Comparison of methods to identify individuals at increased risk of cardiovascular disease in Italian cohorts Davide Giavarina a, *, Elena Barzon c , Massimo Cigolini b , Gabriella Mezzena a , Giuliano Soffiati a a Department of Clinical Pathology, ‘‘San Bortolo’’ Hospital, Via Rodolfi 37, I-36100 Vicenza, Italy b Observatory of Clinical Epidemiology ‘‘Sen. Delio Giacometti’’, Arzignano (VI), Italy c Department of Clinical Pathology, ‘‘S.Maria degli Angeli’’ Hospital, Pordenone (PN), Italy Received 8 June 2005; received in revised form 2 November 2005; accepted 23 December 2005 KEYWORDS Risk assessment; Coronary heart disease; Primary prevention; Guidelines; Cost-effectiveness Abstract Background and aims: Guidelines for the primary prevention of cardio- vascular disease recommend the use of risk-assessment methods to identify high risk patients who can benefit from lifestyle changes and/or drug treatment. Although all these risk-prediction methods are based on the same principle, they produce different risk estimates. The aim of this study was to compare the most recent and widely used cardiovascular risk-prediction methods and the respective guidelines when applied to Italian cohorts. Methods and results: Seven different risk-assessment methods were applied to two groups of subjects, 536 healthy individuals and 426 diabetic patients. Sensitivity and specificity of Framingham-based risk-assessment methods were calculated using the Framingham full equation as the reference standard. The extent of con- cordance among the different risk-assessment methods was determined by kappa test. By using NCEP-ATPIII risk calculator, modified Sheffield tables, Joint European Societies charts, Joint British Societies charts, Italian CUORE Project charts, Euro- pean SCORE charts and New Zealand National Heart Foundation charts in the group of 536 healthy subjects, lipid-lowering treatment would be recommended in 17.5%, 12.7%, 12.1%, 8.6%, 5.0%, 4.7%, and 1.1% subjects, respectively. By using the same risk-assessment methods in the group of 426 diabetic patients, treatment would be recommended for 100%, 82.9%, 66.9%, 77.7%, 43.0%, 74.9%, and 47.4% patients, respectively. The Joint British charts and the modified Sheffield tables showed the closest agreement with the reference standard. * Corresponding author. Tel.: þ39 0444 753372; fax: þ39 0444 752501. E-mail address: davide.giavarina@ulssvicenza.it (D. Giavarina). 0939-4753/$ - see front matter ª 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.numecd.2005.12.008 Nutrition, Metabolism & Cardiovascular Diseases (2007) 17, 311e318 www.elsevier.com/locate/nmcd