Currently, the management of pain after cardiac surgery is based on the concept of multimodal analgesia: Combined use of non-opioid analgesics associated with morphine intravenous analgesia by a system controlled by the patient (patient-controlled analgesia-PCA).

The combination of paracetamol and morphine PCA is very effective on pain at rest, but is limited on pain mobilization and causes the problem of side effects associated with opioid (overdose, sedation, respiratory depression, gastrointestinal intolerance, urinary retention ...) which are contributing factors to increase the length of stay in Intensive Care Unit, an additional cost of care and an increase postoperative morbidity and mortality.

Methods that have proved their effectiveness on pain and mobilization used in postoperative cardiac surgery are: anti-inflammatory drugs (NSAIDs) and / or loco-regional analgesia techniques. NSAIDs enhance analgesia produced by PCA Morphine and allow a reduction in morphine consumption, improved postoperative pain, decreased sedation and decreased postoperative morbidity and mortality.

Adverse effects of NSAIDs are commensurate with their time and exposure dose. Consequently, NSAIDs, in the absence of against-indications, should always be prescribed and used at the lowest effective dose and for the shortest possible time.

Some studies have suggested that lower doses of NSAIDs didn't appear to affect their effectiveness. At present, the investigators have no studies that address the hypothesis from which minimum dose of ketoprofen analgesic effect is obtained.

The investigators hypothesis is that lower dose ketoprofen may have efficacy on pain in the postoperative mobilization of cardiac surgery. The investigators want to find, in their study, this "optimal" ketoprofen dose which would be the minimum dose for clinical efficacy demonstrated dose.

This optimal dose could reduce the number of adverse effects of NSAIDs, but their study will probably not have enough power to prove it. NSAID use at these low doses, in postoperative cardiac surgery, could be extended to patient populations most at risk or for a duration longer than 48 hours.

To answer this hypothesis the investigators will consider starting a study of four groups of patients where appropriate doses to patient weight gradually increasing ketoprofen will be used in seeking the minimum effective dose.

Four groups will be determined by randomization. In all these groups, analgesia will be supplemented by a systematic standard self-administered treatment with morphine and paracetamol.

• Group 1 : Placebo group (P). 0 mg/kg ketoprofen IV every 6 hours for 48 hours (or 0 mg/kg every 24 hours) for 48 hours.
• Group 2 : "Ketoprofen quarter dose" (K ¼). 0,125 mg/kg ketoprofen IV every 6 hours (0,5 mg/kg every 24 hours) for 48 hours.
• Group 3 : "Ketoprofen half-dose" (K ½). 0,25 mg/kg ketoprofen IV every 6 hours (1 mg/kg every 24 hours) for 48 hours.
• Group 4 : "Ketoprofen full dose" (KPD). 0,5 mg/kg ketoprofen IV every 6 hours (or 2 mg/kg every 24 hours) for 48 hours.

Inclusion Criteria:

• Patients scheduled for a cardiac surgery (Coronary artery bypass graft, valve replacement)
• Age over 18 years
• Weight between 60 and 100 kg
• Absence of criteria for non-inclusion

Exclusion Criteria:

• Age over 75 years
• Renal insufficiency (MDRD <60 ml / min)
• Hepatic Insufficiency
• Congestive heart failure (EF <40%)
• Insulin-requiring diabetes
• Preoperative coagulation trouble
• History of peptic ulcer or gastrointestinal bleeding
• Allergy to NSAIDs
• Surgery in emergency, aorta surgery, heart transplantation
• Peptic ulcer scalable, history of peptic ulcer or recurrent bleeding (2 or more distinct episodes of bleeding or ulceration objectified)
• Pregnant or lactating women
• Major protected