Why the study?
The effectiveness of the new NCEP III guidelines to identify young adults at risk for a cardiac event was yet to be studied.
Does an XGBoost predictive model improve risk stratification for new-onset cardiovascular events in hypertensive patients compared to traditional algorithms?
Cohort (n=3,588)
No
Does an XGBoost predictive model improve risk stratification for new-onset cardiovascular events in hypertensive patients compared to traditional algorithms?
Captured external expert commentary on this paper, strongest first. Original sources are linked where available.
“First, most heart attacks occur in people with average levels of risk factors who would not be considered eligible for preventive treatment based on current guidelines. One study (1) suggested that only 25% of patients presenting with acute myocardial infarction (MI) would have been identified as high-risk by the National Cholesterol Education Program's Adult Treatment Panel III guidelines prior to the acute event.”
An XGBoost machine learning model utilizing longitudinal clinical data can accurately predict new-onset cardiovascular events in hypertensive patients, outperforming the traditional static SCORE2 algorithm.
OBJECTIVES: The purpose of this study was to investigate the utility of the new National Cholesterol Education Program (NCEP) III guidelines in a group of young adults. BACKGROUND: These guidelines have been hailed as an improvement in their potential to identify individuals at risk for coronary heart disease (CHD) complications. Compared with the NCEP II, the new guidelines will increase the number of patients who qualify for medical management. However, the effectiveness of these guidelines to identify young adults at risk for a cardiac event is yet to be studied. METHODS: A retrospective review of clinical data from young adults (age <or=55 years for men and <or=65 years for women) hospitalized for acute myocardial infarction over a three-year period was conducted. Patients with a history of CHD or CHD equivalent were excluded. Using the NCEP III guidelines, we calculated a 10-year risk for coronary events on all patients. RESULTS: A total of 222 patients met criteria for inclusion. The mean age was 50 years and 25% were women. Mean lipid levels were all within the normal range; however, rates of smoking and obesity were high. When the 10-year risk of these patients was stratified by the number of risk factors and low-density lipoprotein cholesterol level, only 25% met criteria to qualify for pharmacotherapy. For women in this population, only 18% met criteria for treatment. CONCLUSIONS: The new guidelines offer multiple new features but have a tendency to underappreciate the risk for disease in young adults. To improve performance in young adults, statistical adjustments may be necessary.
No takes yet. Share an insight, caveat, or question.
Akosah et al. (2003) conducted a cohort in Arterial Hypertension (n=3,588). XGBoost machine learning prediction model vs. SCORE2 algorithm was evaluated on Prediction of new-onset cardiovascular events (ischemic heart disease, cerebrovascular disease, or heart failure admission) (ROC AUC 0.856).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: