Determining the level of agreement for atherosclerotic cardiovascular disease risk stratification between coronary artery calcium score and traditional cardiovascular risk models.

Date of Award


Document Type


Degree Name

Master of Medicine (MMed)

First Supervisor/Advisor

Prof. Sudhir Vinayak

Second Supervisor/Advisor

Dr. Harun Otieno

Third Supervisor/Advisor

Dr. Kevin Were


Imaging and Diagnostic Radiology (East Africa)


Introduction: Estimating the risk of future cardiovascular (CV) events is an essential step in the management and prevention of cardiovascular diseases. Many cardiovascular risk tools are available such as the Framingham Risk Score (FRS), the American Heart Association/ American College of Cardiology (AHA/ACC) and the International Society of Hypertension and World Health Organization (ISH/WHO) risk charts. Regrettably, none of these tools has yet to be validated based on data from our Kenyan population. This study, therefore, sought to compare the accuracy of FRS and ACC/AHA in a Kenyan population. The risk estimates derived from the CV tools were correlated with coronary artery calcium score (CACS), a well-established reliable predictor of future risk of cardiovascular events.

Objective: Determining the level of agreement between coronary calcium score and traditional cardiovascular risk models for coronary artery disease risk stratification in a multiethnic population at a tertiary healthcare institution in Kenya.

Design: A cross-sectional study

Methods: Data were collected retrospectively from the medical records of 200 patients referred to the Radiology department for a CT coronary artery calcium score.190 patients met the inclusion criteria. Comparisons for risk stratification were made according to FRS, ACC/AHA, WHO/ISH and CACS and the agreement (Kappa) and correlation (Spearman rho) between them were calculated. Statistical significance was set at p

Results: There was poor agreement (Kappa >0.191) between CACS and the clinical CVD riskmodels in the Kenyan population studied. In relation to this, 83.6%, 81.8 % and 66.2% of the intermediate risk group according to FRS, ACC/AHA and WHO/ISH, respectively, were down-classified by CACS. Moreover, 81.6%, 84.6% and 66.7% of those who would qualify for aggressive management as per FRS, ACC/AHA, and WHO/ISH risk-based algorithms, respectively, would not qualify for the same management as per the CACS.

Conclusion: The poor agreement between CACS and these clinical CVD risk scores suggests that the clinical CVD risk tools currently used in our Kenyan population might be incorrectly stratifying CV risk in patients. This highlights the need to externally validate these CVD risk models in our population to better risk predictions and set appropriate population-wide thresholds that are essential to aid better clinical decision making.

This document is available in the relevant AKU library