In cardiac surgery, the assessment of operative risk and quality of care is a major challenge for both patient, and surgical team. It is also important for health care decisionmakers to have predictive tools to compare alternative technics such as conventional cardiac surgery and interventional cardiology. Since 1998, the European System for Cardiac Operative Risk Evaluation (EuroSCORE), updated in 2012 (EuroSCORE II) is the most universally used system in this purpose. Its success is the result of a good balance between predictive capability and simplicity. It consists almost exclusively of clinical variables.

However, the objectivity and the predictive ability of some of those clinical items remain controversial, particularly those addressing severity of illness for high-risk patients. For instance, the degree of priority is submitted to the subjective assessment by the surgical team at the time of surgery. Objective data describing the severity of patients arriving in the operating room are still missing.

Many biomarkers are relevant in qualifying severity of syndromes: shock (PH, lactates, LDH), heart disease (N-terminal pro b-type natriuretic peptide (NTproBNP), troponin T and I), respiratory disorder (blood gaz analysis), liver insufficiency (TP, factor V), renal impairment (serum creatinine, creatinine clearance), inflammatory condition (fibrinogen, CRP), or the underlying medical condition such as diabetes (HbA1c, microalbuminuria) and nutritional status (albumin).

In ICU, many scores use biological data to measure, on daily basis, the severity of the patient status. Their routine use is simplified by applications available on smartphones. They are drawn into hospital information systems.

In cardiac surgery, some studies seem to demonstrate the measurement of some preoperative biological variables (eg NT Pro-BNP ...) in risk prediction. In terms of diabetes, HbA1c is of particular interest because it detects underlying diabetes if unknown (emergent situation) or reflects its poor control before surgery. This criterion could be more reliable than the simple information of patient treated by insulin (EuroSCORE II criterion).

Finally, the reinforcement of existing scores with biological variables is recommended by the group of recommendations in prognostic research strategy (PROGRESS 2014), rather than creating new scores ex nihilo.

The hypothesis is that adding biological data collected at the time of arrival of the patient in the operating room would better qualify the patients' severity condition and therefore increase the risk prediction of early mortality and severe morbidity after cardiac surgery. The purpose of this study is to test this hypothesis and especially test whether the biological data would increase the EuroSCORE II performance, by improving the prediction for high-risk patients.

Main objective: To improve the EuroSCORE II predictive power on the 90-day mortality and/or severe hospital morbidity, incorporating immediate preoperative biological data.

Secondary objectives:

1. Analyze EuroSCORE II calibration by risk group (external validation).
2. Studying the improvement of the EuroSCORE II-bio predictive power by risk class, especially high risk.
3. Refine selection and definition of discriminating criteria for diabetics for predicting surgical risk in cardiac surgery (the HbA1c value, the existence of microalbuminuria, the nature of treatment etc.).
4. Compare the risk profiles of operated patients in the French Overseas Departments to those from different sites of Western Europe.

Inclusion Criteria:

• Patient > 18 years old at the time of inclusion
• All patients operated on extracorporeal circulation or beating heart
• Patient affiliated to a social security scheme
• Patient having been informed of the research objectives
• Patient not objecting to participate in research

Exclusion Criteria:

• Patient receiving heart transplant
• Patient < 18 years old
• Patient refusing to participate in the study
• Pregnant women
• Protected adult (person under guardianship and trusteeship) or deprived of liberty

All patients needed cardiac surgery during 1 year at the different sites (in France (8 sites), and in two French Overseas Departments (2 sites))