Page 139 - 15Diarrhoeaandvomiting

Diarrhoea and vomiting caused by gastroenteritis in children under 5 years

with BSS (median 56 hours) when compared with the children receiving placebo (median 72 hours)
(P = 0.03). Children treated with BSS had a statistically significant lower output of stool plus urine
(mean 386 g/kg, SD 248 g/kg) than the children receiving placebo (mean 438 g/kg, SD 272 g/kg),
(WMD −52 g/kg; 95% CI −100 to −4 g/kg). There was no statistically significant difference between
the study groups in the mean total intake of ORS solution. While no defined adverse reactions were
observed, it was reported that two children had ‘black tongue’ during treatment.
The second RCT (n = 142) 173 examined the effects of treating children aged between 4 and
36 months and suffering from diarrhoea and dehydration with BSS. The study compared
the effects of administering BSS (100 mg/kg) for 5 days (n = 72) with placebo (n = 70). The
exclusion criteria were symptoms that had lasted for more than 72 hours, blood in stools, severe
malnutrition, antibiotic use in the previous 48 hours, a salicylate intake greater than 20 mg/kg
in the previous 12 hours, allergy to bismuth or to salicylate or acute illness inconsistent with a
diarrhoeal state. The method of randomisation was not reported but allocation concealment was
adequate. Patients were monitored in hospital for at least 5 days and then were followed for 3
more days (whether they remained in hospital or were discharged). Outcomes measured were
disease duration (hours), time to last loose/watery stool, time until last unformed stool, duration
of hospital stay and IV fluid intake (ml/kg). [EL = 1+]
The groups were similar at baseline and the loss to follow-up was 13.4%. The use of BSS
demonstrated statistically significant benefits compared with placebo in terms of shortening the
duration of diarrhoea. The mean time to last loose/watery stool was shorter in the group treated
with BSS (mean 73.4 hours) compared with the group receiving placebo (mean 107.5 hours)
(P < 0.02). The mean time until last unformed stool was also statistically significantly shorter
in the group treated with BSS (mean 130.4 hours) compared with the group receiving placebo
(mean 170 hours) (P < 0.01). The need for IV rehydration therapy was lower in the group treated
with BSS than in the placebo group. The difference was reported as statistically significant (data
for this outcome have been extracted from a histogram and therefore are estimates). At day 3, the
intervention group received on average 30 ml/kg and the control group 45 ml/kg. At day 5, the
intervention group intake was on average 20 ml/kg and the control group 42 ml/kg. There was a
statistically significant reduction in the duration of hospitalisation among the BSS group compared
with the placebo group. The mean of hospital stay for the intervention group was 6.9 days, while
for the control group it was 8.5 days (P = 0.01). No adverse reactions were observed.
An RCT with three treatment arms (n = 215) compared the effects of administering BSS to treat
172
diarrhoea with that of placebo. Participants were boys aged between 6 and 59 months who had
acute diarrhoea (three or more watery stools in the preceding 24 hours). The exclusion criteria were
presence of blood in the stools, diarrhoea for more than 5 days, use of antibiotics, antidiarrhoeal
medication or any treatment with acetylsalicylic acid in the 72 hours before admission, clinical
evidence of another illness requiring antibiotic therapy, severe malnutrition, allergy to salicylate or
to bismuth or exclusive breastfeeding. The boys were randomised to treatment with BSS 100 mg/
kg (n = 108), BSS 150 mg/kg (n = 108) or placebo (n = 107) every 4 hours for 5 days or until the
diarrhoea stopped. The methods of randomisation and allocation concealment were adequate.
The participants were followed up while in hospital and the outcomes assessed were duration of
diarrhoea (proportion of patients with diarrhoea by day 5), total stool output (ml/kg), total volume
of vomitus (ml/kg), total intake of rehydration (ml/kg) and duration of hospital stay (days). [EL = 1+]
The groups were similar at baseline and the loss to follow-up was 8% of 275 patients initially
enrolled. Diarrhoea stopped within 5 days of admission in 76/85 (89%) children treated with
100 mg/kg BSS, in 73/83 (88%) children treated with 150 mg/kg BSS and in 62/84 (74%) children
receiving placebo. By day 5, diarrhoea had ceased in statistically significantly more children in
each of the two intervention groups compared with the control group (100 mg/kg BSS RR 1.21;
95% CI 1.05 to 1.40, and 150 mg/kg BSS RR 1.19; 95% CI 1.03 to 1.38). The mean total stool
output was 182 ml/kg (SD 197 ml/kg) in children treated with 100 mg/kg BSS (n = 85), 174 ml/
kg (SD 159 ml/kg) in children treated with 150 mg/kg BSS (n = 83) and 260 ml/kg (SD 254 ml/kg)
in children receiving placebo (n = 84). The mean total stool output was statistically significantly
reduced in each of the two intervention groups compared with the control group (100 mg/
kg BSS WMD −78 ml/kg; 95% CI −147 to −9 ml/kg, and 150 mg/kg BSS WMD −86 ml/kg;
95% CI −150 to −22 ml/kg). The mean total volume of vomitus was 11.6 ml/kg (SD 19.6 ml/kg)
in children treated with 100 mg/kg BSS (n = 85), 8.7 ml/kg (SD 18.3 ml/kg) in children treated

114