Sodium-glucose co-transporter inhibitors have demonstrated reduced incidence of cardiovascular (CV) events, which cannot be fully explained by effects on HbA1c, body weight or blood pressure. Type 2 diabetes mellitus (T2DM) economic models link treatment-related changes in risk factor levels to long-term incidence of CV events via risk equations (RE). This study assessed the performance of commonly used CV RE in T2DM to the CV outcomes trial DECLARE-TIMI 58. Ten sets of RE were coded in MS-Excel (ADVANCE, ARIC, BRAVO, Framingham, Fremantle, New Zealand CVD, QRisk3, Swedish NDR, UKPDS 68 and 82) to predict annual probabilities of CV events and applied to baseline characteristics and 1-year treatment effects (HbA1c, weight and blood pressure) for dapagliflozin (DAPA) versus placebo (PLA) from DECLARE-TIMI 58. Predicted event rates for myocardial infarction (MI), heart failure (HF) and stroke and the relative risks (RR) associated with DAPA versus PLA were estimated and compared to the closest reported trial outcome. Of the ten sets of RE, seven predicted composite CV outcomes, demanding assumptions to predict risk of individual events; only UKPDS and BRAVO predicted all three events. When comparing trial observations against predictions averaged across RE, correlation was greater for absolute event rates (R2: 0.92) than RRs (R2: 0.42). ARIC, Fremantle and New Zealand RE considerably overpredicted risk of MI and stroke compared to the trial. RE typically underpredicted the effect of DAPA on HF (average predicted RR: 0.89 versus reported hazard ratio 0.73, CI: 0.61 to 0.88) and MI (average predicted RR: 0.93 versus reported hazard ratio 0.89, CI: 0.77 to 1.01). Existing methods of CV risk prediction evaluated in this study did not accurately predict the treatment benefits observed in DECLARE-TIMI 58. The development of DECLARE-TIMI 58-specific risk equations may be required to robustly evaluate the long-term clinical and economic impact of DAPA in T2DM.


P. McEwan: Consultant; Self; AstraZeneca. Employee; Self; Health Economics and Outcomes Research Ltd. V. Foos: Employee; Self; Health Economics and Outcomes Research Ltd., IQVIA. H. Bennett: Consultant; Self; AstraZeneca. Employee; Self; Health Economics and Outcomes Research Ltd. B. Kartman: Employee; Self; AstraZeneca. Stock/Shareholder; Self; AstraZeneca. C. Edmonds: Employee; Self; AstraZeneca. I.A. Gause-Nilsson: Employee; Self; AstraZeneca.



Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at