Identification of prognostic factors for pediatric myocarditis with a random forests algorithm-assisted approach.

Background: Pediatric myocarditis is a rare disease with substantial mortality. Little is known regarding its prognostic factors. We hypothesize that certain comorbidities and procedural needs may increase risks of poor outcomes. This study aims to identify prognostic factors for pediatric myocardit...

Descripción completa

Detalles Bibliográficos
Publicado en:Pediatric Research Vol. 90; no. 2; pp. 427 - 431
Autores principales: Chou, Fu-Sheng, Ghimire, Laxmi V.
Formato: research tables/charts Journal Article
Publicado: Springer Nature Aug2021
Acceso en línea:Ver este registro en EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=152603936&site=ehost-live
header:
  @attributes:
    shortDbName: ccm
    uiTerm: 152603936
    longDbName: CINAHL Complete
    uiTag: AN
  controlInfo:
    bkinfo:
    dissinfo:
    jinfo:
      jid:
        00313998
        4FF
      jtl: Pediatric Research
      issn: 00313998
      maglogo: N
    pubinfo:
      dt: Aug2021
      vid: 90
      iid: 2
      pid: 237
      pub: Springer Nature
      place: New York, New York
    artinfo:
      ui:
        152603936
        152603936
        NLM33208880
        152603936
        10.1038/s41390-020-01268-7
        NLM33208880
        152603936
      ppf: 427
      ppct: 4
      formats:
      tig:
        atl: Identification of prognostic factors for pediatric myocarditis with a random forests algorithm-assisted approach.
      aug:
        au:
          Chou, Fu-Sheng
          Ghimire, Laxmi V.
        affil: Department of Pediatrics, Loma Linda University School of Medicine, 92354, Loma Linda, CA, USA
      sug:
        subj:
          Myocardial Diseases Epidemiology
          Risk Factors
          Risk Assessment
          Comorbidity
          Time Factors
          Prognosis
          Myocardial Diseases Diagnosis
          Resource Databases
          Data Mining
          Prevalence
          Myocardial Diseases Therapy
          Age of Onset
          United States
          Human
          Length of Stay
          Myocardial Diseases Mortality
          Comparative Studies
          Multicenter Studies
          Evaluation Research
      ab: Background: Pediatric myocarditis is a rare disease with substantial mortality. Little is known regarding its prognostic factors. We hypothesize that certain comorbidities and procedural needs may increase risks of poor outcomes. This study aims to identify prognostic factors for pediatric myocarditis.Methods: The national Kids' Inpatient Database was used in the study. A random forests algorithm was implemented for mortality prediction based on comorbidities and procedures. Linear regression analysis was then performed to quantify their associations with mortality and length of stay.Results: The prevalence of pediatric myocarditis among all pediatric hospitalizations doubled from 2003 to 2016. The mortality rate peaked in 2006 (6.7%) and declined steadily thereafter, with a rate of 3.2% in 2016. Brain injury (including encephalopathy, cerebral edema, and intracranial hemorrhage), acute kidney injury, dysrhythmias, coagulopathy, sepsis, and ECMO use were all independent prognostic factors associated with increased mortality and prolonged hospital stay.Conclusion: Prognostic factor identification may not be straightforward in rare diseases such as pediatric myocarditis due to small cohort size in each treating facility. Findings from this report provide insights into the prognostic factors for pediatric myocarditis, and may allow clinicians to be better prepared when informing patients and their families regarding disease outcomes.Impact: The rate of hospitalization due to pediatric myocarditis was increasing but the mortality rate was declining over the past decade. End organ damage, including the brain and the kidney, was associated with mortality and prolonged hospital stay in pediatric myocarditis. Tachyarrhythmias and cardiac function compromise requiring ECMO were also associated with mortality and prolonged hospital stay. A data science approach combining machine learning algorithms and conventional regression modeling using a large dataset may facilitate risk factor identification and outcome correlation in rare diseases, as illustrated in this study.
      pubtype: Academic Journal
      doctype:
        research
        tables/charts
        Journal Article
      ougenre: Article
    language: English
    refInfo:
    holdings:
      @attributes:
        islocal: N