Page 136 - 15Diarrhoeaandvomiting

Other therapies

the duration of the intervention varied from 2 to 6 days among six studies, and the remaining
three did not report such information. Participants were children aged between 1 and 60 months,
and were inpatients and/or outpatients. Definitions for the outcome measures and the resolution
of diarrhoea were different among the studies. The methodological quality and conduct of the
studies were not uniform. An adequate generation of the allocation sequence was reported
only for three trials and the allocation concealment was appropriate only in one. Three trials
were double-blinded and in only five was an intention-to-treat analysis performed. Data were
extracted for the following outcomes: duration of diarrhoea, frequency of stools, vomiting and
adverse events. The review also reported the proportion of patients without diarrhoea by days 3
and 5 as well as the proportion of those presenting diarrhoea for more than 7 days. All the trials
had better than 80% follow-up. [EL = 1+]

Data from six trials (1076 patients) were pooled in a meta-analysis that showed a statistically
significant reduction in the duration of diarrhoea when smectite was administered
(WMD −22.7 hours; 95% CI −24.8 to −20.6 hours). Two studies that provided data on the
number of stools were pooled. No statistically significant difference was found between the
intervention and the control group in the number of stools during the first 24 hours of follow-up.
However, children treated with smectite had a statistically significant reduction in the number
of stools from the first 24 hour period onwards: at 24 to 48 hours WMD −0.62; 95% CI −1.0 to
−0.2), and at 48 to 72 hours WMD −0.58; 95% CI −0.9 to −0.3). The reviewers pooled the data
from four trials together, finding that by day 3 the proportion of children treated with smectite and
without diarrhoea was statistically significantly higher than the proportion of children that were
not treated with smectite and were without diarrhoea (RR 1.64; 95% CI 1.36 to 31.98). When
results were pooled using a fixed effects model, statistically significantly more children treated
with smectite (122/128) were cured by day 5 compared with those receiving placebo (106/126)
(RR 1.24 (95% CI 1.08 to 1.42). However, as significant heterogeneity was observed (χ² = 8.01;
P = 0.02; I² = 75%), the reviewers also presented results using a random effects model, where
the difference was not statistically significant (RR 1.19; 95% CI 0.93 to 1.53). A funnel plot and
regression asymmetry test (P = 0.23 and the 95% CI included 0) did not show any publication or
other small-sample bias.
One trial showed a reduction in the risk of having diarrhoea for more than 7 days for those
children receiving smectite when compared with the control group (RR 0.60; 95% CI 0.42 to
0.85). No statistically significant difference in the number of vomiting episodes between the
two groups was found when the results of two studies were combined. One RCT reported the
duration of vomiting and showed no statistically significant difference between the two groups.
Another trial found no statistically significant differences between the treatment groups for the
incidence of vomiting on day 1 and or on day 3 of the intervention. No statistically significant
differences in adverse events associated with treatments were identified and three RCTs reported
no adverse events associated with short-term treatment with smectite.

Evidence summary
There was evidence from a well-conducted systematic review [EL = 1+] for the effectiveness of
smectite in the treatment of children with diarrhoea. The review showed that children receiving
smectite had a reduction in the frequency and duration of diarrhoea, experienced a higher
resolution of diarrhoea by day 3 and were less likely to have diarrhoea for more than 7 days. On
the other hand, the review showed no statistically significant differences between the smectite
group and the control group in the number of episodes or duration of vomiting or in the resolution
of diarrhoea by day 5. Statistically significant differences in adverse effects between the smectite
and control groups were not reported and some trials reported that no adverse events occurred.

8.2.2 Antisecretory agents
Evidence overview

Two randomised placebo-controlled trials of racecadotril 169,170 and three trials of BSS were
identified. 171–173

111