We studied the evolution of resistance to quinolones in Escherichia coli from 1992 to 1997 in Barcelona, Spain. An increasing proportion of quinolone-resistant E. coli (QREC) infections was observed. QREC strains were more common in patients with nosocomial infections but also increased in patients with community-acquired infections (9% in 1992 to 17% in 1996). Seventy (12%) of 572 episodes of E. coli bacteremia were due to QREC. Factors significantly associated with QREC bacteremia were the presence of underlying disease, recent exposure to antibiotics, and bacteremia of unknown origin. In the multivariate analysis, only prior exposure to antimicrobial agents ( P < 0.001; odds ratio [OR] = 2), specifically, to quinolones ( P < 0.001; OR = 14), and the presence of a urinary catheter ( P < 0.001; OR = 2) were significantly associated with QREC bacteremia. Among 16 QREC isolates from cultures of blood of community origin selected at random, 13 different pulsed-field gel electrophoresis patterns were recognized, showing the genetic diversity of these isolates and in turn indicating the independent emergence of QREC in the community. The prevalence of QREC in the feces of healthy people was unexpectedly high (24% in adults and 26% in children). A survey of the prevalence of QREC of avian and porcine origin revealed a very high proportion of QREC in animal feces (up to 90% of chickens harbored QREC). The high prevalence of QREC in the stools of healthy humans in our area could be linked to the high prevalence of resistant isolates in poultry and pork.

The spacer oligonucleotide typing (spoligotyping) method was evaluated for its ability to differentiate Mycobacterium bovis strains. This method detects the presence or absence of spacers of the direct repeat locus of the M. bovis genome. The spacers in the direct repeat locus are amplified by PCR and are detected by hybridization of the biotin-labelled PCR product with a membrane containing oligonucleotides derived from spacer sequences that have previously been bound to a membrane. One hundred eighty-two M. bovis isolates from domestic animals (cattle, goat, sheep, and cats) and wild animals (deer and wild boar) were spoligotyped, and the results were compared with those obtained by IS6110 restriction fragment length polymorphism analysis. Two rather homogeneous clusters of isolates containing 20 and 4 types, respectively, were identified by spoligotyping. The first cluster included isolates from cattle, cats, and feral animals. By spoligotyping, isolates from the Spanish wild boar and deer had the same pattern as some bovine isolates, suggesting transmission between these animals and cattle and highlighting the importance of the study of these reservoirs. The second cluster included all the caprine and ovine isolates. Within each cluster, the patterns of the different strains differed only slightly, suggesting that the spoligotypes may be characteristic of strains from particular animal species. Spoligotyping proved to be useful for studying the epidemiology of bovine M. bovis isolates, especially of those isolates containing only a single copy of IS6110. In view of our results, we suggest fingerprinting all M. bovis strains by the spoligotyping method initially and then by IS6110 restriction fragment length polymorphism typing of the strains belonging to the most common spoligotypes.