International comparisons of laboratory values from the 4CE collaborative to predict COVID-19 mortality

The Consortium For Clinical Characterization Of COVID-19 By EHR (4CE)

doi:10.1038/s41746-022-00601-0

International comparisons of laboratory values from the 4CE collaborative to predict COVID-19 mortality

The Consortium For Clinical Characterization Of COVID-19 By EHR (4CE)

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

12 Scopus citations

Abstract

Given the growing number of prediction algorithms developed to predict COVID-19 mortality, we evaluated the transportability of a mortality prediction algorithm using a multi-national network of healthcare systems. We predicted COVID-19 mortality using baseline commonly measured laboratory values and standard demographic and clinical covariates across healthcare systems, countries, and continents. Specifically, we trained a Cox regression model with nine measured laboratory test values, standard demographics at admission, and comorbidity burden pre-admission. These models were compared at site, country, and continent level. Of the 39,969 hospitalized patients with COVID-19 (68.6% male), 5717 (14.3%) died. In the Cox model, age, albumin, AST, creatine, CRP, and white blood cell count are most predictive of mortality. The baseline covariates are more predictive of mortality during the early days of COVID-19 hospitalization. Models trained at healthcare systems with larger cohort size largely retain good transportability performance when porting to different sites. The combination of routine laboratory test values at admission along with basic demographic features can predict mortality in patients hospitalized with COVID-19. Importantly, this potentially deployable model differs from prior work by demonstrating not only consistent performance but also reliable transportability across healthcare systems in the US and Europe, highlighting the generalizability of this model and the overall approach.

Original language	English (US)
Article number	74
Journal	npj Digital Medicine
Volume	5
Issue number	1
DOIs	https://doi.org/10.1038/s41746-022-00601-0
State	Published - Dec 2022

Funding

G.W. reports funding from NCATS UL1TR002541, NCATS UL1TR000005, and NLM R01LM013345. S.M. and J.K. report funding from NCATS 5UL1TR001857-05 and NHGRI 5R01HG009174-04. Z.X. reports funding from NINDS R01NS098023. G.O. reports funding from NIH grants NIEHS P30ES017885 and NCI U24CA210967. S.V. reports funding from NLM R01LM012095 and NCATS UL1TR001857. A.S. reports funding from NHLBI K23HL148394 and L40HL148910, and NCATS UL1TR001420. B.A. reports funding from NHLBI U24 HL148865. D.B. and R.F. report funding from NCATS UL1TR001881. T.G. and T.G. report funding from 01ZZ1801E German Federal Ministry of Education and Research. D.H. reports funding from NCATS UL1TR002240. M.K. reports funding from NHGRI 5T32HG002295-18. D.K. reports funding from MIRACUM Consortium grant 01ZZ1801A. Y.L. reports funding from NLM R01LM01333. J.M. reports funding from NCATS UL1TR001878. D.M. reports funding from NCATS UL1-TR001878 Institutional Clinical and Translational Science Award (University of Pennsylvania). L.P. reports funding from NCATS CTSA Award #UL1TR002366.

ASJC Scopus subject areas

Medicine (miscellaneous)
Health Informatics
Computer Science Applications
Health Information Management

Access to Document

10.1038/s41746-022-00601-0

Cite this

@article{198d0aa4e7084b9ea617e3dad908333a,

title = "International comparisons of laboratory values from the 4CE collaborative to predict COVID-19 mortality",

abstract = "Given the growing number of prediction algorithms developed to predict COVID-19 mortality, we evaluated the transportability of a mortality prediction algorithm using a multi-national network of healthcare systems. We predicted COVID-19 mortality using baseline commonly measured laboratory values and standard demographic and clinical covariates across healthcare systems, countries, and continents. Specifically, we trained a Cox regression model with nine measured laboratory test values, standard demographics at admission, and comorbidity burden pre-admission. These models were compared at site, country, and continent level. Of the 39,969 hospitalized patients with COVID-19 (68.6% male), 5717 (14.3%) died. In the Cox model, age, albumin, AST, creatine, CRP, and white blood cell count are most predictive of mortality. The baseline covariates are more predictive of mortality during the early days of COVID-19 hospitalization. Models trained at healthcare systems with larger cohort size largely retain good transportability performance when porting to different sites. The combination of routine laboratory test values at admission along with basic demographic features can predict mortality in patients hospitalized with COVID-19. Importantly, this potentially deployable model differs from prior work by demonstrating not only consistent performance but also reliable transportability across healthcare systems in the US and Europe, highlighting the generalizability of this model and the overall approach.",

author = "{The Consortium For Clinical Characterization Of COVID-19 By EHR (4CE)} and Weber, {Griffin M.} and Chuan Hong and Zongqi Xia and Palmer, {Nathan P.} and Paul Avillach and Sehi L{\textquoteright}Yi and Keller, {Mark S.} and Murphy, {Shawn N.} and Alba Guti{\'e}rrez-Sacrist{\'a}n and Bonzel, {Clara Lea} and Arnaud Serret-Larmande and Antoine Neuraz and Omenn, {Gilbert S.} and Shyam Visweswaran and Klann, {Jeffrey G.} and South, {Andrew M.} and Loh, {Ne Hooi Will} and Mario Cannataro and Beaulieu-Jones, {Brett K.} and Riccardo Bellazzi and Giuseppe Agapito and Mario Alessiani and Aronow, {Bruce J.} and Bell, {Douglas S.} and Vincent Benoit and Bourgeois, {Florence T.} and Luca Chiovato and Kelly Cho and Arianna Dagliati and DuVall, {Scott L.} and Barrio, {Noelia Garc{\'i}a} and Hanauer, {David A.} and Ho, {Yuk Lam} and Holmes, {John H.} and Issitt, {Richard W.} and Molei Liu and Yuan Luo and Lynch, {Kristine E.} and Maidlow, {Sarah E.} and Alberto Malovini and Mandl, {Kenneth D.} and Chengsheng Mao and Matheny, {Michael E.} and Moore, {Jason H.} and Morris, {Jeffrey S.} and Michele Morris and Mowery, {Danielle L.} and Ngiam, {Kee Yuan} and Patel, {Lav P.} and Sanchez-Pinto, {L. Nelson}",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s).",

year = "2022",

month = dec,

doi = "10.1038/s41746-022-00601-0",

language = "English (US)",

volume = "5",

journal = "npj Digital Medicine",

issn = "2398-6352",

publisher = "Nature Publishing Group",

number = "1",

}

TY - JOUR

T1 - International comparisons of laboratory values from the 4CE collaborative to predict COVID-19 mortality

AU - The Consortium For Clinical Characterization Of COVID-19 By EHR (4CE)

AU - Weber, Griffin M.

AU - Hong, Chuan

AU - Xia, Zongqi

AU - Palmer, Nathan P.

AU - Avillach, Paul

AU - L’Yi, Sehi

AU - Keller, Mark S.

AU - Murphy, Shawn N.

AU - Gutiérrez-Sacristán, Alba

AU - Bonzel, Clara Lea

AU - Serret-Larmande, Arnaud

AU - Neuraz, Antoine

AU - Omenn, Gilbert S.

AU - Visweswaran, Shyam

AU - Klann, Jeffrey G.

AU - South, Andrew M.

AU - Loh, Ne Hooi Will

AU - Cannataro, Mario

AU - Beaulieu-Jones, Brett K.

AU - Bellazzi, Riccardo

AU - Agapito, Giuseppe

AU - Alessiani, Mario

AU - Aronow, Bruce J.

AU - Bell, Douglas S.

AU - Benoit, Vincent

AU - Bourgeois, Florence T.

AU - Chiovato, Luca

AU - Cho, Kelly

AU - Dagliati, Arianna

AU - DuVall, Scott L.

AU - Barrio, Noelia García

AU - Hanauer, David A.

AU - Ho, Yuk Lam

AU - Holmes, John H.

AU - Issitt, Richard W.

AU - Liu, Molei

AU - Luo, Yuan

AU - Lynch, Kristine E.

AU - Maidlow, Sarah E.

AU - Malovini, Alberto

AU - Mandl, Kenneth D.

AU - Mao, Chengsheng

AU - Matheny, Michael E.

AU - Moore, Jason H.

AU - Morris, Jeffrey S.

AU - Morris, Michele

AU - Mowery, Danielle L.

AU - Ngiam, Kee Yuan

AU - Patel, Lav P.

AU - Sanchez-Pinto, L. Nelson

PY - 2022/12

Y1 - 2022/12

N2 - Given the growing number of prediction algorithms developed to predict COVID-19 mortality, we evaluated the transportability of a mortality prediction algorithm using a multi-national network of healthcare systems. We predicted COVID-19 mortality using baseline commonly measured laboratory values and standard demographic and clinical covariates across healthcare systems, countries, and continents. Specifically, we trained a Cox regression model with nine measured laboratory test values, standard demographics at admission, and comorbidity burden pre-admission. These models were compared at site, country, and continent level. Of the 39,969 hospitalized patients with COVID-19 (68.6% male), 5717 (14.3%) died. In the Cox model, age, albumin, AST, creatine, CRP, and white blood cell count are most predictive of mortality. The baseline covariates are more predictive of mortality during the early days of COVID-19 hospitalization. Models trained at healthcare systems with larger cohort size largely retain good transportability performance when porting to different sites. The combination of routine laboratory test values at admission along with basic demographic features can predict mortality in patients hospitalized with COVID-19. Importantly, this potentially deployable model differs from prior work by demonstrating not only consistent performance but also reliable transportability across healthcare systems in the US and Europe, highlighting the generalizability of this model and the overall approach.

AB - Given the growing number of prediction algorithms developed to predict COVID-19 mortality, we evaluated the transportability of a mortality prediction algorithm using a multi-national network of healthcare systems. We predicted COVID-19 mortality using baseline commonly measured laboratory values and standard demographic and clinical covariates across healthcare systems, countries, and continents. Specifically, we trained a Cox regression model with nine measured laboratory test values, standard demographics at admission, and comorbidity burden pre-admission. These models were compared at site, country, and continent level. Of the 39,969 hospitalized patients with COVID-19 (68.6% male), 5717 (14.3%) died. In the Cox model, age, albumin, AST, creatine, CRP, and white blood cell count are most predictive of mortality. The baseline covariates are more predictive of mortality during the early days of COVID-19 hospitalization. Models trained at healthcare systems with larger cohort size largely retain good transportability performance when porting to different sites. The combination of routine laboratory test values at admission along with basic demographic features can predict mortality in patients hospitalized with COVID-19. Importantly, this potentially deployable model differs from prior work by demonstrating not only consistent performance but also reliable transportability across healthcare systems in the US and Europe, highlighting the generalizability of this model and the overall approach.

UR - http://www.scopus.com/inward/record.url?scp=85131887608&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85131887608&partnerID=8YFLogxK

U2 - 10.1038/s41746-022-00601-0

DO - 10.1038/s41746-022-00601-0

M3 - Article

C2 - 35697747

AN - SCOPUS:85131887608

SN - 2398-6352

VL - 5

JO - npj Digital Medicine

JF - npj Digital Medicine

IS - 1

M1 - 74

ER -

International comparisons of laboratory values from the 4CE collaborative to predict COVID-19 mortality

Abstract

Funding

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this