Abstract 16946: Accuracy of Cardiovascular (cv) Risk Stratification by Canadian Primary Care Physicians: Preliminary Results From the Primary Care Audit of Global Risk Management (PARADIGM) Study
Background: Optimal clinical management requires proper CV risk stratification. Guidelines recommend the use of validated tools (eg. Framingham Risk Score (FRS)) but such tools are not widely adopted by primary care physicians (PCP). The PARADIGM study prospectively evaluated the methods and accuracy of CV risk stratification used by PCP in Canada.
Results: Between March 2009 and January 2010, 105 PCP enrolled 3015 subjects into the PARADIGM registry. Inclusion criteria were men ≥40y or women ≥50y undergoing CV risk assessment, absence of vascular disease, diabetes or known high FRS, and absence of lipid-lowering drugs. PCP estimated CV risk using their method of choice. Risk was determined centrally through FRS assessment of all subjects. Patient details were collected at baseline, with a follow-up visit at 2 months to determine CV risk. Mean age was 56.3y and 41.2% were female. Mean BP was 127/79 mmHg, BMI 27.8kg/m2, and waist circumference 94.8 cm. Hypertension was present in 30.4%, past/current smoking in 34.7%, and family history of premature CVD in 24.3%. Lipid values (mmol/L) were: total cholesterol 5.6, HDL 1.4, TG 1.6, and LDL 3.6. Fasting glucose was 5.4 mmol/L. FRS was routinely used for risk stratification by 34% of PCP and 28.5% used clinical judgement. Patient sample was estimated to be in the following risk categories through PCP and central risk assessment respectively: low in 1531(50.8%) and 1859(61.6%), intermediate in 1093(36.3%) and 787(26.1%), and high in 390(12.9%) and 368(12.2%). There was only fair agreement (57.9%) between PCP and central risk assessment (kappa 0.27). For PCP routinely using FRS, agreement was moderate (kappa 0.43), and only slight for other methods of risk stratification (kappa 0.20). Concordance between PCP and central risk was 65.7%, 50.4%, and 34.2% for low, intermediate, and high risk subjects respectively.
Conclusion: Agreement between physician and central risk assessment is only fair, improves only marginally in PCP who routinely use the FRS, and is poor for all other methods of risk assessment. Physicians correctly identified one third of high risk subjects, and half of intermediate risk subjects. Physician education regarding the proper application of the FRS, and better risk assessment tools are warranted.
- © 2010 by American Heart Association, Inc.