Comparison And Application Of CVD Risk Prediction Models Among Patients With Hypertension

Part Ⅰ Comparison of the results of CVD risk prediction models among patients with htpertension in Jiangsu ProvinceObjective: To explore the difference between China-PAR and Framingham risk score in the risk assessment of 10-year cardiovascular diseases among patients with hypertension in Jiangsu Province.Methods: A total of 10498 hypertensive patients aged 30 and above were selected from the Standardized Management Areas for Hypertension in Jiangsu Province which was established in 2010.Baseline information was collected by questionnaire survey,physical examination and laboratory test,China-PAR and Framingham risk score were used to evaluate the 10-year CVD risk,and the prediction results of the two models were calculated and compared.Results: The 10-year risk of CVD predicted by China-PAR and Framingham risk score in the total population was 12.22% and 17.74%,respectively.The predicted risk of Framingham risk score was higher than China-PAR,and the difference was statistically significant(P<0.05).The average risk calculated by China-PAR and Framingham risk score was 14.66% and 25.35% for men,10.45% and12.22% for women.The gender difference was statistically significant(P<0.05).The predicted risks of both models were higher in men than in women.After stratified analysis by age,the prediction results in the same model were similar to the total population,indicating that men were at higher risk than women.Regardless of gender,the risk of 60 years old and above is higher than that of 30-44 years old and 45-59 years old.For the same gender,in different age groups,the predicted risk of Framingham was higher than that of China-PAR.The difference was statistically significant(P<0.05).China-PAR and Framingham risk score predicted that 5.3% and 6.3% of the patients with essential hypertension in Jiangsu Province were in the low risk group,of which 2.2% and 2.1% for men,7.6%and 9.3% for women.The middle risk group accounted for 29.2% and 24.6% respectively,of which14.4% and 8.3% for men,40.0% and 36.4% for women.High risk group accounted for 65.5% and69.1% respectively,of which 83.4% and 89.6% for men and 52.4% and 54.3% for women.By analyzing the distribution characteristics of CVD risk factors in different risk groups of the two models,we found that family history of CVD was statistical significance among different risk groups in China-PAR model,while there was no statistical difference in Framingham risk score.The results of cross analysis of the 10-year risk of CVD predicted by China-PAR and Framingham risk score showed that 60.7% of the participants were divided into the same risk score group,and the Kappa value of the consistency test was 0.427,indicating the general consistency.According to gender analysis,48.1% of men and 69.9% of women were divided into the same risk group respectively by the two models.Compared with the China-PAR,50% of men were divided into higher risk score group by Framingham risk score.Among them,42.1% menwere predicted belong to 10%-19.9%the risk of CVD by China-PAR,but the risk predicted by Framingham risk score was ≥20%.The Kappa value of the consistency test of the two prediction methods in male participants was 0.240,indicating poor consistency.More female participants were divided into the same risk score group by two prediction models than male subjects.The Kappa value of the two prediction models in the female participants was 0.531,indicating moderate consistency.Conclusion: The risks of 10-year CVD calculated by China-PAR and Framingham risk score were different.The risk predicted by Framingham risk score is higher than that predicted by China-PAR,and the consistency of predicted rusults between the two models was moderate.PartⅡ The application of CVD risk prediction models among patients with hypertension in Jiangsu ProvinceObjective: To compare the predicted risk calculated by China-PAR and Framingham risk score and actual risk of cardiovascular diseases in hypertensive patients in Jiangsu Province,and to evaluate the effect of standardized management for hypertension in community.Methods: Five provincial standardized management areas of hypertension were established in Jiangsu Province in January 2010.From December 2010 to December 2018,a comprehensive followup was conducted and the accidents,such as CVD,cancer,death,and so on,were collected.The hypertension patients were standardized managed by community doctors according to the “Basic edition of Chinese hypertension prevention and control guidelines”.With the lowest risk group as a reference,the HR and its’ 95% CI of CVD incidence in different risk groups in China-PAR and Framingham risk score were calculated by using multivariate Cox proportional hazard model.The CVD risk predicted by the two models and the actual observed CVD incidence was compared,which used to evaluate the effect of standardized management among patients with hypertension in community.Results: 10498 participants were included in this study with a total of 86917 person-years and an average period of 8.3 years of follow-up.During the follow-up period,totally 693 CVD incidences were identified,with an incidence rate of 6.6%.After adjusting for age,body mass index(BMI)and alcohol consumption,Cox proportional hazards regression analysis results showed that the risks of CVD in China-PAR predicted model in 5%-9.9%,10%-19.9% and ≥20% groups were 2.38 times(HR=2.38,95% CI: 1.10-5.18),3.85 times(HR=3.85,95% CI: 1.75-8.46),and 5.30 times(HR=5.30,95% CI: 2.35-11.96)compared with the group of predicted risk <5%,respectively.The results of gender stratification analysis are similar to the total population.After adjusting for age,waist circumference(WC),BMI and alcohol consumption,the results of Cox proportional hazards regression analysis showed that the risks of CVD in Framingham predicted score in 5%-9.9%,10%-19.9%,and ≥ 20% were 1.26 times(HR=1.26,95% CI: 0.72-2.20),1.80 times(HR=1.80,95% CI: 1.03-3.15),2.06 times(HR=2.06,95% CI: 1.15-3.68)compared with group of predicted risk <5%,respectively.After gender stratification analysis,there was no statistically significant risk association in the male population,while the results in the female population were similar to the total population.According to the China-PAR,the actual incidence rate and the predicted incidence rate in risk group <5% was 1.26% vs 3.28%,5%-9.9% group was 4.01% vs 7.31%,10%-19.9% group was 7.72% vs 13.54%,and ≥20% group was 10.86% vs 24.27%,the predicted incidence rate of each group was higher than the actual incidence rate,which was 2.60 times,1.82 times,1.75 times and 2.23 times,respectively.The results of gender stratification analysis were similar to the total population.After the actual incidence rates were adjusted by Kaplan-Meier,the overestimation of the results in ChinaPAR model still existed;moreover,there was an increasing trend.According to the Framingham risk score,the actual incidence rate and the predicted incidence rate in risk group <5% was 2.28% vs 3.79%,5%-9.9% group was 4.11% vs 7.69%,10%-19.9% group was 7.03% vs 14.32%,and ≥20% group was 8.89% vs 32.38%,the predicted incidence rate of each group was higher than the actual incidence rate,which was 1.66 times,1.87 times,2.04 times and 3.64 times,respectively.The results of gender stratification analysis were similar to the total population.The actual incidence rates were adjusted by Kaplan-Meier,the overestimation of the CVD risk in Framingham risk score was existed.Conclusion: Compared with the actual incidence rate,the predicted risk calculated by China PAR and Framingham risk score was overestimated,suggesting that the effect of standardized management for hypertension in community in Jiangsu Province was better.