Equalization of four cardiovascular risk algorithms after systematic recalibration: individual-participant meta-analysis of 86 prospective studies

Lisa Pennells, Stephen Kaptoge, Angela Wood, Mike Sweeting, Xiaohui Zhao, Ian White, Stephen Burgess, Peter Willeit, Thomas Bolton, Karel G.M. Moons, Børge Grønne Nordestgaard, Torben Jørgensen, Emanuele Di Angelantonio*, Emerging Risk Factors Collaboration, Jørgen Jespersen, Else-Marie Bladbjerg

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

23 Downloads (Pure)

Abstract

Aims: There is debate about the optimum algorithm for cardiovascular disease (CVD) risk estimation. We conducted head-to-head comparisons of four algorithms recommended by primary prevention guidelines, before and after 'recalibration', a method that adapts risk algorithms to take account of differences in the risk characteristics of the populations being studied.

Methods and results: Using individual-participant data on 360 737 participants without CVD at baseline in 86 prospective studies from 22 countries, we compared the Framingham risk score (FRS), Systematic COronary Risk Evaluation (SCORE), pooled cohort equations (PCE), and Reynolds risk score (RRS). We calculated measures of risk discrimination and calibration, and modelled clinical implications of initiating statin therapy in people judged to be at 'high' 10 year CVD risk. Original risk algorithms were recalibrated using the risk factor profile and CVD incidence of target populations. The four algorithms had similar risk discrimination. Before recalibration, FRS, SCORE, and PCE over-predicted CVD risk on average by 10%, 52%, and 41%, respectively, whereas RRS under-predicted by 10%. Original versions of algorithms classified 29-39% of individuals aged ≥40 years as high risk. By contrast, recalibration reduced this proportion to 22-24% for every algorithm. We estimated that to prevent one CVD event, it would be necessary to initiate statin therapy in 44-51 such individuals using original algorithms, in contrast to 37-39 individuals with recalibrated algorithms.

Conclusion: Before recalibration, the clinical performance of four widely used CVD risk algorithms varied substantially. By contrast, simple recalibration nearly equalized their performance and improved modelled targeting of preventive action to clinical need.

Original languageEnglish
JournalEuropean Heart Journal
Volume40
Issue number7
Pages (from-to)621-631
ISSN0195-668X
DOIs
Publication statusPublished - 14. Feb 2019

Fingerprint

Meta-Analysis
Prospective Studies
Hydroxymethylglutaryl-CoA Reductase Inhibitors
Health Services Needs and Demand
Population Characteristics
Primary Prevention
Guidelines

Keywords

  • Calibration
  • Cardiovascular disease
  • Discrimination
  • Risk algorithms
  • Risk prediction

Cite this

Pennells, Lisa ; Kaptoge, Stephen ; Wood, Angela ; Sweeting, Mike ; Zhao, Xiaohui ; White, Ian ; Burgess, Stephen ; Willeit, Peter ; Bolton, Thomas ; Moons, Karel G.M. ; Nordestgaard, Børge Grønne ; Jørgensen, Torben ; Di Angelantonio, Emanuele ; Emerging Risk Factors Collaboration ; Jespersen, Jørgen ; Bladbjerg, Else-Marie. / Equalization of four cardiovascular risk algorithms after systematic recalibration : individual-participant meta-analysis of 86 prospective studies. In: European Heart Journal. 2019 ; Vol. 40, No. 7. pp. 621-631.
@article{a50d4695914446c9accaac860d015cf8,
title = "Equalization of four cardiovascular risk algorithms after systematic recalibration: individual-participant meta-analysis of 86 prospective studies",
abstract = "Aims: There is debate about the optimum algorithm for cardiovascular disease (CVD) risk estimation. We conducted head-to-head comparisons of four algorithms recommended by primary prevention guidelines, before and after 'recalibration', a method that adapts risk algorithms to take account of differences in the risk characteristics of the populations being studied.Methods and results: Using individual-participant data on 360 737 participants without CVD at baseline in 86 prospective studies from 22 countries, we compared the Framingham risk score (FRS), Systematic COronary Risk Evaluation (SCORE), pooled cohort equations (PCE), and Reynolds risk score (RRS). We calculated measures of risk discrimination and calibration, and modelled clinical implications of initiating statin therapy in people judged to be at 'high' 10 year CVD risk. Original risk algorithms were recalibrated using the risk factor profile and CVD incidence of target populations. The four algorithms had similar risk discrimination. Before recalibration, FRS, SCORE, and PCE over-predicted CVD risk on average by 10{\%}, 52{\%}, and 41{\%}, respectively, whereas RRS under-predicted by 10{\%}. Original versions of algorithms classified 29-39{\%} of individuals aged ≥40 years as high risk. By contrast, recalibration reduced this proportion to 22-24{\%} for every algorithm. We estimated that to prevent one CVD event, it would be necessary to initiate statin therapy in 44-51 such individuals using original algorithms, in contrast to 37-39 individuals with recalibrated algorithms.Conclusion: Before recalibration, the clinical performance of four widely used CVD risk algorithms varied substantially. By contrast, simple recalibration nearly equalized their performance and improved modelled targeting of preventive action to clinical need.",
keywords = "Calibration, Cardiovascular disease, Discrimination, Risk algorithms, Risk prediction",
author = "Lisa Pennells and Stephen Kaptoge and Angela Wood and Mike Sweeting and Xiaohui Zhao and Ian White and Stephen Burgess and Peter Willeit and Thomas Bolton and Moons, {Karel G.M.} and Nordestgaard, {B{\o}rge Gr{\o}nne} and Torben J{\o}rgensen and {Di Angelantonio}, Emanuele and {Emerging Risk Factors Collaboration} and J{\o}rgen Jespersen and Else-Marie Bladbjerg",
note = "Artiklen har en lang r{\ae}kke forfattere under Emerging Risk Factors Collaboration, men jeg har kun angivet forfattere tilknyttet SDU.",
year = "2019",
month = "2",
day = "14",
doi = "10.1093/eurheartj/ehy653",
language = "English",
volume = "40",
pages = "621--631",
journal = "European Heart Journal",
issn = "0195-668X",
publisher = "Heinemann",
number = "7",

}

Pennells, L, Kaptoge, S, Wood, A, Sweeting, M, Zhao, X, White, I, Burgess, S, Willeit, P, Bolton, T, Moons, KGM, Nordestgaard, BG, Jørgensen, T, Di Angelantonio, E, Emerging Risk Factors Collaboration, Jespersen, J & Bladbjerg, E-M 2019, 'Equalization of four cardiovascular risk algorithms after systematic recalibration: individual-participant meta-analysis of 86 prospective studies', European Heart Journal, vol. 40, no. 7, pp. 621-631. https://doi.org/10.1093/eurheartj/ehy653

Equalization of four cardiovascular risk algorithms after systematic recalibration : individual-participant meta-analysis of 86 prospective studies. / Pennells, Lisa; Kaptoge, Stephen; Wood, Angela; Sweeting, Mike; Zhao, Xiaohui; White, Ian; Burgess, Stephen; Willeit, Peter; Bolton, Thomas; Moons, Karel G.M.; Nordestgaard, Børge Grønne; Jørgensen, Torben; Di Angelantonio, Emanuele; Emerging Risk Factors Collaboration ; Jespersen, Jørgen; Bladbjerg, Else-Marie.

In: European Heart Journal, Vol. 40, No. 7, 14.02.2019, p. 621-631.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Equalization of four cardiovascular risk algorithms after systematic recalibration

T2 - individual-participant meta-analysis of 86 prospective studies

AU - Pennells, Lisa

AU - Kaptoge, Stephen

AU - Wood, Angela

AU - Sweeting, Mike

AU - Zhao, Xiaohui

AU - White, Ian

AU - Burgess, Stephen

AU - Willeit, Peter

AU - Bolton, Thomas

AU - Moons, Karel G.M.

AU - Nordestgaard, Børge Grønne

AU - Jørgensen, Torben

AU - Di Angelantonio, Emanuele

AU - Emerging Risk Factors Collaboration

AU - Jespersen, Jørgen

AU - Bladbjerg, Else-Marie

N1 - Artiklen har en lang række forfattere under Emerging Risk Factors Collaboration, men jeg har kun angivet forfattere tilknyttet SDU.

PY - 2019/2/14

Y1 - 2019/2/14

N2 - Aims: There is debate about the optimum algorithm for cardiovascular disease (CVD) risk estimation. We conducted head-to-head comparisons of four algorithms recommended by primary prevention guidelines, before and after 'recalibration', a method that adapts risk algorithms to take account of differences in the risk characteristics of the populations being studied.Methods and results: Using individual-participant data on 360 737 participants without CVD at baseline in 86 prospective studies from 22 countries, we compared the Framingham risk score (FRS), Systematic COronary Risk Evaluation (SCORE), pooled cohort equations (PCE), and Reynolds risk score (RRS). We calculated measures of risk discrimination and calibration, and modelled clinical implications of initiating statin therapy in people judged to be at 'high' 10 year CVD risk. Original risk algorithms were recalibrated using the risk factor profile and CVD incidence of target populations. The four algorithms had similar risk discrimination. Before recalibration, FRS, SCORE, and PCE over-predicted CVD risk on average by 10%, 52%, and 41%, respectively, whereas RRS under-predicted by 10%. Original versions of algorithms classified 29-39% of individuals aged ≥40 years as high risk. By contrast, recalibration reduced this proportion to 22-24% for every algorithm. We estimated that to prevent one CVD event, it would be necessary to initiate statin therapy in 44-51 such individuals using original algorithms, in contrast to 37-39 individuals with recalibrated algorithms.Conclusion: Before recalibration, the clinical performance of four widely used CVD risk algorithms varied substantially. By contrast, simple recalibration nearly equalized their performance and improved modelled targeting of preventive action to clinical need.

AB - Aims: There is debate about the optimum algorithm for cardiovascular disease (CVD) risk estimation. We conducted head-to-head comparisons of four algorithms recommended by primary prevention guidelines, before and after 'recalibration', a method that adapts risk algorithms to take account of differences in the risk characteristics of the populations being studied.Methods and results: Using individual-participant data on 360 737 participants without CVD at baseline in 86 prospective studies from 22 countries, we compared the Framingham risk score (FRS), Systematic COronary Risk Evaluation (SCORE), pooled cohort equations (PCE), and Reynolds risk score (RRS). We calculated measures of risk discrimination and calibration, and modelled clinical implications of initiating statin therapy in people judged to be at 'high' 10 year CVD risk. Original risk algorithms were recalibrated using the risk factor profile and CVD incidence of target populations. The four algorithms had similar risk discrimination. Before recalibration, FRS, SCORE, and PCE over-predicted CVD risk on average by 10%, 52%, and 41%, respectively, whereas RRS under-predicted by 10%. Original versions of algorithms classified 29-39% of individuals aged ≥40 years as high risk. By contrast, recalibration reduced this proportion to 22-24% for every algorithm. We estimated that to prevent one CVD event, it would be necessary to initiate statin therapy in 44-51 such individuals using original algorithms, in contrast to 37-39 individuals with recalibrated algorithms.Conclusion: Before recalibration, the clinical performance of four widely used CVD risk algorithms varied substantially. By contrast, simple recalibration nearly equalized their performance and improved modelled targeting of preventive action to clinical need.

KW - Calibration

KW - Cardiovascular disease

KW - Discrimination

KW - Risk algorithms

KW - Risk prediction

U2 - 10.1093/eurheartj/ehy653

DO - 10.1093/eurheartj/ehy653

M3 - Journal article

C2 - 30476079

VL - 40

SP - 621

EP - 631

JO - European Heart Journal

JF - European Heart Journal

SN - 0195-668X

IS - 7

ER -