioration.

Asset Subtitle

Worldwide Data, Infection, Pediatrics, 2022

Asset Caption

INTRODUCTION: SARS-CoV-2 can cause severe pediatric disease via acute COVID-19 and multisystem inflammatory syndrome in children (MIS-C). We sought to determine the characteristics, changes over time, outcomes, and severity risk factors of SARS-CoV-2 infected children within the National COVID Cohort Collaborative (N3C).
METHODS: We performed an analysis of an EHR-based registry of children < 19-years-old at initial SARS-CoV-2 testing at 45 N3C institutions with encounter end dates before 5/27/2021. Outcomes included case incidence and severity over time, risk factors for higher severity disease, vital sign and lab trajectories, clinical outcomes, and acute COVID-19 vs. MIS-C differences for children with SARS-CoV-2.
RESULTS: 728,047 children were tested for SARS-CoV-2 and 91,865 (12.6%) were positive. Of 5,213 (6%) hospitalized children, 685 (13%) met criteria for severe disease: invasive ventilation (7%), vasoactive-inotropic support (7%), ECMO (0.6%), or death/discharge to hospice (1.1%). Male sex, African American race, older age, and most pediatric complex chronic condition (PCCC) categories were associated with higher peak clinical severity (p≤0.05). Vital signs (all p≤0.002) and many lab tests from the day of hospital admission were predictive of peak disease severity. Children with severe (vs. moderate) disease were more likely to be treated with antimicrobial (71% vs. 32%, p < 0.001) and immunomodulatory (53% vs. 16%, p < 0.001) medications. Compared to children with acute COVID-19, those with MIS-C were more likely to be male, African American, 1-to-12-years-old, and less likely to have asthma, diabetes, or a PCCC (p0.04). Children with MIS-C demonstrated a more inflammatory lab profile and severe clinical phenotype with higher rates of invasive ventilation (12% vs. 6%) and vasoactive-inotropic support (31% vs. 6%) compared to those with acute COVID-19, respectively (p < 0.03).
CONCLUSIONS: In the largest U.S. SARS-CoV-2-positive pediatric cohort to date, we identified differences in demographics, comorbidities, and initial vital sign and laboratory values between severity subgroups, indicating that early identification of children likely to progress to severe disease might be feasible using readily available data from the day of admission. Further work is needed to translate this knowledge into improved outcomes.

Meta Tag

Content Type Presentation

Knowledge Area Infection

Knowledge Area Pediatrics

Knowledge Area Worldwide Data

Knowledge Level Advanced

Membership Level Select

Tag Epidemiology Outcomes

Tag Pediatrics

Tag COVID-19

Tag Infectious Diseases

Year 2022

Keywords

pediatrics

National COVID Cohort Collaborative

severe disease

MIS-C

risk factors

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.