Improved 5-year prediction of all-cause mortality by coronary CT angiography applying the CONFIRM score.

Document Type

Article

Publication Date

3-1-2017

Publication Title

Eur Heart J Cardiovasc Imaging

Abstract

Aims: To investigate the long-term performance of the CONFIRM score for prediction of all-cause mortality in a large patient cohort undergoing coronary computed tomography angiography (CCTA).

Methods and results: Patients with a 5-year follow-up from the international multicentre CONFIRM registry were included. The primary endpoint was all-cause mortality. The predictive value of the CONFIRM score over clinical risk scores (Morise, Framingham, and NCEP ATP III score) was studied in the entire patient population as well as in subgroups. Improvement in risk prediction and patient reclassification were assessed using categorical net reclassification index (NRI) and integrated discrimination improvement (IDI). During a median follow-up period of 5.3 years, 982 (6.5%) of 15 219 patients died. The CONFIRM score outperformed the prognostic value of the studied three clinical risk scores (c-indices: CONFIRM score 0.696, NCEP ATP III score 0.675, Framingham score 0.610, Morise score 0.606; c-index for improvement CONFIRM score vs. NCEP ATP III score 0.650, P < 0.0001). Application of the CONFIRM score allowed reclassification of 34% of patients when compared with the NCEP ATP III score, which was the best clinical risk score. Reclassification was significant as revealed by categorical NRI (0.06 with 95% CI 0.02 and 0.10, P = 0.005) and IDI (0.013 with 95% CI 0.01 and 0.015, P < 0.001). Subgroup analysis revealed a comparable performance in a variety of patient subgroups.

Conclusions: The CONFIRM score permits a significantly improved prediction of mortality over clinical risk scores for >5 years after CCTA. These findings are consistent in a large variety of patient subgroups.

Volume

18

Issue

3

First Page

286

Last Page

293

DOI

10.1093/ehjci/jew195

ISSN

2047-2412

PubMed ID

28363203

Share

COinS