New CV Risk Model Predicts Recurrent MACE at 1 Year in Post-MI Patients

The risk model includes 19 variables, however, and one physician fears it might be too complicated for everyday use.

New CV Risk Model Predicts Recurrent MACE at 1 Year in Post-MI Patients

The latest addition to the array of cardiovascular risk prediction tools available to clinicians draws on 19 different variables and, according to the results of a new study, clearly identifies acute MI survivors at risk of major adverse cardiovascular events 1 year after their initial attack.

The risk score, which includes traditional cardiovascular risk factors such as age, hypertension, and renal function, as well as others less commonly used such as college degree and white blood cell (WBC) count, had “good discrimination, calibration, and fit,” write Yun Wang, PhD (Harvard T.H. Chan School of Public Health, Boston, MA), and colleagues August 10, 2018, in JAMA: Open Network. 

For patients at the highest risk based on the extensive risk-assessment model, the observed rate of cardiovascular events at 1 year reached as high as 32.7%. For those at the lowest risk, the observed rate of cardiovascular events was 1.2%.

The risk model and its corresponding risk scores allow clinicians to identify patients who are at heightened risk of 1-year cardiovascular events and might also help patients understand their risk of not only death but also adverse events that could impair their quality of life,” write Wang and colleagues. “The ability to identify individuals with the highest risk of long-term cardiovascular events after acute MI may aid in the provision of targeted, intensive, and higher-quality longitudinal care following discharge.”

Moving Beyond Morality Alone

Traditional studies on long-term outcomes after an acute MI predominately focus on mortality, say the researchers, but other endpoints, such as recurrent MI, stroke, and heart failure, negatively impact quality of life. Given that, they designed the new model to assess the risk of long-term outcomes beyond mortality, so as “to capture the complete health care experience of patients, particularly in non-Western countries.”

The study consisted of 4,227 patients participating in the China Patient-Centered Evaluative Assessment of Cardiac Events (PEACE) Prospective AMI study. All patients were discharged from acute-care hospitals between 2013 and 2014 and randomly divided into three exclusive cohorts: the training (n = 2,113), test (n = 1,057), and validation (n = 1,057) samples. For the training, test, and validation groups, the 1-year rate of cardiovascular events was 8.1%, 9.0%, and 6.4%, respectively (P = 0.36).

The 19 risk factors, which are based on 15 distinct variables, include patient demographics (age and college degree), comorbidities (prior acute MI, prior ventricular tachycardia or fibrillation, hypertension, angina), hospital diagnoses and test results (ejection fraction < 40% or no ejection fraction measurement, renal dysfunction, heart rate > 90 beats per minute, blood glucose > 216 mg/dL, systolic blood pressure < 100 mm Hg, WBC 6000-12,000/µL, and WBC > 12,000/µL), access to care (prearrival medical assistance and time from symptom onset to admission > 4 hours), and in-hospital complications.

In combining the variables into a single risk score, patients were classified as high, average, or low risk for recurrent cardiovascular events at 1 year. Age, ejection fraction, WBC, prior ventricular tachycardia/fibrillation, prior angina, and heart rate were the five risk factors associated with the greatest risk of cardiovascular events at 1 year.  

The C statistic, which is a measure of discrimination, was 0.79 (a value of 1 indicates the model perfectly predicts who will have an event and who will not), and the time-dependent ROC curve ranged from 0.79 to 0.75. The risk model performed similarly in the test and validation samples as it did in the training cohort.

Andrew Foy, MD (Penn State Health, Hershey, PA), who was not involved in the study, praised the investigators for their excellent methodology, as well as for identifying several variables not universally recognized as risk factors for recurrent major adverse cardiovascular events. For example, elevated WBC count is not typically used in risk-assessment models, but Foy said he considers it an important piece of information.

However, Foy told TCTMD that the risk model might simply be too complicated for routine clinical use. He added that the GRACE risk score also predicts the risk of recurrent cardiovascular events, not just mortality, but said that risk score is easier to use.

“I think a weakness of this model is the amount of variables accounted for,” he said. “Adding more independent variables increases the predictive accuracy to some extent—not sure how much though—at the expense of functionality. It would be interesting to directly compare it to a less complex model.”

Foy is also uncertain how information gained from the risk model would translate into better care. The management of risk factors known to increase the risk of cardiac events, such as diabetes and hypertension, would be recommended independently of the individuals’ 1-year risk of recurrent cardiovascular events. “The fact that a person is at higher risk for major adverse cardiovascular events at 1-year compared to another really shouldn’t change physician or the health system’s behavior toward that individual,” said Foy.

  • Yang and Foy report no relevant conflicts of interest.

We Recommend