Evidence review

This evidence review is based on the 2 largest studies (Prado 2002 and Vetel et al. 1999) comparing racecadotril with an active comparator for treating acute diarrhoea in adults. Although other published studies are available they were excluded because they included fewer patients (less than 70) or used placebo as a comparator.

A multinational comparison of racecadotril and loperamide in the treatment of acute watery diarrhoea in adults ( Prado 2002 )

  • Design: single-blind, randomised controlled trial in 21 centres in 14 countries (Brazil, Cameroon, Costa Rica, Guatemala, Indonesia, the Ivory Coast, Kenya, Nigeria, Mexico, Morocco, Pakistan, the Philippines, Tunisia and Vietnam).

  • Population: 945 outpatients aged 18 years or over (mean age 36 years, mean weight about 62 kg) with acute watery diarrhoea of presumed infectious origin, present for between 24 hours and 5 days, and at least 3 watery stools within the previous 24 hours (mean number of diarrhoeic stools in the previous 24 hours about 6.5; mean duration of diarrhoea 2.1 days).

  • Intervention and comparison: patients were randomised to racecadotril 100 mg 3 times daily or loperamide 2 mg 3 times daily until diarrhoea resolved (12 hours without stools or 2 consecutive normal stools) or for a maximum of 7 days. Concomitant medication other than paracetamol and oral rehydration salt (ORS) solution was not permitted during the study.

  • Outcomes: the primary outcome was duration of diarrhoea, defined as the time from the first dose of study drug to the appearance of the first formed stool. Secondary outcome measures included overall clinical response (clinical success or failure at 10–14 days) and occurrence and duration of abdominal pain and distension. The main safety outcomes were adverse events and occurrence of constipation (no stools for 36 hours or more).

Table 1 Summary of the study: Prado (2002)

Racecadotril

Loperamide

Analysis a

Randomised

n=473

n=472

Efficacy

n=473

n=471

Primary outcome:

duration of diarrhoea

Not stated

Not stated

Hazard ratios not stated. However, the authors report that the difference in recovery rates at 72 hours is in keeping with equivalence (less than +/−5% difference between the treatments)

Median duration of diarrhoea

55.0 hours (95% CI 50.0 to 65.0 hours)

55.0 hours (95% CI 48.0 to 66.0 hours)

Significance not given

Selected secondary outcomes:

Overall clinical response

92%

93%

Significance not given

Median duration of abdominal pain

11 hours

10 hours

Significance not given

Median duration of abdominal distension

5.4 hours

24.4 hours

p=0.0001

Safety

n=473

n=472

Any adverse event

14.2%

(67/473)

23.9%

(113/472)

p=0.001

Constipation

16%

(74/473)

25%

(116/472)

p=0.001

Abbreviations: CI, confidence interval.

a Initial identification of the intention to treat (ITT) to treat population excluded almost 25% of the treated population, largely because of missing data points. Missing data were replaced using unspecified predetermined criteria and a further analysis performed. It is this post-hoc analysis that is presented here, although the study authors state that the overall conclusion for both ITT analyses did not differ.


Comparison of racecadotril and loperamide in adults with acute diarrhoea (
Vetel et al. 1999 )

  • Design: double-blind, randomised controlled trial in 34 general practices in France.

  • Population: 157 outpatients aged 18 years or over (mean age 41 years) with acute diarrhoea, defined as at least 3 soft or liquid stools for between 24 hours and 5 days (mean number of diarrhoeic stools in the previous 24 hours 5.6; mean duration of diarrhoea about 40 hours).

  • Intervention and comparison: patients were randomised to racecadotril (100 mg 3 times daily) or loperamide (4 mg initially then 2 mg after each diarrhoeic stool) until diarrhoea resolved (12 hours without stools or 2 consecutive normal stools) or for a maximum of 7 days. A double-placebo design was used in the study because the 2 drugs had different dosing regimens: patients randomised to racecadotril also received placebo capsules to take after each diarrhoeic stool and patients randomised to loperamide received placebo capsules to take 3 times a day. Paracetamol could be administered if the patient developed a fever but patients were withdrawn from the study if other concomitant medication was needed.

  • Outcomes: the primary outcome measure was the number of diarrhoeic stools passed by the patient until recovery. Duration of diarrhoea and the change in associated symptoms and signs were also assessed. Safety outcomes included adverse events and the occurrence of constipation (no stools for at least 2 days).

Table 2 Summary of the study: Vetel et al. (1999)

Racecadotril

Loperamide

Analysis a

Randomised

n=82

n=75

Efficacy a

n=77a

n=70a

Primary outcome: mean number of diarrhoeic stools passed until recovery +/− SEM

3.5+/−0.5

2.9+/−0.4

Significance not given

Selected secondary outcomes:

Mean duration of diarrhoea

+/− SEM

14.9+/−2.0 hours

13.7+/−2.2 hours

Significance not given

Safety

Number of patients not given

Any adverse event

7.4%

12.0%

Significance not given

Constipation

9.8%

18.7%

No significant difference (p value not stated)

Abbreviations: SEM, standard error of the mean.

a Vetel et al. state an intention to treat analysis was done. However, 5 patients from each group did not complete their evaluation sheets accurately and could not be evaluated for efficacy.

Clinical effectiveness

Prado (2002) found that racecadotril and loperamide were similarly effective: diarrhoea resolved in a median of 55 hours with both treatments. By days 10–14, 92% of patients taking racecadotril and 93% of patients taking loperamide had recovered.

The median duration of abdominal distension was significantly shorter with racecadotril compared with loperamide (5.4 hours compared with 24.4 hours, p=0.0001) but the median duration of abdominal pain was similar in both treatment groups (11 hours compared with 10 hours respectively, significance not given).

Vetel et al. (1999) found that patients receiving racecadotril and loperamide passed a similar number of stools until their recovery from diarrhoea (mean 3.5 compared with 2.9, significance not given) and the mean duration of diarrhoea was similar in both treatment groups (14.9 hours compared with 13.7 hours respectively, significance not given).

Safety

Prado (2002) found that significantly more patients in the loperamide group had an adverse event: 23.9% compared with 14.2% in the racecadotril group (p=0.001). These were considered to be treatment related in 18% of patients taking loperamide and 9% of patients taking racecadotril (significance not given).

The most common treatment-related adverse events were constipation, abdominal distension, anorexia, headache and abdominal pain. Constipation was experienced by 16% of patients taking racecadotril and 25% of patients taking loperamide (p=0.001) and was considered to be treatment related in 3.4% and 12.5% respectively (significance not given). Constipation and abdominal pain and distension are recognised adverse effects of loperamide.

Vetel et al. (1999) found that 7.4% of patients in the racecadotril group reported adverse events compared with 12.0% in the loperamide group (significance not given). More patients taking loperamide experienced constipation (18.7% compared with 9.8% of patients taking racecadotril). However, the difference was not found to be statistically significant. It is possible that this analysis may not have been sufficiently statistically powered to be able to detect a difference between the groups should one exist.

More than 37 million adults have been treated with racecadotril in Europe and overall exposure is more than 40 million adults worldwide (Abbott Healthcare Products Limited: personal communication January 2013). According to the summary of product characteristics, the adverse effect most commonly reported with racecadotril is headache (frequency of more than 1 in 100). Rash and erythema have been reported uncommonly (frequency of between 1 in 1000 and 1 in 100).

Evidence strengths and limitations

These 2 studies suggest that racecadotril and loperamide are similarly effective in reducing the duration of diarrhoea and the number of stools produced, and racecadotril has fewer adverse effects than loperamide, particularly constipation. However, both studies have limitations affecting their validity and application to UK practice.

The larger study by Prado (2002) was undertaken in 14 countries where the aetiology and severity of diarrhoea is likely to differ from this country, limiting its applicability to UK practice. The mean duration of diarrhoea at study entry was 2 days (whereas according to the Health Protection Agency most infectious diarrhoea in the UK is self-limiting and nearly half of episodes last less than 1 day); only 14% of patients were reported by the study authors as 'Caucasian'; and the average weight of patients was only 62 kg.

The dosage of loperamide used in the Prado study (2 mg 3 times daily) did not reflect UK practice. The BNF recommends 4 mg initially followed by 2 mg after each diarrhoeic stool (usual dose 6–8 mg daily, maximum 16 mg daily). It is possible that loperamide might be more effective than racecadotril 100 mg 3 times a day at dosages greater than 6 mg daily. However, more adverse effects might be seen with higher dosages of loperamide.

Although the investigators responsible for the clinical evaluation of the patients were unaware which medication each patient received, patients were aware of the treatment (single-blinding) which may have introduced bias. In addition, the method of allocation was not described and it is unclear whether allocation was concealed; unconcealed allocation is another potential source of bias.

The smaller study by Vetel et al. (1999) was conducted in France where the aetiology and severity of diarrhoea is likely to be similar to the UK. However, it has limitations which may affect its validity.

No measures of statistical significance or likely certainty, such as p values or confidence intervals, were given for the comparisons between racecadotril and loperamide. Therefore, it is difficult to estimate the true effects of racecadotril. The Vetel study was double blind but it is unclear whether allocation was concealed.

The mean ages of patients included in the 2 studies (Prado 2002 and Vetel et al. 1999) were 36 years and 41 years respectively. Therefore, the results may not be applicable to older people, the group of adults who are at greatest risk of dehydration and death from acute diarrhoea (see the Clinical Knowledge Summary on gastroenteritis). The incidence of severe dehydration or hospitalisation was not assessed in either study.

An additional small double-blind randomised controlled trial in nursing homes in Italy (Gallelli et al. 2010) compared racecadotril (100 mg 3 times daily) and loperamide (4 mg initially then 2 mg after each diarrhoeic stool, maximum 8 mg daily) in 61 older people (mean age 82 years) with acute diarrhoea and no signs of acute dehydration or bacterial infection. It found that racecadotril was statistically significantly more effective than loperamide. The duration of diarrhoea was 36 hours in the racecadotril group and 63 hours in the loperamide group (p<0.01). More patients in the loperamide group experienced adverse events (60% compared with 12%, significance not given); the most frequently reported adverse events were constipation and nausea.

The authors considered that the low efficacy of loperamide in this study could be related to drug interactions: all patients were taking potentially interacting drugs. The study was double blind but details of blinding of the 2 different dosage regimens are not given and it is not clear whether allocation was concealed. The study included only 61 patients. Small studies have limitations and need to be interpreted carefully because they do not normally yield reliable or precise estimates (Hackshaw 2008).