Aims: To validate and assess performance of the current ACC/AHA very high risk criteria in patients with clinically manifest arterial disease.
Methods and results: Data were used from the SMART study (n = 7216) and REACH Registry (n = 48 322), two prospective cohorts of patients with manifest atherosclerotic arterial disease. Prevalence and incidence rates of recurrent major adverse cardiovascular events (MACE) were calculated, according to the ACC/AHA VHR criteria (cardiovascular disease combined with diabetes, smoking, dyslipidaemia, and/or recent recurrent coronary events). Performance of the ACC/AHA criteria was compared with single very high risk factors in terms of C-statistics and Net Reclassification Index. All patients were at VHR according to the ESC guidelines (incidence of recurrent MACE in SMART was 2.4/100PY, with 95% CI 2.3-2.5/100PY and in REACH 5.1/100PY with 95% CI 5.0-5.3/100PY). In SMART 57% of the patients were at VHR according to the ACC/AHA criteria (incidence of recurrent MACE 2.7/100PY, 95% CI 2.5-2.9/100PY) and in REACH this was 64% (5.9/100PY, 95% CI 5.7-6.1/100PY). The C-statistic for the ACC/AHA VHR criteria was 0.53 in REACH and 0.54 in SMART. Very high risk factors with comparable or slightly better performance were eGFR < 45, polyvascular disease and age >70 years. Around two third of the patients meeting the ACC/AHA VHR criteria had a predicted 10-year risk of recurrent MACE <30%.
Conclusion: The ACC/AHA VHR criteria have limited discriminative power. Identifying patients with clinically manifest arterial disease at VHR for recurrent vascular events using eGFR <45, polyvascular disease, or age >70 years performs as well as the ACC/AHA VHR criteria.
Keywords: Cardiovascular events; Risk prediction; Secondary prevention; Very high risk.
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2017. For permissions, please email: [email protected].