Keywords: CVD, SCORE, predictive validity, prevention

Aim:

The objective is to assess and compare the predictive validity of three risk prediction methodologies - the SCORE ESC 2016, the SCORE NFA, and the SCORE ESC 2019 - and to propose the optimal one for application in the Bulgarian population.

Method:

The design is a prospective cohort study, with a retrospective data collection. Respectively, a total of 1908, 1598 and 2458 patients from the three cohorts were followed up over a calendar year. Following the European Recommendations for Cardiovascular Risk Assessment SCORE according to ESC (2016), the National Framework Agreement (NFA) and ESC (2019) patients were divided into four risk categories. Descriptive statistics, non-parametric analysis, Fisher's exact test and ROC curve were used for data analysis - to assess the predictive ability of the studied methods with a guarantee probability of P = 95% and a significance level of α = 0.05.

Results:

The global absolute cardiovascular risk SCORE was calculated according to the three methodologies, resulting in values of 2.312, 2.740, and 4.304, respectively. The PPV of SCORE (2016) and NFA is the same at 63.15, but the PPV of SCORE (2019) is significantly higher at 79.41. Also, yhe study found that the modified additional risk factor is excessive fasting glucose (> 6.1 mmol/l).

Conclusions:

The measured values of PPV and sensitivity of SCORE (NFA) and SCORE ESC (2016) were unsatisfactory, whereas those of SCORE ESC (2019) were the highest. Based on the results and the developed model, it was found that patients who have the modifying additional CV risk factor of abnormal fasting blood glucose (above 6.1 mmol/l) have a 72% probability of being recategorized from the "moderate risk" group to the "high risk" group according to SCORE on NFC.