Development of an adaptive clinical web-based prediction tool for kidney replacement therapy in children with chronic kidney disease

CKiD investigators

doi:10.1016/j.kint.2023.06.020

Development of an adaptive clinical web-based prediction tool for kidney replacement therapy in children with chronic kidney disease

CKiD investigators

Pediatrics

Research output: Contribution to journal › Article › peer-review

4 Scopus citations

Abstract

Clinicians need improved prediction models to estimate time to kidney replacement therapy (KRT) for children with chronic kidney disease (CKD). Here, we aimed to develop and validate a prediction tool based on common clinical variables for time to KRT in children using statistical learning methods and design a corresponding online calculator for clinical use. Among 890 children with CKD in the Chronic Kidney Disease in Children (CKiD) study, 172 variables related to sociodemographics, kidney/cardiovascular health, and therapy use, including longitudinal changes over one year were evaluated as candidate predictors in a random survival forest for time to KRT. An elementary model was specified with diagnosis, estimated glomerular filtration rate and proteinuria as predictors and then random survival forest identified nine additional candidate predictors for further evaluation. Best subset selection using these nine additional candidate predictors yielded an enriched model additionally based on blood pressure, change in estimated glomerular filtration rate over one year, anemia, albumin, chloride and bicarbonate. Four additional partially enriched models were constructed for clinical situations with incomplete data. Models performed well in cross-validation, and the elementary model was then externally validated using data from a European pediatric CKD cohort. A corresponding user-friendly online tool was developed for clinicians. Thus, our clinical prediction tool for time to KRT in children was developed in a large, representative pediatric CKD cohort with an exhaustive evaluation of potential predictors and supervised statistical learning methods. While our models performed well internally and externally, further external validation of enriched models is needed.

Original language	English (US)
Pages (from-to)	985-994
Number of pages	10
Journal	Kidney international
Volume	104
Issue number	5
DOIs	https://doi.org/10.1016/j.kint.2023.06.020
State	Published - Nov 2023

Funding

Data in this paper were collected by the Chronic Kidney Disease in Children (CKiD) prospective cohort study with clinical coordinating centers (principal investigators) at Children's Mercy Hospital and the University of Missouri—Kansas City (BAW) and Children's Hospital of Philadelphia (SLF), Central Biochemistry Laboratory (GJS) at the University of Rochester Medical Center, and data coordinating center (Alvaro Muñoz and DKN) at the Johns Hopkins Bloomberg School of Public Health. The CKiD study is funded by the National Institute of Diabetes and Digestive and Kidney Diseases, with additional funding from the National Institute of Child Health and Human Development, and the National Heart, Lung, and Blood Institute (U01-DK-66143, U01-DK-66174, U24-DK-082194, U24-DK-66116). FMW received support from the NIH T32 institutional training grant (T32 HL007024). The CKID website is located at https://www.statepi.jhsph.edu/ckid, and a list of CKiD collaborators can be found at https://statepi.jhsph.edu/ckid/site-investigators/ and is presented in Supplementary Table S6. The authors acknowledge Drs. Alvaro Muñoz and Christopher Cox for critical insight and feedback on cross-validation methods. The authors deeply appreciate Dr. Franz Schaefer for providing permission to use ESCAPE study data used in the paper by Furth et al.4 for the external validation analysis. Data in this paper were collected by the Chronic Kidney Disease in Children (CKiD) prospective cohort study with clinical coordinating centers (principal investigators) at Children’s Mercy Hospital and the University of Missouri—Kansas City (BAW) and Children’s Hospital of Philadelphia (SLF), Central Biochemistry Laboratory (GJS) at the University of Rochester Medical Center, and data coordinating center (Alvaro Muñoz and DKN) at the Johns Hopkins Bloomberg School of Public Health. The CKiD study is funded by the National Institute of Diabetes and Digestive and Kidney Diseases , with additional funding from the National Institute of Child Health and Human Development , and the National Heart, Lung, and Blood Institute (U01-DK-66143, U01-DK-66174, U24-DK-082194, U24-DK-66116). FMW received support from the NIH T32 institutional training grant (T32 HL007024). The CKID website is located at https://www.statepi.jhsph.edu/ckid , and a list of CKiD collaborators can be found at https://statepi.jhsph.edu/ckid/site-investigators/ and is presented in Supplementary Table S6 . The authors acknowledge Drs. Alvaro Muñoz and Christopher Cox for critical insight and feedback on cross-validation methods. The authors deeply appreciate Dr. Franz Schaefer for providing permission to use ESCAPE study data used in the paper by Furth et al. 4 for the external validation analysis.

Keywords

kidney failure
kidney replacement therapy
pediatric chronic kidney disease
pediatric nephrology
prediction
risk stratification

ASJC Scopus subject areas

Nephrology

Access to Document

10.1016/j.kint.2023.06.020

Cite this

@article{e517c1e8a9924bae99b26d0b3ac256bb,

title = "Development of an adaptive clinical web-based prediction tool for kidney replacement therapy in children with chronic kidney disease",

abstract = "Clinicians need improved prediction models to estimate time to kidney replacement therapy (KRT) for children with chronic kidney disease (CKD). Here, we aimed to develop and validate a prediction tool based on common clinical variables for time to KRT in children using statistical learning methods and design a corresponding online calculator for clinical use. Among 890 children with CKD in the Chronic Kidney Disease in Children (CKiD) study, 172 variables related to sociodemographics, kidney/cardiovascular health, and therapy use, including longitudinal changes over one year were evaluated as candidate predictors in a random survival forest for time to KRT. An elementary model was specified with diagnosis, estimated glomerular filtration rate and proteinuria as predictors and then random survival forest identified nine additional candidate predictors for further evaluation. Best subset selection using these nine additional candidate predictors yielded an enriched model additionally based on blood pressure, change in estimated glomerular filtration rate over one year, anemia, albumin, chloride and bicarbonate. Four additional partially enriched models were constructed for clinical situations with incomplete data. Models performed well in cross-validation, and the elementary model was then externally validated using data from a European pediatric CKD cohort. A corresponding user-friendly online tool was developed for clinicians. Thus, our clinical prediction tool for time to KRT in children was developed in a large, representative pediatric CKD cohort with an exhaustive evaluation of potential predictors and supervised statistical learning methods. While our models performed well internally and externally, further external validation of enriched models is needed.",

keywords = "kidney failure, kidney replacement therapy, pediatric chronic kidney disease, pediatric nephrology, prediction, risk stratification",

author = "{CKiD investigators} and Ng, {Derek K.} and Matheson, {Matthew B.} and Schwartz, {George J.} and Wang, {Frances M.} and Mendley, {Susan R.} and Furth, {Susan L.} and Warady, {Bradley A.} and Sahar Fathallah-Shaykh and Anjali Nayak and Martin Turman and Tom Blydt-Hansen and Cynthia Wong and Steve Alexander and Ora Yadin and Elizabeth Ingulli and Robert Mak and Cheryl Sanchez-Kazi and Asha Moudgil and Samina Muneeruddin and Carolyn Abitbol and Marissa DeFrietas and Chryso Katsoufis and Wacharee Seeherunvong and Larry Greenbaum and Lyndsay Harshman and Priya Verghese and Sonia Krishnan and Amy Wilson and Stefan Kiessling and Margaret Murphy and Siddharth Shah and Janice Sullivan and Sushil Gupta and Samir El-Dahr and Stacy Drury and Nancy Rodig and Allison Dart and Meredith Atkinson and Arlene Gerson and Tej Matoo and Zubin Modi and Jason Thomas and Bradley Warady and Rebecca Johnson and Vikas Dharnidharka and Stephen Hooper and Susan Massengill and Liliana Gomez-Mendez and Matthew Hand and Joann Carlson",

note = "Publisher Copyright: {\textcopyright} 2023 International Society of Nephrology",

year = "2023",

month = nov,

doi = "10.1016/j.kint.2023.06.020",

language = "English (US)",

volume = "104",

pages = "985--994",

journal = "Kidney international",

issn = "0085-2538",

publisher = "Elsevier B.V.",

number = "5",

}

TY - JOUR

T1 - Development of an adaptive clinical web-based prediction tool for kidney replacement therapy in children with chronic kidney disease

AU - CKiD investigators

AU - Ng, Derek K.

AU - Matheson, Matthew B.

AU - Schwartz, George J.

AU - Wang, Frances M.

AU - Mendley, Susan R.

AU - Furth, Susan L.

AU - Warady, Bradley A.

AU - Fathallah-Shaykh, Sahar

AU - Nayak, Anjali

AU - Turman, Martin

AU - Blydt-Hansen, Tom

AU - Wong, Cynthia

AU - Alexander, Steve

AU - Yadin, Ora

AU - Ingulli, Elizabeth

AU - Mak, Robert

AU - Sanchez-Kazi, Cheryl

AU - Moudgil, Asha

AU - Muneeruddin, Samina

AU - Abitbol, Carolyn

AU - DeFrietas, Marissa

AU - Katsoufis, Chryso

AU - Seeherunvong, Wacharee

AU - Greenbaum, Larry

AU - Harshman, Lyndsay

AU - Verghese, Priya

AU - Krishnan, Sonia

AU - Wilson, Amy

AU - Kiessling, Stefan

AU - Murphy, Margaret

AU - Shah, Siddharth

AU - Sullivan, Janice

AU - Gupta, Sushil

AU - El-Dahr, Samir

AU - Drury, Stacy

AU - Rodig, Nancy

AU - Dart, Allison

AU - Atkinson, Meredith

AU - Gerson, Arlene

AU - Matoo, Tej

AU - Modi, Zubin

AU - Thomas, Jason

AU - Warady, Bradley

AU - Johnson, Rebecca

AU - Dharnidharka, Vikas

AU - Hooper, Stephen

AU - Massengill, Susan

AU - Gomez-Mendez, Liliana

AU - Hand, Matthew

AU - Carlson, Joann

PY - 2023/11

Y1 - 2023/11

N2 - Clinicians need improved prediction models to estimate time to kidney replacement therapy (KRT) for children with chronic kidney disease (CKD). Here, we aimed to develop and validate a prediction tool based on common clinical variables for time to KRT in children using statistical learning methods and design a corresponding online calculator for clinical use. Among 890 children with CKD in the Chronic Kidney Disease in Children (CKiD) study, 172 variables related to sociodemographics, kidney/cardiovascular health, and therapy use, including longitudinal changes over one year were evaluated as candidate predictors in a random survival forest for time to KRT. An elementary model was specified with diagnosis, estimated glomerular filtration rate and proteinuria as predictors and then random survival forest identified nine additional candidate predictors for further evaluation. Best subset selection using these nine additional candidate predictors yielded an enriched model additionally based on blood pressure, change in estimated glomerular filtration rate over one year, anemia, albumin, chloride and bicarbonate. Four additional partially enriched models were constructed for clinical situations with incomplete data. Models performed well in cross-validation, and the elementary model was then externally validated using data from a European pediatric CKD cohort. A corresponding user-friendly online tool was developed for clinicians. Thus, our clinical prediction tool for time to KRT in children was developed in a large, representative pediatric CKD cohort with an exhaustive evaluation of potential predictors and supervised statistical learning methods. While our models performed well internally and externally, further external validation of enriched models is needed.

AB - Clinicians need improved prediction models to estimate time to kidney replacement therapy (KRT) for children with chronic kidney disease (CKD). Here, we aimed to develop and validate a prediction tool based on common clinical variables for time to KRT in children using statistical learning methods and design a corresponding online calculator for clinical use. Among 890 children with CKD in the Chronic Kidney Disease in Children (CKiD) study, 172 variables related to sociodemographics, kidney/cardiovascular health, and therapy use, including longitudinal changes over one year were evaluated as candidate predictors in a random survival forest for time to KRT. An elementary model was specified with diagnosis, estimated glomerular filtration rate and proteinuria as predictors and then random survival forest identified nine additional candidate predictors for further evaluation. Best subset selection using these nine additional candidate predictors yielded an enriched model additionally based on blood pressure, change in estimated glomerular filtration rate over one year, anemia, albumin, chloride and bicarbonate. Four additional partially enriched models were constructed for clinical situations with incomplete data. Models performed well in cross-validation, and the elementary model was then externally validated using data from a European pediatric CKD cohort. A corresponding user-friendly online tool was developed for clinicians. Thus, our clinical prediction tool for time to KRT in children was developed in a large, representative pediatric CKD cohort with an exhaustive evaluation of potential predictors and supervised statistical learning methods. While our models performed well internally and externally, further external validation of enriched models is needed.

KW - kidney failure

KW - kidney replacement therapy

KW - pediatric chronic kidney disease

KW - pediatric nephrology

KW - prediction

KW - risk stratification

UR - http://www.scopus.com/inward/record.url?scp=85168516160&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85168516160&partnerID=8YFLogxK

U2 - 10.1016/j.kint.2023.06.020

DO - 10.1016/j.kint.2023.06.020

M3 - Article

C2 - 37391041

AN - SCOPUS:85168516160

SN - 0085-2538

VL - 104

SP - 985

EP - 994

JO - Kidney international

JF - Kidney international

IS - 5

ER -

Development of an adaptive clinical web-based prediction tool for kidney replacement therapy in children with chronic kidney disease

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this