Comparison of heart valve culture between two Danish endocarditis centres

Abstract

Background: The degree to which the results of valve culture depend on different laboratory procedures as well as other factors is unknown. The aim of this study was to compare the results of heart valve culture at 2 different endocarditis centres in order to clarify this. Methods: The study included 223 patients with definitive endocarditis undergoing heart valve surgery at 2 Danish endocarditis centres (96 at the East centre and 127 at the West centre). The following data related to the samples were registered: transportation, time to inoculation, culture media used and duration of incubation, species distribution, and preoperative duration of appropriate antimicrobial treatment (DAAT). 16S polymerase chain reaction (PCR) of valve tissue was used to estimate the frequency of non-cultivable bacteria. Results: Valve culture was positive in 12.5% of cases at the East centre and 36.2% at the West centre (p < 0.001). Valve cultures were positive in 10%, 45%, and 58% of cases of streptococcal, staphylococcal, and enterococcal infections, respectively. The median DAAT was 3 days for valve culture-positive cases and 11 days for valve culture-negative cases (p < 0.001). By logistic regression analysis, DAAT, species distribution, and laboratory protocol were identified as independent risk determinants for positive valve culture. The adjusted odds ratio for positive culture by the West centre protocol was 3.0 (confidence interval 1.3–6.9; East = 1, p < 0.01). The valve culture contamination rate was 4% at the East centre and 31% at the West centre (p < 0.01). Conclusions: Our study showed that the duration of preoperative antibiotic treatment and species distribution must be taken into account when comparing valve culture results. Differences in laboratory protocol explained the 3-fold higher sensitivity of heart valve culture observed in the West centre compared to the East centre.

Keywords::

• Heart valve culture
• laboratory protocol
• endocarditis

Acknowledgements

We thank the Western Danish Research Forum for Health Sciences for their help with the statistical analyses. This work was supported by the Central Denmark Region Health Science Research Foundation.

Declaration of interest: No conflicts of interest.