Dataset Information

Machine Learning Outperforms ACC / AHA CVD Risk Calculator in MESA.

ABSTRACT: Background Studies have demonstrated that the current US guidelines based on American College of Cardiology/American Heart Association (ACC/AHA) Pooled Cohort Equations Risk Calculator may underestimate risk of atherosclerotic cardiovascular disease ( CVD ) in certain high-risk individuals, therefore missing opportunities for intensive therapy and preventing CVD events. Similarly, the guidelines may overestimate risk in low risk populations resulting in unnecessary statin therapy. We used Machine Learning ( ML ) to tackle this problem. Methods and Results We developed a ML Risk Calculator based on Support Vector Machines ( SVM s) using a 13-year follow up data set from MESA (the Multi-Ethnic Study of Atherosclerosis) of 6459 participants who were atherosclerotic CVD-free at baseline. We provided identical input to both risk calculators and compared their performance. We then used the FLEMENGHO study (the Flemish Study of Environment, Genes and Health Outcomes) to validate the model in an external cohort. ACC / AHA Risk Calculator, based on 7.5% 10-year risk threshold, recommended statin to 46.0%. Despite this high proportion, 23.8% of the 480 "Hard CVD " events occurred in those not recommended statin, resulting in sensitivity 0.76, specificity 0.56, and AUC 0.71. In contrast, ML Risk Calculator recommended only 11.4% to take statin, and only 14.4% of "Hard CVD " events occurred in those not recommended statin, resulting in sensitivity 0.86, specificity 0.95, and AUC 0.92. Similar results were found for prediction of "All CVD " events. Conclusions The ML Risk Calculator outperformed the ACC/AHA Risk Calculator by recommending less drug therapy, yet missing fewer events. Additional studies are underway to validate the ML model in other cohorts and to explore its ability in short-term CVD risk prediction.

SUBMITTER: Kakadiaris IA

PROVIDER: S-EPMC6404456 | biostudies-literature | 2018 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine Learning Outperforms ACC / AHA CVD Risk Calculator in MESA.

Kakadiaris Ioannis A IA Vrigkas Michalis M Yen Albert A AA Kuznetsova Tatiana T Budoff Matthew M Naghavi Morteza M

Journal of the American Heart Association 20181101 22

Background Studies have demonstrated that the current US guidelines based on American College of Cardiology/American Heart Association (ACC/AHA) Pooled Cohort Equations Risk Calculator may underestimate risk of atherosclerotic cardiovascular disease ( CVD ) in certain high-risk individuals, therefore missing opportunities for intensive therapy and preventing CVD events. Similarly, the guidelines may overestimate risk in low risk populations resulting in unnecessary statin therapy. We used Machin ...[more]

PMID: 30571498

Similar Datasets

Project description:BACKGROUND:The 2017 American College of Cardiology/American Heart Association (ACC/AHA) blood pressure (BP) guideline provides updated recommendations for antihypertensive medication initiation and intensification. OBJECTIVES:Determine the risk for cardiovascular disease (CVD) events among adults recommended and not recommended antihypertensive medication initiation or intensification by the 2017 ACC/AHA BP guideline. METHODS:The authors analyzed data for black and white REGARDS (REasons for Geographic And Racial Differences in Stroke) study participants (age ?45 years). Systolic BP (SBP) and diastolic BP (DBP) were measured twice at baseline (2003 to 2007) and averaged. Participants not taking (n = 14,039) and taking (n = 15,179) antihypertensive medication were categorized according to their recommendations for antihypertensive medication initiation and intensification by the 2017 ACC/AHA guideline. Overall, 4,094 CVD events (stroke, coronary heart disease, and heart failure) occurred by December 31, 2014. RESULTS:Among participants not taking antihypertensive medication, 34.4% were recommended pharmacological antihypertensive treatment initiation. The CVD event rate per 1,000 person-years among participants recommended antihypertensive medication initiation with SBP/DBP ?140/90 mm Hg was 22.7 (95% confidence interval [CI]: 20.3 to 25.0). Among participants with SBP/DBP 130 to 139/80 to 89 mm Hg, the CVD event rate was 20.5 (95% CI: 18.5 to 22.6) and 3.4 (95% CI: 2.4 to 4.4) for those recommended and not recommended antihypertensive medication initiation, respectively. Among participants taking antihypertensive medication, 62.8% were recommended treatment intensification. The CVD event rate per 1,000 person-years among participants recommended treatment intensification was 33.6 (95% CI: 31.5 to 35.6) and 22.4 (95% CI: 20.8 to 23.9) for those with SBP/DBP ?140/90 mm Hg and 130 to 139/80 to 89 mm Hg, respectively. CONCLUSIONS:Implementing the 2017 ACC/AHA guideline would direct antihypertensive medication initiation and intensification to adults with high CVD risk.

Project description:ObjectiveThe aim of the study is to compare maternal and neonatal outcomes among patients who are normotensive, hypertensive by Stage I American College of Cardiology-American Heart Association (ACC-AHA) criteria, and hypertensive by American College of Obstetricians and Gynecologists (ACOG) criteria.Study designSecondary analysis of a prospective first trimester cohort study between 2007 and 2010 at three institutions in Baltimore, MD, was conducted. Blood pressure at 11 to 14 weeks' gestation was classified as (1) normotensive (systolic blood pressure [SBP] <130 mm Hg and diastolic blood pressure [DBP] <80 mm Hg); (2) hypertensive by Stage I ACC-AHA criteria (SBP 130-139 mm Hg or DBP 80-89 mm Hg); or (3) hypertensive by ACOG criteria (SBP ≥140 mm Hg or DBP ≥90 mm Hg). Primary outcomes included preeclampsia, small for gestational age (SGA) neonate, and preterm birth.ResultsAmong 3,422 women enrolled, 2,976 with delivery data from singleton pregnancies of nonanomalous fetuses were included. In total, 20.2% met hypertension criteria (Stage I ACC-AHA n = 254, 8.5%; ACOG n = 347, 11.7%). The Stage I ACC-AHA group's risk for developing preeclampsia was threefold higher than the normotensive group (adjusted relative risk [aRR] 3.70, 95% confidence interval [CI] 2.40-5.70). The Stage I ACC-AHA group had lower preeclampsia risk than the ACOG group but the difference was not significant (aRR 0.87, 95% CI 0.55-1.37). The Stage I ACC-AHA group was more likely than the normotensive group to deliver preterm (aRR 1.44, 95% CI 1.02-2.01) and deliver an SGA neonate (aRR 1.51, 95% CI 1.07-2.12). The Stage I ACC-AHA group was less likely to deliver preterm compared with the ACOG group (aRR 0.65, 95% CI 0.45-0.93), but differences in SGA were not significant (aRR 1.31, 95% CI 0.84-2.03).ConclusionPregnant patients with Stage I ACC-AHA hypertension in the first trimester had higher rates of preeclampsia, preterm birth, and SGA neonates compared with normotensive women. Adverse maternal and neonatal outcomes were numerically lower in the Stage I ACC-AHA group compared with the ACOG group, but these comparisons only reached statistical significance for preterm birth. Optimal pregnancy management for first trimester Stage I ACC-AHA hypertension requires active study.Key points· Women with first trimester American College of Cardiology-American Heart Association (ACC-AHA) Stage I hypertension were more likely to develop preeclampsia, deliver preterm, and deliver a small-for-gestational age neonate than normotensive women.. · Women with first trimester American College of Obstetricians and Gynecologists (ACOG) hypertension (consistent with stage II ACC-AHA hypertension) had the highest numeric rate of adverse outcomes; however, compared with Stage I ACC-AHA hypertension, there was only statistically significant difference for preterm delivery.. · The risk profile for pregnant women with Stage I ACC-AHA hypertension and women with hypertension by conventional ACOG criteria may be more similar than previously understood..

Project description:There is limited information on predicting incident cardiovascular outcomes among high- to very high-risk populations such as the elderly (≥ 65 years) in the absence of prior cardiovascular disease and the presence of non-cardiovascular multi-morbidity. We hypothesized that statistical/machine learning modeling can improve risk prediction, thus helping inform care management strategies. We defined a population from the Medicare health plan, a US government-funded program mostly for the elderly and varied levels of non-cardiovascular multi-morbidity. Participants were screened for cardiovascular disease (CVD), coronary or peripheral artery disease (CAD or PAD), heart failure (HF), atrial fibrillation (AF), ischemic stroke (IS), transient ischemic attack (TIA), and myocardial infarction (MI) for a 3-yr period in the comorbid history. They were followed up for up to 45.2 months. Analyses included descriptive approaches in terms of incidence rates and density ratios, and inferential in terms of main effect statistical/complex machine learning modeling. The contemporary risk factors of interest spanned across the domains of comorbidity, lifestyle, and healthcare utilization history. The cohort consisted of 154,551 individuals (mean age 68.8 years; 62.2% female). The overall crude incidence rate of CVD events was 9.9 new cases per 100 person-years. The highest rates among its component outcomes were obtained for CAD or PAD (3.6 for each), followed by HF (2.2) and AF (1.8), then IS (1.3), and finally TIA (1.0) and MI (0.9).Model performance was modest in terms of discriminatory power (C index: 0.67, 95%CI 0.667-0.674 for training; and 0.668, 95%CI 0.663-0.673 for validation data), equal agreement between predicted and observed events for calibration purposes, and good clinical utility in terms of a net benefit of 15 true positives per 100 patients relative to the All-patient treatment strategy. Complex models based on machine learning algorithms yielded incrementally better discriminatory power and much improved goodness-of-fitness tests from those based on main effect statistical modeling. This Medicare population represents a highly vulnerable group for incident CVD events. This population would benefit from an integrated approach to their care and management, including attention to their comorbidities and lifestyle factors, as well as medication adherence.

Dataset Information

Machine Learning Outperforms ACC / AHA CVD Risk Calculator in MESA.

Publications

Machine Learning Outperforms ACC / AHA CVD Risk Calculator in MESA.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets