Prediction of individual life-years gained without cardiovascular events from lipid, blood pressure, glucose, and aspirin treatment based on data of more than 500 000 patients with Type 2 diabetes mellitus

Literature - Berkelmans GFN, Gudbjörnsdottir S, Visseren FLJ et al. - Eur Heart J 2019; 0, 1–10

Introduction

Patients with type 2 diabetes mellitus (T2DM) are at up to two-fold increased risk for CVD, which results in reduced life expectancy and quality-adjusted life years, compared with individuals without T2DM, regardless of other risk factors [1-3]. Although group level effectiveness of recommended medications for prevention of CVD in T2DM has been proven in high-quality trials, important differences in absolute effectiveness are known to exist between individuals.

This study therefore aimed to develop and externally validate a prediction tool for individualizing CVD prevention with lipid-lowering, antihypertensive, glucose-lowering, and aspirin treatment in patients with T2DM, by predicting treatment effects as reductions in 10-year CVD-risk and lifetime risk, and gain in CVD-free life expectancy (time without recurrent MI or stroke).

Methods and results

Development of the DIAL model

The Diabetes Lifetime-perspective prediction (DIAL) model was developed, using data from patients with T2DM registered in the Swedish National Diabetes Registry (NDR) (n=389.366). Two complementary competing risk adjusted Cox proportional hazard functions were developed, one for the prediction of CVD events using non-vascular mortality as the competing endpoint, and another for prediction of non-vascular mortality using CVD events as the competing endpoint.

External validation was performed using data from the ADVANCE, ACCORD, ASCOT and ALLHAT-LLT trials, the SMART and EPIC-NL cohorts, and the Scottish Care Information (SCI) Diabetes database (total n=197.785). This analysis included patients aged >18 years with a diagnosis of T2DM with or without prevalent CVD. Individuals with a history of cancer and those with stage 4 or 5 CKD were excluded. Eleven CVD risk predictors were selected based on existing diabetes risk scores and availability in routine clinical practice: age, sex, smoking, systolic blood pressure, BMI, HbA1c, eGFR, non- HDL-c, albuminuria, T2DM duration, insulin treatment, and history of CVD. Competing outcomes were CVD events (non-fatal MI, non-fatal stoke, or vascular mortality) and non-vascular mortality (all deaths other than those with an identified CV outcome).

Internal validation of the DIAL model

Predicted 10-year risk for CVD and all-cause mortality showed good agreement with the 10-year observed risk, with c-statistics of 0.83 (95%CI: 0.83-0.84) for 10-year CVD risk, 0.72 (95%CI: 0.72-0.72) for 10 year non-vascular mortality risk and 0.77 (95%CI: 0.76-0.77) for 10-year CVD-free survival.

External validation of the DIAL model

Predicted 5-year risks for CVD and all-cause mortality were in good agreement with the observed 5-year CVD-free survival in Western-Europe, Eastern-Europe, North-America, and Asia and Oceania. The C-statistics of the estimated 5-year CVD risk were between 0.64 and 0.65 in all geographically pooled datasets. For 5-year non-vascular mortality, the c- statistics of the different geographical regions varied between 0.59 and 0.67, and between 0.64 and 0.69 for all-cause mortality.

Individual lifetime estimates and treatment effects for people with T2DM

The DIAL model prediction data have been implemented in the interactive online U-Prevent calculator. Patient characteristics and current medication can be entered in this decision support tool to estimate individual risk and CVD-free survival. Also, the individual effect from medication changes can be modelled in terms of CVD-free life-years gained, absolute risk reduction, and individual number needed to treat.

Conclusion

References

Find this article online at Eur Heart JFind the interactive calculator at U-prevent.com