Dual versus triple antithrombotic therapy in patients undergoing percutaneous coronary intervention-meta-analysis and meta-regression.