Is the SMART risk prediction model ready for real-world implementation? A validation study in a routine care setting of approximately 380 000 individuals

Using machine learning to predict acute myocardial infarction and ischemic heart disease in primary care cardiovascular patients

BACKGROUND

Early recognition, which preferably happens in primary care, is the most important tool to combat cardiovascular disease (CVD). This study aims to predict acute myocardial infarction (AMI) and ischemic heart disease (IHD) using Machine Learning (ML) in primary care cardiovascular patients. We compare the ML-models' performance with that of the common SMART algorithm and discuss clinical implications.

METHODS AND RESULTS

Patient-level medical record data (n = 13,218) collected between 2011-2021 from 90 GP-practices were used to construct two random forest models (one for AMI and one for IHD) as well as a linear model based on the SMART risk prediction algorithm as a suitable comparator. The data contained patient-level predictors, including demographics, procedures, medications, biometrics, and diagnosis. Temporal cross-validation was used to assess performance. Furthermore, predictors that contributed most to the ML-models' accuracy were identified. The ML-model predicting AMI had an accuracy of 0.97, a sensitivity of 0.67, a specificity of 1.00 and a precision of 0.99. The AUC was 0.96 and the Brier score was 0.03. The IHD-model had similar performance. In both ML-models anticoagulants/antiplatelet use, systolic blood pressure, mean blood glucose, and eGFR contributed most to model accuracy. For both outcomes, the SMART algorithm was substantially outperformed by ML on all metrics.

CONCLUSION

Our findings underline the potential of using ML for CVD prediction purposes in primary care, although the interpretation of predictors can be difficult. Clinicians, patients, and researchers might benefit from transitioning to using ML-models in support of individualized predictions by primary care physicians and subsequent (secondary) prevention.

Estimating potential cardiovascular health benefits of improved population level control of LDL cholesterol through a twice-yearly siRNA-based approach: A simulation study of a health-system level intervention

BACKGROUND AND AIMS

Inclisiran, an siRNA therapy, consistently reduces low-density lipoprotein cholesterol (LDL-C) with twice-yearly dosing. Potential cardiovascular benefits of implementing inclisiran at a population level, added to statins, were evaluated through simulation.

METHODS

For each participant in the ORION-10 and ORION-11 trials comparing inclisiran with placebo, baseline 10-year cardiovascular risk was estimated using the SMART equation. The time-adjusted LDL-C difference from baseline observed 90-540 days after baseline was assumed to persist and used to estimate potential reduction in 10-year cardiovascular risk. Impact on 500,000 ORION-like individuals was simulated with Monte-Carlo.

RESULTS

Mean baseline LDL-C and predicted 10-year major vascular risk among patients randomized to inclisiran (n = 1288) versus placebo (n = 1264) were 2.66 mmol/L versus 2.60 mmol/L and 24.9% versus 24.6%, respectively. Placebo-corrected time-adjusted absolute reduction in LDL-C with inclisiran was -1.32 mmol/L (95% CI -1.37 to -1.26; p < 0.001), which predicted a 10-year cardiovascular risk of 18.1% with inclisiran versus 24.7% with placebo (absolute difference [95% CI], -6.99% [-7.33 to -6.66]; p < 0.001) NNT 15. Extrapolating to 500,000 inclisiran-treated individuals, the model predicted large population shifts towards lower quintiles of risk with fewer remaining in high-risk categories; 3350 to 471 (≥80% risk), 11,793 to 3332 (60-<80% risk), 52,142 to 22,665 (40-<60% risk), 197,752 to 141,014 (20-<40% risk), and more moving into the lowest risk category (<20%) from 234,963 to 332,518.

CONCLUSIONS

Meaningful gains in population health might be achieved over 10 years by implementing at-scale approaches capable of providing substantial and sustained reductions in LDL-C beyond those achievable with statins.

Validating the SMART2 Score in a Racially Diverse High-Risk Nationwide Cohort of Patients Receiving Coronary Artery Bypass Grafting

Background We tested the potential of the Secondary Manifestations of Arterial Disease (SMART2) risk score for use in patients undergoing coronary artery bypass grafting. Methods and Results We conducted an external validation of the SMART2 score in a racially diverse high-risk national cohort (2010-2019) that underwent isolated coronary artery bypass grafting. We calculated the preoperative SMART2 score and modeled the 5-year major adverse cardiovascular event (cardiovascular mortality+myocardial infarction+stroke) incidence. We evaluated SMART2 score discrimination at 5 years using c-statistic and calibration with observed/expected ratio and calibration plots. We analyzed the potential clinical benefit using decision curves. We repeated these analyses in clinical subgroups, diabetes, chronic kidney disease, and polyvascular disease, and separately in White and Black patients. In 27 443 (mean age, 65 years; 10% Black individuals) US veterans undergoing coronary artery bypass grafting (2010-2019) nationwide, the 5-year major adverse cardiovascular event rate was 25%; 27% patients were in high predicted risk (>30% 5-year major adverse cardiovascular events). SMART2 score discrimination (c-statistic: 64) was comparable to the original study (c-statistic: 67) and was best in patients with chronic kidney disease (c-statistic: 66). However, it underpredicted major adverse cardiovascular event rates in the whole cohort (observed/expected ratio, 1.45) as well as in all studied subgroups. The SMART2 score performed better in White than Black patients. On decision curve analysis, the SMART2 score provides a net benefit over a wide range of risk thresholds. Conclusions The SMART2 model performs well in a racially diverse coronary artery bypass grafting cohort, with better predictive capabilities at the upper range of baseline risk, and can therefore be used to guide secondary preventive pharmacotherapy.

A Novel Approach to Developing Disease and Outcome-Specific Social Risk Indices

INTRODUCTION

A variety of industry composite indices are employed within health research in risk-adjusted outcome measures and to assess health-related social needs. During the COVID-19 pandemic, the relationships among risk adjustment, clinical outcomes, and composite indices of social risk have become relevant topics for research and healthcare operations. Despite the widespread use of these indices, composite indices are often comprised of correlated variables and therefore may be affected by information duplicity of their underlying risk factors.

METHODS

A novel approach is proposed to assign outcome- and disease group-driven weights to social risk variables to form disease and outcome-specific social risk indices and apply the approach to the county-level Centers for Disease Control and Prevention social vulnerability factors for demonstration. The method uses a subset of principal components reweighed through Poisson rate regressions while controlling for county-level patient mix. The analyses use 6,135,302 unique patient encounters from 2021 across seven disease strata.

RESULTS

The reweighed index shows reduced root mean squared error in explaining county-level mortality in five of the seven disease strata and equivalent performance in the remaining strata compared with the reduced root mean squared error using the current Centers for Disease Control and Prevention Social Vulnerability Index as a benchmark.

CONCLUSIONS

A robust method is provided, designed to overcome challenges with current social risk indices, by accounting for redundancy and assigning more meaningful disease and outcome-specific variable weights.

Treatment gaps in the implementation of LDL cholesterol control among high- and very high-risk patients in Europe between 2020 and 2021: the multinational observational SANTORINI study

Background

European data pre-2019 suggest statin monotherapy is the most common approach to lipid management for preventing cardiovascular (CV) events, resulting in only one-fifth of high- and very high-risk patients achieving the 2019 ESC/EAS recommended low-density lipoprotein cholesterol (LDL-C) goals. Whether the treatment landscape has evolved, or gaps persist remains of interest.

Methods

Baseline data are presented from SANTORINI, an observational, prospective study that documents the use of lipid-lowering therapies (LLTs) in patients ≥18 years at high or very high CV risk between 2020 and 2021 across primary and secondary care settings in 14 European countries.

Findings

Of 9602 enrolled patients, 9044 with complete data were included (mean age: 65.3 ± 10.9 years; 72.6% male). Physicians reported using 2019 ESC/EAS guidelines as a basis for CV risk classification in 52.0% (4706/9044) of patients (overall: high risk 29.2%; very high risk 70.8%). However, centrally re-assessed CV risk based on 2019 ESC/EAS guidelines suggested 6.5% (308/4706) and 91.0% (4284/4706) were high- and very high-risk patients, respectively. Overall, 21.8% of patients had no documented LLTs, 54.2% were receiving monotherapy and 24.0% combination LLT. Median (interquartile range [IQR]) LDL-C was 2.1 (1.6, 3.0) mmol/L (82 [60, 117] mg/dL), with 20.1% of patients achieving risk-based LDL-C goals as per the 2019 ESC/EAS guidelines.

Interpretation

At the time of study enrolment, 80% of high- and very high-risk patients failed to achieve 2019 ESC/EAS guidelines LDL-C goals. Contributory factors may include CV risk underestimation and underutilization of combination therapies. Further efforts are needed to achieve current guideline-recommended LDL-C goals.

Trial registration

ClinicalTrials.gov Identifier: NCT04271280.

Funding

This study is funded by Daiichi Sankyo Europe GmbH, Munich, Germany.

Inverse association between apolipoprotein C-II and cardiovascular mortality: role of lipoprotein lipase activity modulation

AIMS

Apolipoprotein C-II (ApoC-II) is thought to activate lipoprotein lipase (LPL) and is therefore a possible target for treating hypertriglyceridemia. Its relationship with cardiovascular risk has not been investigated in large-scale epidemiologic studies, particularly allowing for apolipoprotein C-III (ApoC-III), an LPL antagonist. Furthermore, the exact mechanism of ApoC-II-mediated LPL activation is unclear.

METHODS AND RESULTS

ApoC-II was measured in 3141 LURIC participants of which 590 died from cardiovascular diseases during a median (inter-quartile range) follow-up of 9.9 (8.7-10.7) years. Apolipoprotein C-II-mediated activation of the glycosylphosphatidylinositol high-density lipoprotein binding protein 1 (GPIHBP1)-LPL complex was studied using enzymatic activity assays with fluorometric lipase and very low-density lipoprotein (VLDL) substrates. The mean ApoC-II concentration was 4.5 (2.4) mg/dL. The relationship of ApoC-II quintiles with cardiovascular mortality exhibited a trend toward an inverse J-shape, with the highest risk in the first (lowest) quintile and lowest risk in the middle quintile. Compared with the first quintile, all other quintiles were associated with decreased cardiovascular mortality after multivariate adjustments including ApoC-III as a covariate (all P < 0.05). In experiments using fluorometric substrate-based lipase assays, there was a bell-shaped relationship for the effect of ApoC-II on GPIHBP1-LPL activity when exogenous ApoC-II was added. In ApoC-II-containing VLDL substrate-based lipase assays, GPIHBP1-LPL enzymatic activity was almost completely blocked by a neutralizing anti-ApoC-II antibody.

CONCLUSION

The present epidemiologic data suggest that increasing low circulating ApoC-II levels may reduce cardiovascular risk. This conclusion is supported by the observation that optimal ApoC-II concentrations are required for maximal GPIHBP1-LPL enzymatic activity.

Estimated cardiovascular benefits of bempedoic acid in patients with established cardiovascular disease

Background and aims

Cardiovascular outcomes trials have demonstrated that lowering low-density lipoprotein cholesterol (LDL-C) reduces the risk for future cardiovascular events. We assessed the potential cardiovascular benefits of bempedoic acid through a simulation study in patients with atherosclerotic cardiovascular disease (ASCVD) and elevated LDL-C.

Methods

The validated SMART prediction model was used to estimate the baseline 10-year risk of three-point major adverse cardiovascular events (cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke) in patients with ASCVD who were enrolled in four Phase 3, randomized, placebo-controlled bempedoic acid studies. The predicted change in 10-year cardiovascular risk associated with bempedoic acid was estimated for each patient based on the Cholesterol Treatment Trialists' model. Data were analyzed in two cohorts: Cohort 1 included mostly patients treated with moderate-high intensity statins, and Cohort 2 included patients who were intolerant of more than low-intensity statin.

Results

A total of 2884 patients were included in Cohort 1 and 226 in Cohort 2. Weighted average baseline 10-year cardiovascular event risk was 26.1% and 31.6% for Cohorts 1 and 2, respectively. The least squares mean percent difference (95% confidence interval (CI) of the predicted absolute change in 10-year cardiovascular event risk with bempedoic acid was -3.3% (-3.7% to -2.9%) for patients in Cohort 1 and -6.0% (-7.7% to -4.3%) for patients in Cohort 2 compared with placebo (p < 0.0001 for both).

Conclusions

Among patients with ASCVD who could potentially benefit from additional LDL-C lowering, our simulation predicted a lower absolute cardiovascular event risk after initiating bempedoic acid as compared with placebo.

Assessing Hospital Resource Utilization with Application to Imaging for Patients Diagnosed with Prostate Cancer

Resource utilization measures are typically modeled by relying on clinical characteristics. However, in some settings, those clinical markers are not available, and hospitals are unable to explore potential inefficiencies or resource misutilization. We propose a novel approach to exploring misutilization that solely relies on administrative data in the form of patient characteristics and competing resource utilization, with the latter being a novel addition. We demonstrate this approach in a 2019 patient cohort diagnosed with prostate cancer (n = 51,111) across 1056 U.S. healthcare facilities using Premier, Inc.’s (Charlotte, NC, USA) all payor databases. A multivariate logistic regression model was fitted using administrative information and competing resources utilization. A decision curve analysis informed by industry average standards of utilization allows for a definition of misutilization with regards to these industry standards. Odds ratios were extracted at the patient level to demonstrate differences in misutilization by patient characteristics, such as race; Black individuals experienced higher under-utilization compared to White individuals (p < 0.0001). Volume-adjusted Poisson rate regression models allow for the identification and ranking of facilities with large departures in utilization. The proposed approach is scalable and easily generalizable to other diseases and resources and can be complemented with clinical information from electronic health record information, when available.