Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV

Rosan A. Van Zoest, Matthew Law, Caroline A. Sabin, Ilonca Vaartjes, Marc Van Der Valk, Joop E. Arends, Peter Reiss, Ferdinand W. Wit, on behalf of the ATHENA National Observational HIV Cohort

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background: People living with HIV (PLWH) experience a higher cardiovascular disease (CVD) risk. Yet, traditional algorithms are often used to estimate CVD risk. We evaluated the performance of 4 commonly used algorithms. Setting: The Netherlands. Methods: We used data from 16,070 PLWH aged ≥18 years, who were in care between 2000 and 2016, had no pre-existing CVD, had initiated first combination antiretroviral therapy >1 year ago, and had available data on CD4 count, smoking status, cholesterol, and blood pressure. Predictive performance of 4 algorithms [Data Collection on Adverse Effects of Anti-HIV Drugs Study (D:A:D); Systematic COronary Risk Evaluation adjusted for national data (SCORE-NL); Framingham CVD Risk Score (FRS); and American College of Cardiology and American Heart Association Pooled Cohort Equations (PCE)] was evaluated using a Kaplan-Meier approach. Model discrimination was assessed using Harrell's C-statistic. Calibration was assessed using observed-versus-expected ratios, calibration plots, and Greenwood-Nam-D'Agostino goodness-of-fit tests. Results: All algorithms showed acceptable discrimination (Harrell's C-statistic 0.73-0.79). On a population level, D:A:D, SCORE-NL, and PCE slightly underestimated, whereas FRS slightly overestimated CVD risk (observed-versus-expected ratios 1.35, 1.38, 1.14, and 0.92, respectively). D:A:D, FRS, and PCE best fitted our data but still yielded a statistically significant lack of fit (Greenwood-Nam-D'Agostino χ2 ranged from 24.57 to 34.22, P < 0.05). Underestimation of CVD risk was particularly observed in low-predicted CVD risk groups. Conclusions: All algorithms perform reasonably well in PLWH, with SCORE-NL performing poorest. Prediction algorithms are useful for clinical practice, but clinicians should be aware of their limitations (ie, lack of fit and slight underestimation of CVD risk in low-risk groups).

Original languageEnglish
Pages (from-to)562-571
Number of pages10
JournalJournal of Acquired Immune Deficiency Syndromes
Volume81
Issue number5
DOIs
Publication statusPublished - 15 Aug 2019

Cite this

Van Zoest, R. A., Law, M., Sabin, C. A., Vaartjes, I., Van Der Valk, M., Arends, J. E., ... on behalf of the ATHENA National Observational HIV Cohort (2019). Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV. Journal of Acquired Immune Deficiency Syndromes, 81(5), 562-571. https://doi.org/10.1097/QAI.0000000000002069
Van Zoest, Rosan A. ; Law, Matthew ; Sabin, Caroline A. ; Vaartjes, Ilonca ; Van Der Valk, Marc ; Arends, Joop E. ; Reiss, Peter ; Wit, Ferdinand W. ; on behalf of the ATHENA National Observational HIV Cohort. / Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV. In: Journal of Acquired Immune Deficiency Syndromes. 2019 ; Vol. 81, No. 5. pp. 562-571.
@article{446dc25ea3c84ba1ac157135315f464b,
title = "Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV",
abstract = "Background: People living with HIV (PLWH) experience a higher cardiovascular disease (CVD) risk. Yet, traditional algorithms are often used to estimate CVD risk. We evaluated the performance of 4 commonly used algorithms. Setting: The Netherlands. Methods: We used data from 16,070 PLWH aged ≥18 years, who were in care between 2000 and 2016, had no pre-existing CVD, had initiated first combination antiretroviral therapy >1 year ago, and had available data on CD4 count, smoking status, cholesterol, and blood pressure. Predictive performance of 4 algorithms [Data Collection on Adverse Effects of Anti-HIV Drugs Study (D:A:D); Systematic COronary Risk Evaluation adjusted for national data (SCORE-NL); Framingham CVD Risk Score (FRS); and American College of Cardiology and American Heart Association Pooled Cohort Equations (PCE)] was evaluated using a Kaplan-Meier approach. Model discrimination was assessed using Harrell's C-statistic. Calibration was assessed using observed-versus-expected ratios, calibration plots, and Greenwood-Nam-D'Agostino goodness-of-fit tests. Results: All algorithms showed acceptable discrimination (Harrell's C-statistic 0.73-0.79). On a population level, D:A:D, SCORE-NL, and PCE slightly underestimated, whereas FRS slightly overestimated CVD risk (observed-versus-expected ratios 1.35, 1.38, 1.14, and 0.92, respectively). D:A:D, FRS, and PCE best fitted our data but still yielded a statistically significant lack of fit (Greenwood-Nam-D'Agostino χ2 ranged from 24.57 to 34.22, P < 0.05). Underestimation of CVD risk was particularly observed in low-predicted CVD risk groups. Conclusions: All algorithms perform reasonably well in PLWH, with SCORE-NL performing poorest. Prediction algorithms are useful for clinical practice, but clinicians should be aware of their limitations (ie, lack of fit and slight underestimation of CVD risk in low-risk groups).",
keywords = "HIV, cardiovascular disease, risk prediction algorithms",
author = "{Van Zoest}, {Rosan A.} and Matthew Law and Sabin, {Caroline A.} and Ilonca Vaartjes and {Van Der Valk}, Marc and Arends, {Joop E.} and Peter Reiss and Wit, {Ferdinand W.} and {on behalf of the ATHENA National Observational HIV Cohort} and S.E. Geerlings and JWR Hovius and TW Kuijpers and FJB Nellen and JM Prins and {Wiersinga de Vreede}, WJ and EJ Peters and {van Agtmael}, MA and MK Bomers and CW Ang and {van Houdt}, R and AM Pettersson and CMJE Vandenbroucke-Grauls and BB Wintermans",
year = "2019",
month = "8",
day = "15",
doi = "10.1097/QAI.0000000000002069",
language = "English",
volume = "81",
pages = "562--571",
journal = "Journal of Acquired Immune Deficiency Syndromes",
issn = "1525-4135",
publisher = "Lippincott Williams and Wilkins",
number = "5",

}

Van Zoest, RA, Law, M, Sabin, CA, Vaartjes, I, Van Der Valk, M, Arends, JE, Reiss, P, Wit, FW & on behalf of the ATHENA National Observational HIV Cohort 2019, 'Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV' Journal of Acquired Immune Deficiency Syndromes, vol. 81, no. 5, pp. 562-571. https://doi.org/10.1097/QAI.0000000000002069

Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV. / Van Zoest, Rosan A.; Law, Matthew; Sabin, Caroline A.; Vaartjes, Ilonca; Van Der Valk, Marc; Arends, Joop E.; Reiss, Peter; Wit, Ferdinand W.; on behalf of the ATHENA National Observational HIV Cohort.

In: Journal of Acquired Immune Deficiency Syndromes, Vol. 81, No. 5, 15.08.2019, p. 562-571.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Predictive Performance of Cardiovascular Disease Risk Prediction Algorithms in People Living With HIV

AU - Van Zoest, Rosan A.

AU - Law, Matthew

AU - Sabin, Caroline A.

AU - Vaartjes, Ilonca

AU - Van Der Valk, Marc

AU - Arends, Joop E.

AU - Reiss, Peter

AU - Wit, Ferdinand W.

AU - on behalf of the ATHENA National Observational HIV Cohort

AU - Geerlings, S.E.

AU - Hovius, JWR

AU - Kuijpers, TW

AU - Nellen, FJB

AU - Prins, JM

AU - Wiersinga de Vreede, WJ

AU - Peters, EJ

AU - van Agtmael, MA

AU - Bomers, MK

AU - Ang, CW

AU - van Houdt, R

AU - Pettersson, AM

AU - Vandenbroucke-Grauls, CMJE

AU - Wintermans, BB

PY - 2019/8/15

Y1 - 2019/8/15

N2 - Background: People living with HIV (PLWH) experience a higher cardiovascular disease (CVD) risk. Yet, traditional algorithms are often used to estimate CVD risk. We evaluated the performance of 4 commonly used algorithms. Setting: The Netherlands. Methods: We used data from 16,070 PLWH aged ≥18 years, who were in care between 2000 and 2016, had no pre-existing CVD, had initiated first combination antiretroviral therapy >1 year ago, and had available data on CD4 count, smoking status, cholesterol, and blood pressure. Predictive performance of 4 algorithms [Data Collection on Adverse Effects of Anti-HIV Drugs Study (D:A:D); Systematic COronary Risk Evaluation adjusted for national data (SCORE-NL); Framingham CVD Risk Score (FRS); and American College of Cardiology and American Heart Association Pooled Cohort Equations (PCE)] was evaluated using a Kaplan-Meier approach. Model discrimination was assessed using Harrell's C-statistic. Calibration was assessed using observed-versus-expected ratios, calibration plots, and Greenwood-Nam-D'Agostino goodness-of-fit tests. Results: All algorithms showed acceptable discrimination (Harrell's C-statistic 0.73-0.79). On a population level, D:A:D, SCORE-NL, and PCE slightly underestimated, whereas FRS slightly overestimated CVD risk (observed-versus-expected ratios 1.35, 1.38, 1.14, and 0.92, respectively). D:A:D, FRS, and PCE best fitted our data but still yielded a statistically significant lack of fit (Greenwood-Nam-D'Agostino χ2 ranged from 24.57 to 34.22, P < 0.05). Underestimation of CVD risk was particularly observed in low-predicted CVD risk groups. Conclusions: All algorithms perform reasonably well in PLWH, with SCORE-NL performing poorest. Prediction algorithms are useful for clinical practice, but clinicians should be aware of their limitations (ie, lack of fit and slight underestimation of CVD risk in low-risk groups).

AB - Background: People living with HIV (PLWH) experience a higher cardiovascular disease (CVD) risk. Yet, traditional algorithms are often used to estimate CVD risk. We evaluated the performance of 4 commonly used algorithms. Setting: The Netherlands. Methods: We used data from 16,070 PLWH aged ≥18 years, who were in care between 2000 and 2016, had no pre-existing CVD, had initiated first combination antiretroviral therapy >1 year ago, and had available data on CD4 count, smoking status, cholesterol, and blood pressure. Predictive performance of 4 algorithms [Data Collection on Adverse Effects of Anti-HIV Drugs Study (D:A:D); Systematic COronary Risk Evaluation adjusted for national data (SCORE-NL); Framingham CVD Risk Score (FRS); and American College of Cardiology and American Heart Association Pooled Cohort Equations (PCE)] was evaluated using a Kaplan-Meier approach. Model discrimination was assessed using Harrell's C-statistic. Calibration was assessed using observed-versus-expected ratios, calibration plots, and Greenwood-Nam-D'Agostino goodness-of-fit tests. Results: All algorithms showed acceptable discrimination (Harrell's C-statistic 0.73-0.79). On a population level, D:A:D, SCORE-NL, and PCE slightly underestimated, whereas FRS slightly overestimated CVD risk (observed-versus-expected ratios 1.35, 1.38, 1.14, and 0.92, respectively). D:A:D, FRS, and PCE best fitted our data but still yielded a statistically significant lack of fit (Greenwood-Nam-D'Agostino χ2 ranged from 24.57 to 34.22, P < 0.05). Underestimation of CVD risk was particularly observed in low-predicted CVD risk groups. Conclusions: All algorithms perform reasonably well in PLWH, with SCORE-NL performing poorest. Prediction algorithms are useful for clinical practice, but clinicians should be aware of their limitations (ie, lack of fit and slight underestimation of CVD risk in low-risk groups).

KW - HIV

KW - cardiovascular disease

KW - risk prediction algorithms

UR - http://www.scopus.com/inward/record.url?scp=85069888409&partnerID=8YFLogxK

U2 - 10.1097/QAI.0000000000002069

DO - 10.1097/QAI.0000000000002069

M3 - Article

VL - 81

SP - 562

EP - 571

JO - Journal of Acquired Immune Deficiency Syndromes

JF - Journal of Acquired Immune Deficiency Syndromes

SN - 1525-4135

IS - 5

ER -