SCCM Resource Library-Acute Kidney Injury and SARS-CoV-19 in Children: Data from the National COVID Cohort Collaborative

Video Transcription

Thank you for allowing me to talk today and share the work that we've done to identify  and characterize acute kidney injury and SARS-CoV-2 in children using data from the National COVID Cohort Collaborative.  My name is Adam DeZorni, and I'm Assistant Professor of Pediatrics and Biomedical Engineering at the University of Rochester.  This work is part of a set of projects completed by Margaret Brennan, an undergraduate student  working in our lab last summer, who was unfortunately unable to present this work today. Additional collaborators are shown here.  I have no relevant financial disclosures related to this work. Much has been learned over the past two years about the association of acute kidney injury,  or AKI, with SARS-CoV-2 infection. In adults, AKI occurs frequently with severe COVID-19 disease and is associated with poor prognosis.  This figure from Hirsch et al., 2020, in the study of adults hospitalized in early 2020 in the New York metropolitan area, stratifies patient disposition by the stage of severity  of AKI from no AKI through stage 3. With worsening AKI severity, the percentage of patients discharged decreased, while the percent admitted and the percent expired increased.  Similarly, and more recently, Moldolina and colleagues reported high rates of AKI among COVID-19 adults, which were not explained by known risk factors, suggesting possibly  direct effect of COVID on the kidneys. Numerous other adult studies point to an increased frequency of AKI among critically ill adults  and suggest COVID-19 is an independent risk factor for AKI development. There are several proposed mechanisms for kidney injury in patients infected with SARS-CoV-2.  Shown here is a figure from Ng et al., 2020, which highlights some of these mechanisms. Although this study focused on adult patients, similar mechanisms are proposed in children.  The SARS-CoV-2 virus enters host cells primarily via the ACE2 receptor, which are abundant in the kidney, leading possibly to direct viral damage to the kidney cells.  The resulting microthrombi and coagulation cascade abnormalities may lead to TMA-related changes, which may be helped by anticoagulation.  Additionally, there is a suggestion that SARS-CoV-2 infections may be a trigger for new autoimmune manifestations, which may include kidney disease in at-risk people.  Additionally, the usual suspects, causes of AKI, such as hypovolemia caused by increased  insensible losses leading to a pre-renal isotemia and medication-induced injury, are both possible in the setting of SARS-CoV-2 infection.  Several studies have examined the association of AKI with SARS-CoV-2 infection in children. One of the first reported evaluations from a single center study in the UK demonstrated  29% of children hospitalized with SARS-CoV-2 infection had acute kidney injury. Since the submission of this abstract and this presentation, several additional studies  have been published reporting these associations in children. This includes the SPARC sequential point prevalence study of 57 centers, of which 42 reported  on the total of 331 patients meeting inclusion criteria. Of these, 179, or 54%, had confirmed SARS-CoV-2 infection, and of those, 74, or 41%, met CADIGO creatinine criteria for AKI.  Additionally, Rayana et al. published a meta-analysis of 24 studies looking at pediatric patients  in a hospitalized in the ICU, including a preliminary report of the study above, which reported on 1,249 confirmed COVID-positive patients, with an AKI incidence of 30.5%.  In this study, we make use of the National COVID Cohort Collaborative. This is an NIH and NCATS-supported effort, which supports the collection and harmonization  of electronic health record, or EHR, data from institutions around the country. In addition to the aggregation of data in a central repository, this collaborative manages  data access and use, coordinates the central analytics platform, and facilitates analytics groups to assist with data analysis.  The screen capture at the bottom, taken on February 23rd of this year, shows a snapshot of the current data entered in the N3C dataset.  There are currently 70 participating sites with over 4.3 million COVID-positive cases. In this study, we make use of a subset of these data, which was populated complete  at the time of abstract submission, or July of 2021. The objective of this study was to use this large, multi-institution, harmonized dataset  to determine the association of SARS-CoV-2 infection and acute kidney injury among children.  Specifically, we did not focus only on hospitalized or critically ill children, but included and stratified all children in the N3C cohort.  The study is a retrospective cohort study among the 56 participating N3 sites at the time of abstract submission.  We included children less than or equal to 18 years of age, with a positive PCR or antigen confirming SARS-CoV-2 infection, captured between March 2020 and June 2021.  Our primary outcome measure was the presence of AKI by defined CADEGO creatinine criteria.  Similar to most large, harmonized EHR datasets today, the N3C cohort does not include granular information about intake and output sufficient to define the full CADEGO criteria.  To apply these criteria require a baseline creatinine measurement, such that a change from baseline can be reported.  We used a similar method to that used in the SPARC study analyses to identify baseline creatinine.  If a subject did not have a baseline creatinine within the prior three months from documented  SARS-CoV-2 infection, we applied Haast et al.'s height-independent measure for estimating baseline creatinine.  Additional secondary outcomes applied were the severity of COVID-19 disease, which was categorized using the WHO clinical progression scale with one minor adjustment, similar to  what was done with the other N3C pediatric analyses. We separated mild into mild cases not presenting to the ED, or ambulatory cases, and mild ED  cases presenting to an emergency department but not admitted. Inpatients were considered moderate unless they met criteria for severe infection, including  non-invasive ventilation, mechanical ventilation, or ECMO support. We also separately reported on the use of vasoactive medications, mechanical ventilation,  or ECMO support in this cohort. We include the exposure to nephrotoxic medications as a covariate in these analyses.  Our list of nephrotoxic medications comes from the reference listed here, mapped to medication concepts in the N3C data model.  This first table reports on the total number of patient encounters in our cohort, a little over 90,000, separated by those meeting criteria for AKI and those not meeting criteria.  We separate these patients out by age in the grouping shown here. Each group begins just above the age listed.  For example, the one- to five-year age group includes children from one year and one day through five years.  A chi-square test of independence was performed and showed that the relation between age group  and AKI status was significant, with younger infants and adolescents having higher frequency of AKI compared to toddlers and school-age children.  Similar, we separate these patients by gender. A chi-square test of independence shows that this relation between gender and AKI status  was also significant, with males having a slightly lower frequency of AKI compared to females.  The degree of significance is likely impacted by the large sample size, since these proportions do not differ greatly in magnitude.  We separate patients by disease severity, shown here. As you recall, mild patients have ambulatory mild disease, except those who present to the ED, as indicated by mild ED.  Moderate disease indicates where inpatient hospitalization without more than nasal cannula  oxygen support, whereas severe disease indicates additional oxygen support or extracorporeal support.  Although the M3C does not include unit or department-level information, at many institutions, patients in the severe category would be managed in the pediatric intensive care unit.  We report approximately 75% of these patients with acute kidney disease. As the disease severity increases, the proportion of patients with AKI increases.  Not surprisingly, a chi-square test of independence shows that this relation between disease severity and AKI status was significant.  Additionally, we examined the relationship between other outcomes and AKI. Among patients requiring inotropic support, mechanical ventilation, and ECMO, a substantially  greater fraction met criteria for acute kidney injury. This figure shows the average or arithmetic mean creatinine value on the y-axis by hospital  day on the x-axis, separated by patients who meet and do not meet AKI criteria by status.  Comparison of creatinine value by a two-way ANOVA, including day of hospital visit following positive SARS-CoV-2 test and AKI status, was significant.  Patients meeting AKI criteria had a significantly higher creatinine value on each day of the hospital visit.  These two figures show the proportion of patients within each disease severity category by month  of the study from March 2020 through June 2021, separated by those meeting AKI criteria on the right versus those not meeting criteria on the left.  Visually, one can observe a much higher proportion of patients in the light green or mild disease  severity category across all time points among those patients not meeting AKI criteria on the left.  We hypothesize that the relative increase in mild ED, moderate, and severe patients early in the pandemic actually reflects a lack of mild or ambulatory patients, in part  due to early lack of ambulatory testing in that early phase of the pandemic when patients presented sicker for testing.  Lastly, this figure shows the count of medication administrations on the Y-axis for each of  the common drugs identified in this data set across the X-axis among hospitalized COVID-positive patients.  Among the roughly 5,000 patients who were hospitalized with SARS-CoV-2, 1,026, or 20%, received nephrotoxic medications during their hospitalization.  Among these, 877, or 85%, met criteria for AKI. This is higher than the 2,183, or 52%, of hospitalized children who met AKI criteria and did not receive nephrotoxic medication.  In summary, this project made use of a large, multi-institutional, EHR-extracted, and harmonized  data set, the National COVID Cohort Collaborative, to better understand the epidemiology and association of SARS-CoV-2 infection with acute kidney injury.  Such data sets have advantages over single-center data as well as administrative databases, which rely on coded diagnoses. However, there are limitations as well.  This data set is harmonized centrally and may include misclassifications on specific center tests.  Additionally, we are limited to mostly calculating baseline creatinine values, as most subjects did not have a pretest creatinine value present in the data set.  Using this data set, we showed that among children with SARS-CoV-2, acute kidney injury frequency was associated with worsened disease severity and was common among hospitalized  and critically ill children. We report a frequency of near 75% among hospitalized patients, higher than that reported in prior smaller studies.  Among the reasons for this difference may be in the requirement to mostly calculate baseline creatinine values in this study.  Additionally, we identified that AKI was common among inpatients exposed to nephrotoxic medications.  This finding is not surprising, given that exposure to nephrotoxic medications in the setting of other critical illnesses results in an increased risk of AKI.  Based on these results, clinicians treating pediatric patients with SARS-CoV-2 infection should be mindful of how common end-organ failure can be, and should always practice  renal-friendly management techniques. This work was made possible through the support of those listed here, as well as many others.  Specifically, I'd like to thank the Department of Pediatrics at the University of Rochester,  including Drs. Cholette and Brophy, as well as Dr. Hill in the Department of Public Health Sciences for her assistance with the N3C data set.  Additionally, I'd like to thank our external collaborators, including Tell and Bennett and Peter DeWitt at Colorado, and Rajat Basu at Northwestern Lurie Children's.  Thank you very much for your time today.

Video Summary

In this video transcript, Adam DeZorni discusses the association between acute kidney injury (AKI) and SARS-CoV-2 infection in children. He highlights that AKI occurs frequently in adults with severe COVID-19 and is associated with a poor prognosis. The proposed mechanisms for kidney injury include direct viral damage, microthrombi formation, coagulation cascade abnormalities, and autoimmune manifestations. Several studies have reported high rates of AKI among children with SARS-CoV-2 infection. DeZorni's study uses data from the National COVID Cohort Collaborative to determine the association of SARS-CoV-2 infection and AKI in children. The analysis shows that AKI is more common in younger infants and adolescents and is associated with worsened disease severity. It also indicates that AKI is common among inpatients exposed to nephrotoxic medications. Clinicians should be aware of the high incidence of AKI in pediatric COVID-19 patients and practice renal-friendly management techniques.

Asset Subtitle

Worldwide Data, Crisis Management, Pediatrics, 2022

Asset Caption

INTRODUCTION: Acute Kidney Injury (AKI) occurs frequently in adult patients with SARS-CoV-19 and confers a higher risk of mortality. Harmonized, multi-institution datasets such as the National COVID Cohort Collaborative (N3C), a repository of over 6.3 million patients from 56 sites, offer a unique opportunity to assess AKI and clinical outcomes. We sought to determine the association of SARS-CoV-19 infection and AKI among children within N3C.
METHODS: We conducted a retrospective cohort study on children ≤18 years old with SARS-CoV-2 positive PCR or antigen test. We stratified disease severity using the World Health Organization (WHO) Clinical Progression Scale. We defined AKI by KDIGO creatinine criteria and examined exposure to nephrotoxic medications. Outcomes included case incidence, mortality and length of stay. Univariate analyses compared AKI versus no AKI across severity groups.
RESULTS: As of 6/23/2021, a total of 90,169 children tested positive for SARS-CoV-2 in N3C. Among 5,193 (5.7%) hospitalized children, 3,060 (59%) met criteria for AKI. Only 69 children died, of whom 55 (80%) met AKI criteria. AKI was associated with disease severity (χ2, p < 0.001), with 457/602 (76%) of severe patients, but only 9,817/75,209 (13%) of mild patients, meeting AKI criteria. AKI was common among patients requiring vasoactive medication (375/474, 79%), mechanical ventilation (242/318, 76%) and ECMO (27/29, 93%). Length of stay was significantly longer among hospitalized children with AKI (mean [SD] 2.9 [7.3] days vs 0.5 [18] days). Only 1,026 (20%) of hospitalized children were administered nephrotoxic medications. Of these, 877 (85.5%) met criteria for AKI. The most common nephrotoxic medications given were vancomycin, ibuprofen, and ketorolac.
CONCLUSIONS: In a large multi-institutional sample of children infected with SARS-CoV-2, we demonstrate increased frequency of AKI with greater disease severity. Additionally, AKI was common (~85%) among inpatients exposed to nephrotoxic medications. Collaborative datasets such as N3C, along with associated shared data analytics tools, provide valuable resources for observational research in pediatric critical care and AKI.

Meta Tag

Content Type Presentation

Knowledge Area Pediatrics

Knowledge Area Crisis Management

Knowledge Area Worldwide Data

Knowledge Level Foundational

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Tag Epidemiology Outcomes

Tag Pediatrics

Tag COVID-19

Year 2022

Keywords

acute kidney injury

SARS-CoV-2 infection

children

COVID-19

prognosis

	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.