Page 138 - 15Diarrhoeaandvomiting

Other therapies

Treatment groups were broadly comparable at baseline. Using the full data set (n = 167), the
mean hourly rate of stool production in the first 24 hours was statistically significantly lower in the
racecadotril group (n = 85) (10.5 g/hour) compared with the placebo group (n = 82) (16 g/hour)
(read from graph). The authors estimated that the treatment difference showed that stool output
was approximately 65% of that with placebo (95% CI 44% to 95%; P = 0.025). The mean hourly
rate of stool production in the first 24 hours was also found to be lower in the racecadotril group
(n = 58) (11 g/hour) compared with the placebo group (n = 63) (17.5 g/hour) in the per-protocol
population (read from graph). The authors estimated that the treatment difference showed that
stool output was approximately 65% of that with placebo (95% CI 36% to 90%; P = 0.015). In
the full data set analysis, mean stool output in the first 48 hours was shown to be statistically
significantly reduced in the racecadotril group (n = 84) (9 g/hour) compared with the placebo
group (n = 82) (15 g/hour) (read from graph). The authors estimated that the treatment difference
showed a 60% reduction in stool output in those receiving racecadotril compared with placebo
(95% CI 43% to 88%; P = 0.009). Using covariate analysis in the per-protocol population, the
mean hourly rate of stool production in the first 48 hours was found to be statistically significantly
lower in the racecadotril group (n = 53) (8 g/hour) compared with the placebo group (n = 63)
(16 g/hour) (read from graph). The authors estimated that this indicated a treatment difference of a
50% reduction in stool output (95% CI 33% to 75%; P = 0.001). The authors stated that this effect
was independent of rotavirus status (P = 0.5 for interaction) and that racecadotril was similarly
statistically significantly effective in the rotavirus-positive (8 g/hour versus placebo 19 g/hour)
and rotavirus-negative (6 g/hour versus placebo 13 g/hour) groups (P = 0.001).

Evidence summary
There was evidence from two randomised placebo-controlled trials for the effectiveness of
racecadotril in the treatment of diarrhoea in gastroenteritis. One trial conducted in Peru [EL = 1+]
and one poorly reported European multicentre trial [EL = 1−] found that children under 4 years
given racecadotril (1.5 mg/kg three times daily) and oral rehydration had a reduced total and
average hourly stool output 48 hours after starting treatment compared with children given
placebo and oral rehydration. The effect on total stool output was independent of rotavirus status.
One trial also reported that the rate of stool output was reduced at 24 hours. [EL = 1−] The
average stool output before recovery was found to be reduced, irrespective of the child’s rotavirus
status in one of the trials, which also reported a higher diarrhoeal cure rate at 5 days for children
given racecadotril. [EL = 1+] A ‘cured’ child had passed two consecutive formed stools or no
passage of stool for 12 hours.

8.2.2.2 Bismuth subsalicylate

Three RCTs investigating BSS were identified from the searches. 171–173 One was conducted in
Bangladesh, 171 one in Chile 173 and one in Peru. 172 Two trials had two treatment arms 171,173 and
the third had three. Data were collected from 808 children across the three trials although the
outcomes considered varied and were as follows: onset of persistent diarrhoea, duration of
diarrhoea, intake of oral or IV rehydration and total administration of rehydration solutions, total
stool output, total volume of vomitus and duration of hospitalisation.
171
The largest RCT (n = 451) compared the effects of administering BSS (100 mg/kg per day) to children
presenting with acute diarrhoea with the administration of placebo. The participants were children
aged between 4 and 36 months with a history of acute watery diarrhoea. The exclusion criteria
were use of antimicrobials within the previous 48 hours, blood in the stool, severe malnutrition,
other systemic illness, salicylates intake in the previous 24 hours, allergy to salicylates, or varicella
or measles in the previous 3 months. The methods of randomisation and allocation concealment
were adequate. Children were followed up for the duration of the hospitalisation and another
4 days. The outcomes measured were onset of persistent diarrhoea, duration of acute diarrhoea
(median), total intake of ORS solution and total stool and urine output. [EL = 1+]

The groups were similar at baseline and the loss to follow-up was 8%. The difference in the
proportions of children who developed persistent diarrhoea between treatment groups was not
statistically significant. The duration of diarrhoea was also found to be similar in children receiving
BSS and placebo. However, when considering those children positive to rotavirus, the authors
found a statistically significant reduction in the duration of diarrhoea among the children treated

113