We searched the Cochrane
Infectious Diseases Group Specialized Register (August 2004), CENTRAL (The Cochrane Library Issue 3, 2004), MEDLINE (1966 to August 2004), EMBASE (1974 to August 2004), LILACS (1982 to August 2004), conference proceedings, reference lists, and contacted researchers.
SELECTION CRITERIA: Thirty-three trials met the inclusion criteria; 22 had unclear allocation concealment, and 29 were not blinded. Three trials exclusively included children, and two studied outpatients. The main analysis examined clinical failure, microbiological failure, and relapse. Compared with
chloramphenicol,
fluoroquinolones were not statistically significantly different for clinical (544 participants) or microbiological failure (378 participants) in adults; they reduced clinically diagnosed relapse in adults (OR 0.14, 0.04 to 0.50; 467 participants, 6 trials), but this was not statistically significant in participants with blood culture-confirmed relapse (121 participants, 2 trials). Compared with
co-trimoxazole, we detected no statistically significant difference (82 participants, 2 trials). Among adults,
fluoroquinolones reduced clinical failure compared with
ceftriaxone (OR 0.08, 0.01 to 0.45; 120 participants, 3 trials), but showed no difference for microbiological failure or relapse. We detected no statistically significant difference between
fluoroquinolones and
cefixime (80 participants, 1 trial) or
azithromycin (152 participants, 2 trials). In trials of hospitalized children,
fluoroquinolones were not statistically significantly different from
ceftriaxone (60 participants, 1 trial, involving
norfloxacin) or
cefixime (82 participants, 1 trial).
Norfloxacin had more clinical failures than other
fluoroquinolones (417 participants, 5 trials). Trials comparing different durations of
fluoroquinolone treatment showed no statistically significant differences (693 participants, 8 trials).
AUTHORS' CONCLUSIONS: Many trials were small, and methodological quality varied widely. Although
enteric fever most commonly affects children, trials in this group were particularly sparse. Insufficient data in all comparisons preclude any firm conclusions to be made regarding superiority of
fluoroquinolones over first-line
antibiotics in children and adults.