Validation of an imaging based cardiovascular risk score in a Scottish population. Issue 98 (January 2018)

• Record Type: Journal Article
• Authors: Kockelkoren, Remko; Jairam, Pushpa M; Murchison, John T.; Debray, Thomas P.A.; Mirsadraee, Saeed; van der Graaf, Yolanda; Jong, Pim A.de; van Beek, Edwin J.R.
• Is Part Of: European journal of radiology. Issue 98(2018)
• Page Start: 143
• Page End: 149
• Publication Date: 2018-01
• Subjects: DSC descending aorta -- ICD international classification of disease -- LAD left anterior descending artery -- MV mitral valve -- PROVIDI PROgnostic Value of unrequested Information in Diagnostic Imaging; Cardiovascular disease -- Vascular calcification -- Epidemiology -- Risk prediction -- Multidetector computed tomography; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2017.11.016
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: A model was previously developed determining CVD risk using routine-care CT data. External validation was performed in a Scottish population to determine applicability. The radiological risk score showed adequate performance. It offers a novel strategy for identifying patients with high CVD risk. Abstract: Objectives: A radiological risk score that determines 5-year cardiovascular disease (CVD) risk using routine care CT and patient information readily available to radiologists was previously developed. External validation in a Scottish population was performed to assess the applicability and validity of the risk score in other populations. Methods: 2915 subjects aged ≥40 years who underwent routine clinical chest CT scanning for non-cardiovascular diagnostic indications were followed up until first diagnosis of, or death from, CVD. Using a case-cohort approach, all cases and a random sample of 20% of the participant's CT examinations were visually graded for cardiovascular calcifications and cardiac diameter was measured. The radiological risk score was determined using imaging findings, age, gender, and CT indication. Results: Performance on 5-year CVD risk prediction was assessed. 384 events occurred in 2124 subjects during a mean follow-up of 4.25 years (0–6.4 years). The risk score demonstrated reasonable performance in the studied population. Calibration showed good agreement between actual and 5-year predicted risk of CVD. The c-statistic was 0.71 (95%CI:0.67-0.75). Conclusions: The radiological CVD risk score performed adequately in the Scottish population offering a potential novel strategy for identifying patients at high risk for developing cardiovascular disease using routine care CT data.