Four trials were included in the review (n=381 children). All trials reported blinding of the participants and included more than 80% of participants in the main analysis. Three trials reported adequate sequence generation, allocation concealment and blinding of service providers and outcome assessors.
Two trials assessed gentamicin and found no significant difference compared with placebo in the number of children with diarrhoea after seven days of treatment. There were also no significant differences for weight gain, fluid intake, energy intake or fluid output.
One trial compared metronidazole with nalidixic acid, metronidazole alone and placebo and found no significant difference for most outcomes except the number of stools, which was lower in the combined treatment group compared with metronidazole alone (MD -1.10, 95% CI -2.07 to -0.13).
One trial compared sulphamethoxazole-trimethoprim with placebo and reported: less diarrhoea after seven days of treatment (RR 0.4, 95% CI 0.16 to 0.99); a lower stool output (MD 179.4g, 95% CI -340.2 to -18.6); and a reduced risk of acquiring a hospital infection (RR 0.06, 95% CI 0.01 to 0.54).
One trial reported no adverse effects; the other trials did not report on adverse effects.