Methods: Patients were retrospectively identified from 2007-2015 using appropriate IE-related ICD-9 codes. Cases that met definite Modified Duke Criteria for IE were further classified as either PVE or NVE, and were reviewed for epidemiology, causative organism(s), affected valves and associations, risk factors, dental procedures in the past 6 months, and 30-day mortality.
Results: 363 admissions met criteria for definite endocarditis, with 261 NVE cases and 59 PVE cases. Forty-three cases that were either associated with an infection involving both native and prosthetic valves or intracardiac devices were omitted from this study. Most risk factors, such as hemodialysis and intravenous drug use did not show any significant difference amongst the two groups. IE involving the aortic valve as well as a previous history of IE were more likely to be seen in PVE (both p<0.0001). Dental procedures done in the preceding 6 months before IE admission were more likely to be associated with PVE than NVE (p=0.0043). PVE showed a higher likelihood of 30 day mortality compared to NVE (p=0.067). The causative organisms of PVE were more likely to be caused by common gut pathogens such as Klebsiella and Enterobacter species.
Conclusion: PVE cases had a significantly higher chance of involving the aortic valve as well as having a history of IE. PVE cases were also significantly more likely to be associated with a dental procedure done in the preceding 6 months than with the NVE cases. This implies that patients with prosthetic valves, who are currently covered under the 2007 AHA guidelines to receive prophylaxis prior to dental procedures, are still at a high risk of developing PVE. It may be prudent to reconsider adding a post-procedure dose of antibiotics, instead of a single pre-procedure dose, to extend the protection of this high risk population with prosthetic valves. Furthermore, PVE cases showed higher rates of 30 day mortality compared to NVE with near significance, which is likely multifactorial.
H. Y. A. Lo,
N. Khardori, None