The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort studyCitation formats

  • External authors:
  • Carlo Berzuini
  • Cathal John Hannan
  • Andrew Thomas King
  • Claire O'Leary
  • James Galea
  • Megan Wright
  • Omar Pathmanaban
  • Luisa Bernardinelli
  • Hiren Patel

Standard

The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort study. / Berzuini, Carlo; Hannan, Cathal John ; King, Andrew Thomas; Vail, Andy; O'Leary, Claire; Brough, David; Galea, James; Ogungbenro, Kayode; Wright, Megan ; Pathmanaban, Omar; Hulme, Sharon; Allan, Stuart; Bernardinelli, Luisa; Patel, Hiren.

In: BMJ Open, 02.08.2020.

Research output: Contribution to journalArticlepeer-review

Harvard

APA

Berzuini, C., Hannan, C. J., King, A. T., Vail, A., O'Leary, C., Brough, D., Galea, J., Ogungbenro, K., Wright, M., Pathmanaban, O., Hulme, S., Allan, S., Bernardinelli, L., & Patel, H. (Accepted/In press). The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort study. BMJ Open.

Vancouver

Author

Berzuini, Carlo ; Hannan, Cathal John ; King, Andrew Thomas ; Vail, Andy ; O'Leary, Claire ; Brough, David ; Galea, James ; Ogungbenro, Kayode ; Wright, Megan ; Pathmanaban, Omar ; Hulme, Sharon ; Allan, Stuart ; Bernardinelli, Luisa ; Patel, Hiren. / The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort study. In: BMJ Open. 2020.

Bibtex

@article{ed75394e19c24b128853f3676c9bef7a,
title = "The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort study",
abstract = "Objectives Being able to predict which COVID-19 patients are going to deteriorate is important to help identify patients for clinical and research practice. Clinical prediction models play a critical role in this process, but current models are of limited value because they are typically restricted to baseline predictors and don{\textquoteright}t always use contemporary statistical methods. We sought to explore the benefits of incorporating dynamic changes in routinely measured biomarkers, non-linear effects and applying {\textquoteleft}state of the art{\textquoteright} statistical methods in the development of a prognostic model to predict death in hospitalised COVID-19 patients. DesignData were analysed from COVID-19 admissions to three hospital sites. Exploratory data analysis included a graphical approach to partial correlations. Dynamic biomarkers were considered up to five days following admission rather than depending solely on baseline or single time-point data. Marked departures from linear effects of covariates were identified by employing smoothing splines within a generalised additive modelling framework.Setting3 secondary and tertiary level centres in Greater Manchester, UK.Participants392 hospitalised patients with a diagnosis of COVID-19Results392 patients with a COVID-19 diagnosis were identified. Area under the receiver operating characteristic (ROC) curve increased from 0.73 using admission data alone to 0.75 when also considering results of baseline blood samples and to 0.83 when considering dynamic values of routinely collected markers. There was clear non-linearity in the association of age with patient outcome.ConclusionsThis study shows that clinical prediction models to predict death in hospitalised COVID-19 patients can be improved by taking into account both non-linear effects in covariates such as age and dynamic changes in values of biomarkers.",
author = "Carlo Berzuini and Hannan, {Cathal John} and King, {Andrew Thomas} and Andy Vail and Claire O'Leary and David Brough and James Galea and Kayode Ogungbenro and Megan Wright and Omar Pathmanaban and Sharon Hulme and Stuart Allan and Luisa Bernardinelli and Hiren Patel",
year = "2020",
month = aug,
day = "2",
language = "English",
journal = "BMJ Open",
issn = "2044-6055",
publisher = "BMJ ",

}

RIS

TY - JOUR

T1 - The value of dynamic clinical and biomarker data for mortality risk prediction in COVID-19: A multi-centre retrospective cohort study

AU - Berzuini, Carlo

AU - Hannan, Cathal John

AU - King, Andrew Thomas

AU - Vail, Andy

AU - O'Leary, Claire

AU - Brough, David

AU - Galea, James

AU - Ogungbenro, Kayode

AU - Wright, Megan

AU - Pathmanaban, Omar

AU - Hulme, Sharon

AU - Allan, Stuart

AU - Bernardinelli, Luisa

AU - Patel, Hiren

PY - 2020/8/2

Y1 - 2020/8/2

N2 - Objectives Being able to predict which COVID-19 patients are going to deteriorate is important to help identify patients for clinical and research practice. Clinical prediction models play a critical role in this process, but current models are of limited value because they are typically restricted to baseline predictors and don’t always use contemporary statistical methods. We sought to explore the benefits of incorporating dynamic changes in routinely measured biomarkers, non-linear effects and applying ‘state of the art’ statistical methods in the development of a prognostic model to predict death in hospitalised COVID-19 patients. DesignData were analysed from COVID-19 admissions to three hospital sites. Exploratory data analysis included a graphical approach to partial correlations. Dynamic biomarkers were considered up to five days following admission rather than depending solely on baseline or single time-point data. Marked departures from linear effects of covariates were identified by employing smoothing splines within a generalised additive modelling framework.Setting3 secondary and tertiary level centres in Greater Manchester, UK.Participants392 hospitalised patients with a diagnosis of COVID-19Results392 patients with a COVID-19 diagnosis were identified. Area under the receiver operating characteristic (ROC) curve increased from 0.73 using admission data alone to 0.75 when also considering results of baseline blood samples and to 0.83 when considering dynamic values of routinely collected markers. There was clear non-linearity in the association of age with patient outcome.ConclusionsThis study shows that clinical prediction models to predict death in hospitalised COVID-19 patients can be improved by taking into account both non-linear effects in covariates such as age and dynamic changes in values of biomarkers.

AB - Objectives Being able to predict which COVID-19 patients are going to deteriorate is important to help identify patients for clinical and research practice. Clinical prediction models play a critical role in this process, but current models are of limited value because they are typically restricted to baseline predictors and don’t always use contemporary statistical methods. We sought to explore the benefits of incorporating dynamic changes in routinely measured biomarkers, non-linear effects and applying ‘state of the art’ statistical methods in the development of a prognostic model to predict death in hospitalised COVID-19 patients. DesignData were analysed from COVID-19 admissions to three hospital sites. Exploratory data analysis included a graphical approach to partial correlations. Dynamic biomarkers were considered up to five days following admission rather than depending solely on baseline or single time-point data. Marked departures from linear effects of covariates were identified by employing smoothing splines within a generalised additive modelling framework.Setting3 secondary and tertiary level centres in Greater Manchester, UK.Participants392 hospitalised patients with a diagnosis of COVID-19Results392 patients with a COVID-19 diagnosis were identified. Area under the receiver operating characteristic (ROC) curve increased from 0.73 using admission data alone to 0.75 when also considering results of baseline blood samples and to 0.83 when considering dynamic values of routinely collected markers. There was clear non-linearity in the association of age with patient outcome.ConclusionsThis study shows that clinical prediction models to predict death in hospitalised COVID-19 patients can be improved by taking into account both non-linear effects in covariates such as age and dynamic changes in values of biomarkers.

M3 - Article

JO - BMJ Open

JF - BMJ Open

SN - 2044-6055

ER -