Ja. Kellogg et al., CLINICAL COMPARISON OF ISOLATOR AND THIOL BROTH WITH ESP AEROBIC AND ANAEROBIC BOTTLES FOR RECOVERY OF PATHOGENS FROM BLOOD, Journal of clinical microbiology, 32(9), 1994, pp. 2050-2055
The recovery of pathogens and the speed of their detection were determ
ined for our conventional blood culture system (an Isolator [Wampole]
and a 100-ml Thiol bottle [Difco]) compared with automated ESP aerobic
and anaerobic bottles (80 ml each; Difco). Each of the four culture d
evices was inoculated with approximately 10 ml of blood from symptomat
ic patients weighing more than 80 lb (ca. 36 kg). From 7,070 sets of c
ultures for 2,841 patients, 607 clinically significant isolates were r
ecovered: 456 (75.1%) from the Isolator, 353 (58.2%) from Thiol, 377 (
62.1%) from ESP aerobic bottles, and 346 (57.0%) from ESP anaerobic bo
ttles. Of the 607 isolates, 149 (24.5%) were detected only with the co
nventional system (Isolator and/or Thiol), and 65 (10.7%) were detecte
d only with the ESP two-bottle system (P < 0.001). Our conventional sy
stem allowed for detection of significantly more isolates of members o
f the family Enterobacteriaceae (P < 0.001), Staphylococcus aureus (P
< 0.01), Staphylococcus spp. (coagulase-negative) (P < 0.01), and Ente
rococcus spp. (P < 0.05), and ESP facilitated detection of significant
ly more isolates of S. pneumoniae (P < 0.01). When all four devices in
a culture set were positive for the same isolate, no microbial specie
s or group was detected significantly earlier (greater than or equal t
o 24 h) by either blood culture system. The Isolator contamination rat
e (4.8%) was greater than or equal to 6 times the rate for any of the
bottles. Of pathogens detected by the Isolator, 50% were recovered in
counts of less than or equal to 1.0 CFU/ml and 18% were recovered only
as a single colony. The ESP system offered an automated, less labor-i
ntensive blood culture system for which routine subcultures were not r
equired, but the important considerations of culturing large volumes o
f blood and of obtaining at least two sets from each patient in our po
pulation were reemphasized.