Which Office-Based Cardiovascular Risk Score is Suitable for Pokhareli Nepalese: Globorisk, WHO CVD, or Framingham?
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Introduction
Cardiovascular diseases (CVDs) are the leading cause of death. In resource-limited settings, office-based methods offer cost-effective alternatives to laboratory-based methods for primary prevention. Although the WHO Charts are used in the PEN (Package of Essential Non-communicable Diseases) program, their predictability is limited. In Pokhara Metropolitan, Nepal, this study compared the risk scores and agreement levels of three office-based CVD risk prediction models—WHO CVD, Globorisk, and Framingham Risk Score (FRS).
Methods
A community-based, cross-sectional study was conducted among 532 individuals (≥30 years) from Pokhara. Sociodemographic data and risk factors were collected through interviews, and anthropometric measurements. Descriptive statistics, t-tests, ANOVA, and Cohen’s kappa were used to compare risk categories and agreement levels, while linear regression analyzed trends across algorithms.
Results
CVD risk estimates differed significantly: WHO CVD (4.51 ± 3.46), Globorisk (7.35 ± 7.14), and FRS (9.59 ± 8.34), (F-statistics = 78.04, p <.01). Globorisk and the FRS showed very good model fits (R²= 85% and 90% respectively), whereas the WHO CVD showed an excellent model-fit (R²= 94%), doubting overfitting. The level of agreement was fair between Globorisk and WHO CVD (Kappa = 0.327, p <.01), slight between WHO and FRS (Kappa = 0.19, p < .01), and moderate between Globorisk and FRS (Kappa = 0.48, p < .01). The highest agreement was between Globorisk and FRS, particularly for females, and the lowest between WHO and FRS, especially for males. Ethnicity, education, marital status, and socio-economic factors were associated with CVD risk.
Conclusion
CVD risk predictions varied, with FRS predicting the highest risk and WHO CVD the lowest; with a difference of more than five percent. Variation was highest in low-risk (23.9%) and lowest in moderate-risk (10.4%). Ethnicity, occupation, education, marital status, and socio-economic factors should be considered before using appropriate algorithms.
SUMMARY
What is already known on this topic
WHO/ISH 2007 (World Health Organization/International Society of Hypertension) cardiovascular disease (CVD) risk charts are currently used in Nepal’s Primary Health Care Centers. These charts have been updated in 2019 (as WHO CVD 2019 risk charts). However, these both charts still have limitations, particularly in underestimating the risk.
What this study adds
This study compared risk scores provided by WHO CVD with other non-laboratory algorithms like Globorisk and the Framingham Risk Score (FRS). Our analysis included Cohen’s kappa statistics to find the agreement level among the risk categories and found differences in the mean risk scores among them. Ethnicity, educational level, and the marital status should also be considered when assessing CVD risk.
How this study might affect research, practice, or policy
Our findings highlight the variability in CVD risk predictions, with Globorisk demonstrating the most consistent risk estimates, while FRS overestimated and WHO CVD underestimated moderate and high risk. Globorisk may be a better predicting office-based model for Pokhareli Nepalese.
