Details

Autor(en) / Beteiligte

• Barr, Elizabeth L
• Barzi, Federica
• Rohit, Athira
• Cunningham, Joan
• Tatipata, Shawn
• McDermott, Robyn
• Hoy, Wendy E
• Wang, Zhiqiang
• Bradshaw, Pamela
• Dimer, Lyn
• Thompson, Peter
• Brimblecombe, Julie
• O’Dea, Kerin
• Connors, Christine
• Guthridge, Steven
• Brown, Alex D
• Cass, Alan
• Shaw, Jonathan E
• Maple-Brown, Louise J

Titel

Abstract 14374: Performance of ACC/AHA Cardiovascular Disease Risk Equations in Indigenous Australians With and Without Diabetes: The CRISP Study Collaboration

Ist Teil von

• Circulation (New York, N.Y.), 2018-11, Vol.138 (Suppl_1 Suppl 1), p.A14374-A14374

Ort / Verlag

by the American College of Cardiology Foundation and the American Heart Association, Inc

Erscheinungsjahr

2018

Quelle

Elektronische Zeitschriftenbibliothek - Freely accessible e-journals

Beschreibungen/Notizen

• IntroductionAtherosclerotic cardiovascular disease (ASCVD) remains a major public health issue for many Indigenous populations, including Indigenous Australians because of earlier onset diabetes and longer duration of risk factor exposure at any given age. Using data from the Cardiovascular Risk in IndigenouS People (CRISP) study, we assessed the performance of the ACC/AHA ASCVD risk equations.MethodsWe undertook an individual-participant meta-analysis on 2,066 Indigenous Australians (57% women) from five population-based cohorts aged 40-79 years without an ASCVD history. Baseline data were harmonised and 312 prospective ASCVD outcomes were collected using data-linkage to hospital databases and the National Death Index. Predicted 10-year ASCVD risks (95%CI) were determined using the ACC/AHA equations for non-Hispanic white and African American men and women. χ tests compared predicted and observed risks. Area under the receiver operating characteristic curve (AUC) assessed discrimination. Analyses were stratified by diabetes status (n=768 with diabetes).ResultsThe observed 10-year ASCVD probability for Indigenous Australians was consistently underestimated by both ASCVD risk equations (Table). The equation for non-Hispanic white Americans underestimated ASCVD risks by 15% in men and 48% in women with diabetes, and by 38% in men and 50% in women without diabetes. While the degree of underestimation was similar when applying the non-Hispanic white or African American equation for men, the gaps between observed and predicted risks were greater for women when using the equation for non-Hispanic white Americans. Discrimination was fair (AUC=0.64-0.65 for men; 0.71-0.72 for women).ConclusionsThese findings highlight that ASCVD risk prediction needs to be improved to better detect Indigenous Australians at risk of ASCVD, and may have implications for other Indigenous populations who have similar risk factor profiles.