Radiometric detection of bacteremia: requirement for terminal subcultures

A radiometric procedure for rapid detection of bacteria in clinical blood specimens was utilized over a period of 1 year in this laboratory. Although in initial studies it was felt that all positive bacteremias would be detected by radiometric examination of cultures for 14CO2 evolution over a 7-day period, we found in the present study that a significant number of bacteria were not detected, except by blind subculturing on day 7 before discarding the culture sample. Microorganisms were detected in 490 individual specimens from 348 patients after examination of 6,200 individual blood specimens, both anaerobically and aerobically. All but 30 of the positive specimens were detected by the radiometric procedure, with an average detection time of 30.5 h. Thirty organisms, representing 6% of the total organisms isolated, were not detected by the BACTEC apparatus. The predominating organisms missed by the radiometric method were group D streptococci, both enterococcal and non-enterococcal species. These isolates represented two-thirds of the total number of organisms not detected by the radiometric procedure. A majority of bacteria detected only upon terminal subculture were isolates from a very small number of patients, suggesting that these organisms might have specific properties that preclude their detection by radiometric assay. Nevertheless, our study results indicate that it is essential that all radiometrically examined blood culture specimens be subcultured prior to discard in order to lessen the likelihood of missing a microbial pathogen.