Clinical Features and Survival Analysis of Patients after Mechanical Heart Valve Replacement, with an Emphasis on Prosthetic Valve Thrombosis

Abstract

Background: Valvular heart diseases are highly prevalent in the world, and surgical valve replacement has improved patients’ survival. Objectives: To describe clinical and laboratory data of patients undergoing mechanical valve replacement, and to determine the incidence of prosthetic valve thrombosis (PVT). Methods: Retrospective cohort study with a follow-up of up to nine years. The study variables were collected from conventional and electronic medical charts. Statistical calculations were performed using the Jamovi software version 1.2.2.; a p<0.05 was considered statistically significant. Kaplan Meier curves were constructed, and Cox regression analysis was performed for analysis of factors related to mortality. Results: A total of 473 patients were included, mean age of 46.9 ±11.3 years. Rheumatic disease was the most common etiology. In a mean follow-up period of 4.43 years, mortality rate was 16.1%. Patients with aortic prosthesis showed higher survival than patients with double implant (mitral and aortic) (p=0.026). Of the factors adjusted for mortality, only functional class and chronic renal failure showed statistically significant association. The incidence of PVT was 0.24/100 patients/year, and the first event occurred more than 1000 days after the implant. Smoking and pannus formation were significantly associated with PVT. No differences were found in INR variability between patients with and without thrombosis by prosthetic position, but significant differences were found in INR before thrombosis as compared with patients without thrombosis (INR= 2.20 [1.80-2.20] vs. 2.80 [2.20-3.40]; p= 0.040). The incidence of stroke and bleeding was 4.4% and 5.2% respectively. Conclusions: The study population was young, and rheumatic valve disease was common in this group. The prevalence of PVT was similar to that described in the literature, despite the low income and low educational level of our sample.