A 4-month trial involving 2,563 routine clinical specimens was conducted to compare the improved BACTEC TB system (12B medium) with the conventional Lowenstein-Jensen (LJ) media for the isolation, identification, and susceptibility testing of mycobacteria. One hundred sixty-two mycobacterial isolates were recovered, 147 (91%) with BACTEC and 118 (73%) with LJ media. Of these, 62 were Mycobacterium tuberculosis complex strains, 59 (95%) of which were isolated with BACTEC and 54 (87%) of which were isolated with LJ media. Of the remaining 100 isolates, which were mycobacteria other than tuberculosis (MOTT), BACTEC and LJ media detected 88 and 64%, respectively. The contamination rate was significantly higher in BACTEC (5%) than in LJ media (3.3%). The mean isolation time for M. tuberculosis complex with BACTEC was 15.5 days, compared with 25.6 days with LJ. For MOTT, the mean isolation times were 5.8 and 21.4 days, respectively. Identification of 32 M. tuberculosis complex isolates and 38 isolates of MOTT by the BACTEC NAP (p-nitroa-acetylamino- ,-hydroxypropiophenone) inhibition test gave 100% agreement with conventional biochemical identifications. The results of susceptibility testing of 18 M. tuberculosis complex isolates with BACTEC agreed completely with those obtained by the resistance ratio method.

No abstract

DNA hybridization assays for genes encoding the Escherichia coli adhesins K88 and K99 were developed. These assays were used to screen a variety of animal E. coli strains, and the results were compared with results obtained by serological methods. Ail methods compared well for determining the presence of adhesin K99. However, the DNA hybridization detected more K88-positive strains than did either enzyme immunoassay or slide agglutination. Agreement was improved when strains were grown on blood or nutrient agar instead of Minca medium.