Title : Comparison of cardiovascular risk assessment tools: Laboratory vs non-laboratory tools
Abstract:
Background: Cardiovascular disease risk assessment tools are algorithms used to predict an individual’s chance of developing a cardiovascular disease within a specific period. It has been shown that awareness of one’s risk of developing a particular disease often result in positive behavior aimed at preventing that disease. Nonetheless, this positive behavioral change would be most beneficial if the prediction tool is as accurate as possible, thus the need to compare various tools and their performance. The aim of this research is to evaluate correlation of lab-based tool and non-lab-based tool to validate the use of the non-lab-based tool in resource-poor setting.
Methods: This is a hospital-based cross-sectional study among 94 diabetic patients attending a diabetic clinic in Kumasi, Ghana. Data collected includes sociodemographic information, anthropometry, medical history and lipid profile which were then used to compute the cardiovascular risk score using the pooled cohort equation (PCE) and WHO non-laboratory scoring tools.
Result: The mean score according to the PCE calculator and WHO/ISH non-laboratory chart was 13.5% [CI 95: 10.8 - 16.1] and 7.2% [CI 95: 6.2 - 8.1] respectively. Mean PCE score was significantly higher than mean WHO score, p-value < 0.001. There was a strong relationship between the unclassified risk scores for WHO and PCE calculators (r=0.899, adjusted-p<0.001, rho=0.934, adjusted-p<0.001). Using a cut-off of 10% to dichotomize the risk into low and moderate/high, the Cohen’s Kappa analysis showed moderate agreement between the two algorithms (Agreement =73.4% Kappa=0.455 [95% CI: 0.285 – 0.624], p<0.001). Using the PCE risk prediction as the reference with a cutoff of 10%, the non-laboratory WHO/ISH showed good discrimination, the overall AUC was 0.95 [CI 95: 0.91 – 0.99]. The model correctly classified 89.36% of the participants with a sensitivity of 86.67% and specificity of 91.84%.
Conclusion: There is significant association between pooled cohort equation and non-laboratory WHO risk assessment though non-lab based tool underestimated the CVD risk compared to PCE. The non-lab WHO CVD risk calculator can be used in resource-poor settings where serum cholesterol is not readily available.
