Abstract 15850: Twenty Years Single Centre Experience with Mechanical Heart Valves: A critical review of the Anticoagulation Policy.
Objective: Since the start of the department of cardiac surgery three different mechanical valves (SJM Med, Carbomedics, ATS Med) have been routinely implanted at our institution. The objective of this study was to retrospectively analyze the twenty-year clinical results of all mechanical valves and to challenge our anticoagulation policy.
Methods: From January 1990 to December 2008, 2097 mechanical valves were inserted into 1887 consecutive patients (1344 aortic, 717 mitral, 27 tricuspid and 10 pulmonary). Mean age was 61.2(±12.9) years with a majority of male patients (64,3%). Seventy-two percent of the patients were in NYHA functional class III or higher. Most frequent co-morbidities included atrial fibrillation (594) and coronary disease (527). Follow-up (99.7% completeness) yielded 13328 pt/years and ranged from 12 to 241 months (average=91.2 months). During follow-up 113957 INR values were gathered.
Results: Early 90-day mortality was 5.2% (98 patients, 14 valve-related). The majority of the 723 late deaths were cardiac (324). Survival (Kaplan-Meier) was significantly better for aortic valve patients compared to mitral (Mantel-Cox, p<0.0001). Multivariate analysis (Cox regression) selected age above 70 (p<0.0001), NYHA class ≥III (p<0.0001), non-sinus rhythm (p<0.0001), COPD (p=0.0001), CRF (p=0.001), long-acting coumadins (p=0.001), concomitant CABG (p=0.008), mitral (p=0.004) and tricuspid valve replacement (p=0.023) as significant risk factors for death. Erratic INR (p=0.001), non-sinus rhythm (p=0.04) and NYHA ≥III (p=0.025) were considered independent risk factors for thromboembolism. Long-acting coumadins (p<0.0001), age above 70 (p=0.024) and erratic INR (p=0.012) were risk factors for bleeding. Overall linearized incidences (100/pt/yrs) were: valve thrombosis 0.31, thromboembolism 0.98, bleeding 0.87, endocarditis 0.15, and paravalvular leakage 0.25.
Conclusions: Our twenty-year experience showed excellent clinical outcomes with no structural failure and an acceptable incidence of adverse events. Ideal INR ranges were defined at 2–2.5±0.4 for aortic and 3–3.5±0.5 for mitral valve replacement.
- © 2010 by American Heart Association, Inc.