Evaluation of VITAL Automated Blood Culture System

Lee, Myung Hee

Abstract

BACKGROUND: An evaluation was performed to assess the performance of continuously monitoring VITAL automated blood culture system (bio-Merieux, Marcy-l'Etoile, France) and to investigate the value of performing final subcultures at the end of the protocol.
METHODS
A retrospective study was conducted over a period of one year (October 1996 to September 1997) with 7,078 blood culture bottles sent to Microbiology Department of Korea Veterans Hospital. Not only VITAL positive bottles but also all the VITAL negative bottles were observed under a microscope after Gram staining and subcultured aerobically and anaerobically at the end of 5-day protocol. All isolates were identified by the conventional method and ATB system.
RESULTS
Among total 7,078 bottles, 688 bottles (9.72%) were declared positive by the system, of which 68 (0.96%) proved to be false positive. The final blind subculture permitted the detection of 96 falsely negative bottles (1.38%). The average time to detection was 38h 08, and 20% of 444 samples having microorganisms were detected during the first 12 h, 45% during the first 24 h, 63% within 48 h, and 87% within 120 h. 58 samples (13%), which contained 20 cases of Gram positive cocci and 20 cases of yeasts- especially 65% of C. parapsilosis, were declared negative by the system but gave a positive subculture. Among the positive bottles, 86.3% were detected by the slope algorithm, 10.6% by the delta algorithm, and 3.1% by the threshold algorithm.
CONCLUSIONS
I conclude that the VITAL system must be modified to improve the detection of staphylococci and yeasts and a more sensitive computer algorithm may be required so that the terminal subcultures will not be necessary. Each laboratory must decide the value of terminal blind subcultures on the basis of patient population and the microorganisms that are most frequently isolated in their institution.