Diagnostic models of the pre-test probability of stable coronary artery disease: A systematic review

Abstract

A comprehensive search of PubMed and Embase was performed in January 2015 to examine the available literature on validated diagnostic models of the pre-test probability of stable coronary artery disease and to describe the characteristics of the models. Studies that were designed to develop and validate diagnostic models of pre-test probability for stable coronary artery disease were included. Data regarding baseline patient characteristics, procedural characteristics, modeling methods, metrics of model performance, risk of bias, and clinical usefulness were extracted. Ten studies involving the development of 12 models and two studies focusing on external validation were identified. Seven models were validated internally, and seven models were validated externally. Discrimination varied between studies that were validated internally (C statistic 0.66-0.81) and externally (0.49-0.87). Only one study presented reclassification indices. The majority of better performing models included sex, age, symptoms, diabetes, smoking, and hyperlipidemia as variables. Only two diagnostic models evaluated the effects on clinical decision making processes or patient outcomes. Most diagnostic models of the pre-test probability of stable coronary artery disease have had modest success, and very few present data regarding the effects of these models on clinical decision making processes or patient outcomes.

Machine learning of clinical variables and coronary artery calcium scoring for the prediction of obstructive coronary artery disease on coronary computed tomography angiography: analysis from the CONFIRM registry

“…[23][24][25] Additionally, several models have been validated in more than one external population, with a trend towards lower discriminative ability over the past few years. 26 Indeed, differences in derivation (utilization of various imaging modalities as well as different cut-off values for the definition of obstructive CAD, utilization of imputation methodologies for missing values), model complexity as well as inconsistent external validation often exist, which limit their utilization in routine practice. In an ever-changing environment where populations are longitudinally evolving as a result of changing dietary habits, environmental exposures, primordial and preventative practices, there is a need for comprehensive models that evolve over time.…”

“…The utility of likelihood analysis in the diagnosis of CAD, either on invasive coronary angiography or CCTA, has been an area of growing interest amongst the cardiovascular imaging community given the low prevalence of obstructive CAD in patients undergoing CCTA. 13,14,26 The overarching goals of cardiovascular imaging, specifically within the context of suspected CAD, are to identify individuals with high-risk anatomy and/or myocardial ischaemia that would improve prognosis with revascularization therapy as well as to improve the utilization of preventative therapies. Data from the National Cardiovascular Data Registry (NCDR) reveal that the diagnostic yield of invasive coronary angiography is low with 149 739/398 978 (37.6%) having obstructive CAD despite an 83.9% precatheterization rate of noninvasive testing.…”

Validated models for pre-test probability of stable coronary artery disease: a systematic review suggesting how to improve validation procedures

“…Unlike previous works, 20 external validation was primarily considered. We also included internal validation but limited it to k-fold cross-validation as a technique inspired by the same purposes of external validation.…”

“…Reasons of exclusion were: (i) acute coronary syndrome, unstable chest pain, a history of myocardial infarction or previous revascularisation; (ii) models that included a diagnostic procedure that do not reflect the usual practices of the first-line assessment; 3,10 (iii) models based on a single predictive variable; (iv) lack of clearly stated discrimination power. Unlike previous works, 20 external validation was primarily considered. We also included internal validation but limited it to k- fold cross-validation as a technique inspired by the same purposes of external validation.…”

Machine learning and medicine: book review and commentary

“…That also would require well controlled clinical studies to justify its use, and it seems that few such studies have been done with classifier based tests. A 2017 systematic review [ 9 ] concluded that mathematical models for pre-test prediction of outcomes of tests for stable CAD in cardiology had “only modest success”. No machine learning-based models met the inclusion criteria for acceptance in that review A 2017 study by Korley et al [ 10 ] assessed use of clinical risk factors (such as in the Z-Aldesani database) for diagnosing CAD as a pre-test selection tool.…”