Effect of trimethoprim-sulfamethoxazole prophylaxis on antimicrobial resistance of fecal Escherichia coli in HIV-infected patients in Tanzania.

BACKGROUND

Trimethoprim-sulfamethoxazole (SXT) reduces morbidity and mortality among HIV-infected persons in Africa, but its impact on antimicrobial resistance is of concern.

METHODS

HIV-uninfected (group A), HIV-infected but not requiring SXT (group B), and HIV-infected and eligible for SXT (group C) adults were recruited into a prospective observational cohort study in Moshi, Tanzania. Stool was examined for Escherichia coli nonsusceptible to SXT at baseline and at weeks 1, 2, 4, and 24. General estimating equation models were used to assess differences in susceptibility over time and cross-resistance to other antimicrobials.

RESULTS

Of 181 subjects, 118 (65.1%) were female and the median (range) age was 36 (20 to 72) years. At baseline, E. coli nonsusceptible to SXT was isolated from 23 (53.5%) of 43 patients in group A, 25 (67.6%) of 37 patients in group B, and 37 (64.9%) of 57 patients in group C. The odds ratios (P value) for SXT nonsusceptibility in group C at weeks 1, 2, 4, and 24 compared with baseline were 3.4 (0.013), 3.0 (0.019), 2.9 (0.030), and 1.5 (0.515), respectively. SXT nonsusceptibility was associated with nonsusceptibility to ampicillin, chloramphenicol, ciprofloxacin, and nalidixic acid (P <or= 0.006).

CONCLUSION

In Tanzania, carriage of fecal E. coli nonsusceptible to SXT is common before SXT prophylaxis. Initiation of SXT leads to further loss of susceptibility to SXT and to other antimicrobials.