Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate

Felipe Kenji Nakano; Anna Åkesson; Jasper de Boer; Klest Dedja; Robbe D’hondt; Fateme Nateghi Haredasht; Jonas Björk; Marie Courbebaisse; Lionel Couzi; Natalie Ebert; Björn O. Eriksen; R. Neil Dalton; Laurence Derain-Dubourg; Francois Gaillard; Cyril Garrouste; Anders Grubb; Lola Jacquemont; Magnus Hansson; Nassim Kamar; Christophe Legendre; Karin Littmann; Christophe Mariat; Toralf Melsom; Lionel Rostaing; Andrew D. Rule; Elke Schaeffner; Per-Ola Sundin; Arend Bökenkamp; Ulla Berg; Kajsa Åsling-Monemi; Luciano Selistre; Anders Larsson; Ulf Nyman; Antoine Lanot; Hans Pottel; Pierre Delanaye; Celine Vens

doi:10.1038/s41598-024-77618-w

Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate

Felipe Kenji Nakano^*
, Anna Åkesson
, Jasper de Boer
, Klest Dedja
, Robbe D’hondt
, Fateme Nateghi Haredasht
, Jonas Björk
, Marie Courbebaisse
, Lionel Couzi
, Natalie Ebert
, Björn O. Eriksen
, R. Neil Dalton
, Laurence Derain-Dubourg
, Francois Gaillard
, Cyril Garrouste
, Anders Grubb
, Lola Jacquemont
, Magnus Hansson
, Nassim Kamar
, Christophe Legendre

Karin Littmann, Christophe Mariat, Toralf Melsom, Lionel Rostaing, Andrew D. Rule, Elke Schaeffner, Per-Ola Sundin, Arend Bökenkamp, Ulla Berg, Kajsa Åsling-Monemi, Luciano Selistre, Anders Larsson, Ulf Nyman, Antoine Lanot, Hans Pottel, Pierre Delanaye, Celine Vens

^*Corresponding author for this work

KU Leuven
Lund University
Assistance publique – Hôpitaux de Paris
IMMUNOLOGY from CONCEPT and EXPERIMENTS to TRANSLATION
Charité – Universitätsmedizin Berlin
University of Tromsø – The Arctic University of Norway
Evelina London Children's Healthcare
Hôpital Édouard Herriot
CHU de Clermont-Ferrand
CHU de Nantes
Karolinska Institutet
Université Toulouse III - Paul Sabatier
Université Paris Cité
Hôpital Nord, CHU de Saint-Étienne
CHU Grenoble-Alpes
Mayo Clinic Rochester, MN
Örebro University
Vrije Universiteit Amsterdam
Mestrado em Ciências da Saúde – Universidade Caxias do Sul Foundation CAPES
Uppsala University
Université de Caen Normandie
Centre Georges-François Leclerc
University of Liege
Hôpital Universitaire Carémeau

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

15 Downloads (Pure)

Abstract

In clinical practice, the glomerular filtration rate (GFR), a measurement of kidney functioning, is normally calculated using equations, such as the European Kidney Function Consortium (EKFC) equation. Despite being the most general equation, EKFC, just like previously proposed approaches, can still struggle to achieve satisfactory performance, limiting its clinical applicability. As a possible solution, recently machine learning (ML) has been investigated to improve GFR prediction, nonetheless the literature still lacks a general and multi-center study. Using a dataset with 19,629 patients from 13 cohorts, we investigate if ML can improve GFR prediction in comparison to EKFC. More specifically, we compare diverse ML methods, which were allowed to use age, sex, serum creatinine, cystatin C, height, weight and BMI as features, in internal and external cohorts against EKFC. The results show that the most performing ML method, random forest (RF), and EKFC are very competitive where RF and EKFC achieved respectively P10 and P30 values of 0.45 (95% CI 0.44;0.46) and 0.89 (95% CI 0.88;0.90), whereas EKFC yielded 0.44 (95% CI 0.43; 0.44) and 0.89 (95% CI 0.88; 0.90), considering the entire cohort. Small differences were, however, observed in patients younger than 12 years where RF slightly outperformed EKFC.

Original language	English
Article number	26383
Journal	Scientific reports
Volume	14
Issue number	1
DOIs	https://doi.org/10.1038/s41598-024-77618-w
Publication status	Published - 1 Dec 2024

Access to Document

10.1038/s41598-024-77618-w

Comparison-between-the-ekfc-equation-and-machine-learning-models-to-predict-glomerular-filtration-rate.pdfFinal published version, 2.35 MBLicence: CC BY

Cite this

Nakano, F. K., Åkesson, A., de Boer, J., Dedja, K., D’hondt, R., Haredasht, F. N., Björk, J., Courbebaisse, M., Couzi, L., Ebert, N., Eriksen, B. O., Dalton, R. N., Derain-Dubourg, L., Gaillard, F., Garrouste, C., Grubb, A., Jacquemont, L., Hansson, M., Kamar, N., ... Vens, C. (2024). Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate. Scientific reports, 14(1), Article 26383. https://doi.org/10.1038/s41598-024-77618-w

@article{64271ff54fbb40d9ac32e212435a4e16,

title = "Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate",

abstract = "In clinical practice, the glomerular filtration rate (GFR), a measurement of kidney functioning, is normally calculated using equations, such as the European Kidney Function Consortium (EKFC) equation. Despite being the most general equation, EKFC, just like previously proposed approaches, can still struggle to achieve satisfactory performance, limiting its clinical applicability. As a possible solution, recently machine learning (ML) has been investigated to improve GFR prediction, nonetheless the literature still lacks a general and multi-center study. Using a dataset with 19,629 patients from 13 cohorts, we investigate if ML can improve GFR prediction in comparison to EKFC. More specifically, we compare diverse ML methods, which were allowed to use age, sex, serum creatinine, cystatin C, height, weight and BMI as features, in internal and external cohorts against EKFC. The results show that the most performing ML method, random forest (RF), and EKFC are very competitive where RF and EKFC achieved respectively P10 and P30 values of 0.45 (95\% CI 0.44;0.46) and 0.89 (95\% CI 0.88;0.90), whereas EKFC yielded 0.44 (95\% CI 0.43; 0.44) and 0.89 (95\% CI 0.88; 0.90), considering the entire cohort. Small differences were, however, observed in patients younger than 12 years where RF slightly outperformed EKFC.",

author = "Nakano, \{Felipe Kenji\} and Anna {\AA}kesson and \{de Boer\}, Jasper and Klest Dedja and Robbe D{\textquoteright}hondt and Haredasht, \{Fateme Nateghi\} and Jonas Bj{\"o}rk and Marie Courbebaisse and Lionel Couzi and Natalie Ebert and Eriksen, \{Bj{\"o}rn O.\} and Dalton, \{R. Neil\} and Laurence Derain-Dubourg and Francois Gaillard and Cyril Garrouste and Anders Grubb and Lola Jacquemont and Magnus Hansson and Nassim Kamar and Christophe Legendre and Karin Littmann and Christophe Mariat and Toralf Melsom and Lionel Rostaing and Rule, \{Andrew D.\} and Elke Schaeffner and Per-Ola Sundin and Arend B{\"o}kenkamp and Ulla Berg and Kajsa {\AA}sling-Monemi and Luciano Selistre and Anders Larsson and Ulf Nyman and Antoine Lanot and Hans Pottel and Pierre Delanaye and Celine Vens",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2024.",

year = "2024",

month = dec,

day = "1",

doi = "10.1038/s41598-024-77618-w",

language = "English",

volume = "14",

journal = "Scientific reports",

issn = "2045-2322",

publisher = "Nature Publishing Group",

number = "1",

}

Nakano, FK, Åkesson, A, de Boer, J, Dedja, K, D’hondt, R, Haredasht, FN, Björk, J, Courbebaisse, M, Couzi, L, Ebert, N, Eriksen, BO, Dalton, RN, Derain-Dubourg, L, Gaillard, F, Garrouste, C, Grubb, A, Jacquemont, L, Hansson, M, Kamar, N, Legendre, C, Littmann, K, Mariat, C, Melsom, T, Rostaing, L, Rule, AD, Schaeffner, E, Sundin, P-O, Bökenkamp, A, Berg, U, Åsling-Monemi, K, Selistre, L, Larsson, A, Nyman, U, Lanot, A, Pottel, H, Delanaye, P & Vens, C 2024, 'Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate', Scientific reports, vol. 14, no. 1, 26383. https://doi.org/10.1038/s41598-024-77618-w

TY - JOUR

T1 - Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate

AU - Nakano, Felipe Kenji

AU - Åkesson, Anna

AU - de Boer, Jasper

AU - Dedja, Klest

AU - D’hondt, Robbe

AU - Haredasht, Fateme Nateghi

AU - Björk, Jonas

AU - Courbebaisse, Marie

AU - Couzi, Lionel

AU - Ebert, Natalie

AU - Eriksen, Björn O.

AU - Dalton, R. Neil

AU - Derain-Dubourg, Laurence

AU - Gaillard, Francois

AU - Garrouste, Cyril

AU - Grubb, Anders

AU - Jacquemont, Lola

AU - Hansson, Magnus

AU - Kamar, Nassim

AU - Legendre, Christophe

AU - Littmann, Karin

AU - Mariat, Christophe

AU - Melsom, Toralf

AU - Rostaing, Lionel

AU - Rule, Andrew D.

AU - Schaeffner, Elke

AU - Sundin, Per-Ola

AU - Bökenkamp, Arend

AU - Berg, Ulla

AU - Åsling-Monemi, Kajsa

AU - Selistre, Luciano

AU - Larsson, Anders

AU - Nyman, Ulf

AU - Lanot, Antoine

AU - Pottel, Hans

AU - Delanaye, Pierre

AU - Vens, Celine

PY - 2024/12/1

Y1 - 2024/12/1

N2 - In clinical practice, the glomerular filtration rate (GFR), a measurement of kidney functioning, is normally calculated using equations, such as the European Kidney Function Consortium (EKFC) equation. Despite being the most general equation, EKFC, just like previously proposed approaches, can still struggle to achieve satisfactory performance, limiting its clinical applicability. As a possible solution, recently machine learning (ML) has been investigated to improve GFR prediction, nonetheless the literature still lacks a general and multi-center study. Using a dataset with 19,629 patients from 13 cohorts, we investigate if ML can improve GFR prediction in comparison to EKFC. More specifically, we compare diverse ML methods, which were allowed to use age, sex, serum creatinine, cystatin C, height, weight and BMI as features, in internal and external cohorts against EKFC. The results show that the most performing ML method, random forest (RF), and EKFC are very competitive where RF and EKFC achieved respectively P10 and P30 values of 0.45 (95% CI 0.44;0.46) and 0.89 (95% CI 0.88;0.90), whereas EKFC yielded 0.44 (95% CI 0.43; 0.44) and 0.89 (95% CI 0.88; 0.90), considering the entire cohort. Small differences were, however, observed in patients younger than 12 years where RF slightly outperformed EKFC.

AB - In clinical practice, the glomerular filtration rate (GFR), a measurement of kidney functioning, is normally calculated using equations, such as the European Kidney Function Consortium (EKFC) equation. Despite being the most general equation, EKFC, just like previously proposed approaches, can still struggle to achieve satisfactory performance, limiting its clinical applicability. As a possible solution, recently machine learning (ML) has been investigated to improve GFR prediction, nonetheless the literature still lacks a general and multi-center study. Using a dataset with 19,629 patients from 13 cohorts, we investigate if ML can improve GFR prediction in comparison to EKFC. More specifically, we compare diverse ML methods, which were allowed to use age, sex, serum creatinine, cystatin C, height, weight and BMI as features, in internal and external cohorts against EKFC. The results show that the most performing ML method, random forest (RF), and EKFC are very competitive where RF and EKFC achieved respectively P10 and P30 values of 0.45 (95% CI 0.44;0.46) and 0.89 (95% CI 0.88;0.90), whereas EKFC yielded 0.44 (95% CI 0.43; 0.44) and 0.89 (95% CI 0.88; 0.90), considering the entire cohort. Small differences were, however, observed in patients younger than 12 years where RF slightly outperformed EKFC.

UR - https://www.scopus.com/pages/publications/85208291511

U2 - 10.1038/s41598-024-77618-w

DO - 10.1038/s41598-024-77618-w

M3 - Article

C2 - 39487227

SN - 2045-2322

VL - 14

JO - Scientific reports

JF - Scientific reports

IS - 1

M1 - 26383

ER -

Comparison between the EKFC-equation and machine learning models to predict Glomerular Filtration Rate

Abstract

Access to Document

Other files and links

Fingerprint

Cite this