Summary of findings
Description of the condition
Schistosomiasis is a parasitic blood fluke infection, of which three species commonly infect humans; Schistosoma mansoni (common in the tropics and sub-tropics), S. haematobium (mostly endemic in Africa and the Middle East) and S. japonicum (endemic in the People's Republic of China and the Philippines) (Engels 2002; WHO 2002; Gryseels 2006; Steinmann 2006; Utzinger 2009). It has been estimated that 779 million people are at risk of schistosomiasis worldwide and 207 million people may be infected (Steinmann 2006). Of these, 120 million people are estimated to be symptomatic and 20 million suffer from long-term complications (Chitsulo 2000; WHO 2002; van der Werf 2003). In global burden of disease estimates, schistosomiasis causes 1.7 to 4.5 million disability-adjusted life years (DALYs) (WHO 2002; WHO 2004; Hotez 2006; Steinmann 2006; Utzinger 2009). Some suggest that this value may underestimate the true burden of schistosomiasis (WHO 2002; van der Werf 2003; King 2005; King 2007; King 2008a; King 2010).
People infected with S. mansoni excrete the fluke eggs in their faeces, and faecal contamination of freshwater allows these eggs to hatch into larvae (miracidia) which penetrate a specific freshwater snail (the intermediate host). Within the snail, the miracidia develop into cercariae (the infective larvae), which can penetrate a person’s skin upon contact with contaminated water bodies.
Following infection, the worms migrate through the human venous system, via the right chamber of the heart and the lungs, and through the mesenteric arteries and the liver via the portal vein, before finally settling in the superior mesenteric veins which drain the large intestine. Here, male and female worms mature, pair up and the female worms start to produce eggs (≂ 300 per day) (Davis 2009). An adult worm usually lives for three to five years, but some can live up to 30 years (Gryseels 2006). The eggs produced by the worms traverse the intestinal wall to be excreted in the faeces, and in the process some become trapped and initiate inflammatory reactions, which cause the underlying pathology and symptomatic illness (Richter 2003a; King 2008b). Early symptoms depend on the severity of infection (Gryseels 1987), and if treatment is not provided early, chronic illness and long-term serious disease can follow.
Symptoms and effects
Schistosomiasis mansoni can present as an acute or chronic illness. The acute illness, or Katayama syndrome, is caused by migrating and maturing schistosomula that may result in a systemic hypersensitivity reaction characterized by fever, feeling of general discomfort (malaise), muscle pain (myalgia), fatigue, non-productive cough, diarrhoea (with or without blood), and pain in the upper right part of the abdomen just below the rib cage. Chronic and advanced disease results from the host's immune response to schistosome eggs deposited in tissues and the granulomatous reaction evoked by the antigens they secrete and is characterized by non-specific intestinal symptoms, such as abdominal pain, diarrhoea and blood in the stool (Gryseels 1992; Gray 2011; Gryseels 2012).
Inflammatory reactions in the liver lead to hepatosplenic schistosomiasis, a key feature of chronic infection, which can manifest within a couple of months for heavy infections or many years after light infections. The chronic inflammation produces fibrotic lesions, which in turn lead to liver cirrhosis that progressively occludes the portal system giving rise to portal hypertension. The portal hypertension eventually leads to enlargement of hepatic arteries, and the associated oesophageal varices may rupture with heavy blood loss, haemorrhagic shock and death. The patient may also suffer repeated episodes of variceal bleeding – the primary cause of death in hepatic schistosomiasis (Andersson 2007). Severity of disease depends upon the intensity and duration of infection (Naus 2003), but recent evidence suggests the presence of the infection alone determines morbidity (King 2008a).
S. mansoni infection overlaps in distribution with S. haematobium in some areas of sub-Saharan Africa resulting in mixed infections (WHO 2002). Unlike S. mansoni, the main early symptoms of S. haematobium infection are blood in urine (haematuria) and painful urination (dysuria). Chronic and advanced disease is insidious and may result in structural damage to the bladder wall which may eventually lead to kidney failure.
Definitive diagnosis of S. mansoni infection is by microscopy for parasite eggs in the stool. Quantitative methods are recommended for epidemiological purposes because they allow estimation of intensity and evaluation of the impact of control programmes not only in terms of cure rate but also egg reduction rate (WHO 1985; Doenhoff 2004; Bergquist 2009). The Kato-Katz technique (Katz 1972) is the most common quantitative technique (Booth 2003). Recently, the FLOTAC technique has been applied for the detection and quantification of S. mansoni eggs in stools with promising results and hence warranting further investigation (Glinz 2010).
Egg output can be influenced by several factors, such as day-to-day, intra-stool, and seasonal variations as well as environmental conditions (Braun-Munzinger 1992; Engels 1996; Engels 1997; Enk 2008). Therefore negative results following microscopic examination of a single stool are unreliable (de Vlas 1992; Kongs 2001; Booth 2003; Enk 2008), and measurement of prevalence and intensity of infection by egg count has shortcomings (Gryseels 1996; de Vlas 1997; Utzinger 2001a). Rectal biopsy is more sensitive than microscopy and is occasionally done when repeated stool examinations are negative for eggs. However, this method is unsuitable for use in population-based control programmes (Allan 2001).
A monoclonal antibody-based dipstick is increasingly being used for the diagnosis of the infection with promising results (Polman 2001; Legesse 2007; Legesse 2008; Caulibaly 2011). A more specific and sensitive diagnostic technique based on polymerase chain reaction (PCR) is increasingly being used in some reference laboratories in Europe (Sandoval 2006; Cnops 2012; Enk 2012). Ultrasound is used for diagnosing and assessing infection-related pathology (Hatz 1990; Mohamed-Ali 1991; Doehring-Schwerdtfeger 1992; Hatz 2001; Richter 2003b).
Clinically, intestinal schistosomiasis is diagnosed on the basis of presence of blood in stool, (bloody) diarrhoea, and abdominal pain, but these are non-sensitive and non-specific (Gryseels 1992; Utzinger 2000c; Danso-Appiah 2004) as diarrhoea or blood in stool can be due to other causes such as hookworm infection, dysentery and typhoid fever.
Description of the intervention
Schistosomiasis control measures implemented before the 1970s – when efficacious antischistosomal drugs were not available – focused mainly on interrupting transmission with molluscicides to kill the intermediate host snails (WHO 1985; Sturrock 2001). The 1970s marked the turning point in schistosomiasis control when efficacious drugs that can be applied in a single oral dose were discovered, shifting the control emphasis from transmission control to chemotherapy-based morbidity control (WHO 1985; Cioli 1995). A body of evidence suggests that morbidity due to schistosomiasis can be prevented and pathology reversed with available antischistosomal treatments (Mohamed-Ali 1991; Doehring-Schwerdtfeger 1992; Savioli 2004; Zhang 2007; Webster 2009; Koukounari 2010).
Mass drug administration, or treatment of infected individuals or entire 'at-risk' populations (eg school-aged children), usually without prior diagnosis - an approach termed 'preventive chemotherapy', is the control strategy currently pursued by the World Health Organization (WHO) and applied in many endemic countries (WHO 2006). Usually, praziquantel at a single 40 mg/kg oral dose is used (Fenwick 2009), but still there are uncertainties regarding this dose. An exception is Brazil where the national policy adopted since 1995 recommends a single oral dose of 60 mg/kg for children aged between two and 15 years, and 50 mg/kg for adolescents and adults (Favre 2009). The recently adopted policy for schistosomiasis control in Brazil disapproves of treatment without prior diagnosis, and therefore the preventive chemotherapy strategy is no longer applied in Brazil (Favre 2009).
Oxamniquine has also been used extensively for the control of schistosomiasis mansoni in different endemic countries, most notably Brazil, where more than 12 million doses of oxamniquine have been administered by the national schistosomiasis control programme (Katz 2008). There are uncertainties around the standard dose of oxamniquine (Foster 1987; Cioli 1995). Therefore, the WHO recommends total doses of 20 to 60 mg/kg (in divided doses of up to 20 mg/kg) (WHO 2001).
More recently, the artemisinin derivatives used in the treatment of malaria have been shown to have antischistosomal properties, particularly against the immature developing stages of the schistosome parasites (Borrmann 2001; Utzinger 2007). Praziquantel, in contrast, acts against the adult worms and the very young schistosomula just after skin penetration (Sabah 1986; Utzinger 2007).
The current emphasis of schistosomiasis control is to reduce the burden of disease in high endemicity areas and to interrupt transmission in low endemicity areas (WHO 2002). Intensity of infection is highest in school-aged children and adolescents, therefore preventive chemotherapy is targeted especially to these at-risk groups (Magnussen 2001; WHO 2002; Savioli 2004; Savioli 2009).
Why it is important to do this review
Currently, entire control and treatment programmes are based on praziquantel and there is risk of drug resistance and perhaps shortages of praziquantel. There is a need to assess alternative drugs or combinations. Still there are uncertainties around effective and safe dosage of praziquantel and standard doses of oxamniquine. There are also uncertainties about adequacy of current adult doses used in children.
To evaluate the effects of antischistosomal drugs, used alone or in combination, for treating S. mansoni infection.
Criteria for considering studies for this review
Types of studies
Randomized controlled trials.
Types of participants
Individuals infected with S. mansoni diagnosed microscopically for the presence of S. mansoni eggs in stool using the Kato-Katz technique (Katz 1972), or any other quantitative diagnostic method, such as the quantitative oogram and FLOTAC techniques.
Types of interventions
The following comparisons are evaluated in this review:
- Antischistosomal drugs alone or in combination versus placebo;
- Antischistosomal drugs alone or in combination versus a different dose of the same antischistosomal drug; and
- Antischistosomal drugs alone or in combination versus different antischistosomal drugs alone or in combination.
Trials that allocated non-schistosomal drug or interventions in addition to the treatment and control of interest were eligible provided the same drug was allocated to both treatment and control groups.
Types of outcome measures
- Parasitological failure, defined as treated individuals who remained positive for S. mansoni eggs in stool using the standard Kato-Katz or other quantitative techniques (follow-up: up to one month).
- Egg reduction rate, defined as percent reduction in S. mansoni egg count after treatment (follow-up: up to 12 months).
- Parasitological failure (follow-up: greater than one month).
- Resolution of symptoms (eg abdominal pain, diarrhoea and bloody diarrhoea).
- Resolution of pathology (eg hepatomegaly, splenomegaly, portal fibrosis, cirrhosis of the liver or colonic polyps) measured by ultrasound, by standard international classification or other standardized methods (CWG 1992).
- Non-serious adverse events.
- Serious adverse events (ie any untoward medical occurrence or effect that at any dose: results in death; is life-threatening; requires hospitalisation or prolongation of existing inpatients' hospitalisation; results in persistent or significant disability or incapacity; is a congenital anomaly or birth defect).
Search methods for identification of studies
We attempted to identify all relevant trials regardless of language or publication status (published, unpublished, in press, under review and in progress).
We searched the following databases using the search terms and strategy described in Table 1: Cochrane Infectious Diseases Group Specialized Register (October 2012); Cochrane Central Register of Controlled Trials (CENTRAL), published in the Cochrane Library; MEDLINE (1966 to October 2012); EMBASE (1974 to October 2012); and LILACS (1982 to October 2012). We also searched the metaRegister of Controlled Trials (mRCT) in October 2012 using ’Schisto * mansoni' as the search term.
Searching other resources
Researchers and organizations
We contacted individual researchers working on antischistosomal drugs, pharmaceutical industries and experts from the UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR) for unpublished data and ongoing trials.
We checked the reference lists of all studies identified by the aforementioned methods for additional relevant studies.
Data collection and analysis
Selection of studies
Vittoria Lutje, the Cochrane Infectious Diseases Group (CIDG) Information Retrieval Specialist, searched the literature and retrieved studies using the search strategy outlined in Table 1. Anthony Danso-Appiah (ADA) screened the results to identify potentially relevant trials, obtained the full trial reports and assessed the eligibility of trials for inclusion in the review using an eligibility form based on the inclusion criteria. Jürg Utzinger (JU) independently verified the eligibility assessment results.
ADA contacted the authors of potentially relevant trials for clarification if eligibility was unclear. We excluded studies that did not meet our inclusion criteria and we have detailed the reasons for exclusion in the Characteristics of excluded studies. This was verified independently by JU and Piero L. Olliaro (PLO). We resolved any discrepancies through discussion between the authors.
Data extraction and management
ADA extracted trial characteristics such as methods, participants, interventions and outcomes, and recorded on standard forms, which were independently verified by JU. ADA and JU resolved discrepancies through discussion, and where necessary contacted a third author (PLO). ADA contacted trial authors for clarification, or insufficient or missing data when necessary.
We extracted the number of participants randomized and the number of patients followed-up in each treatment arm. For dichotomous outcomes, we recorded the number of participants experiencing the event in each treatment group of the trial. For continuous outcomes summarized as geometric means, we extracted means and their standard deviations (SD) on the log scale. If the data were summarized as arithmetic mean, we extracted the means and their SDs. We extracted medians or ranges when they were reported to summarize the data.
For each outcome, we extracted data for each follow-up time reported in the trial report.
Assessment of risk of bias in included studies
ADA assessed the risk of bias of each trial using The Cochrane Collaboration's risk of bias tool (Higgins 2011) and the assessment results were verified independently by Dave Sinclair (DS). Where information in the trial report was unclear, we attempted to contact the trial authors for clarification. We assessed the risk of bias for six domains: sequence generation, allocation concealment, blinding (investigators, outcome assessors and participants), incomplete outcome data, selective outcome reporting and other sources of bias. For each domain, we made a judgment of 'low risk' of bias, 'high risk' of bias or 'unclear'. We resolved any discrepancies by discussion between the authors.
Measures of treatment effect
We presented dichotomous outcomes using risk ratios (RR). Mean differences (MD) were used as the measure of effect for continuous outcomes that were summarized as arithmetic means. We used geometric mean ratios for continuous outcomes that were summarized as geometric means. We presented all results with 95% confidence intervals (CI).
Dealing with missing data
We analysed data based on the number of patients for whom an outcome was recorded (complete case analysis).
Assessment of heterogeneity
We assessed heterogeneity by inspecting the forest plots for overlapping CIs and outlying data; using the Chi
Assessment of reporting biases
We would have attempted to explore publication bias using funnel plots if there were sufficient number of trials in the comparisons.
We used Review Manager (RevMan) to perform the statistical analyses. We stratified the analyses by: comparison; the dose of the drug; and the length of follow-up time. We used meta-analysis to combine the results across trials. When heterogeneity was detected, we used a random-effects meta-analysis approach; otherwise a fixed-effect approach was adopted. We tabulated adverse events and also data that could not be meta-analysed.
Subgroup analysis and investigation of heterogeneity
When heterogeneity was detected, we planned to carry out subgroup analyses to explore potential causes. Subgroupings would be as follows: patient age (children versus adults); and intensity of infection (< 500 eggs per gram of stool versus > 500 eggs per gram of stool).
We conducted a subsidiary, non-randomized comparison of failure rates in children with failure rates in adults for the same drug and same dose (mg/kg) to explore issues around dose applicability in children.
Where data were sufficient we planned to conduct sensitivity analyses to assess the robustness of the results to the risk of bias components.
Description of studies
We identified 52 trials (10,269 participants) which met the inclusion criteria (see Characteristics of included studies). We managed one multicentre trial carried out in Brazil, Mauritania and Tanzania as three separate trials in the analysis (Olliaro 2011 BRA; Olliaro 2011 MRT; Olliaro 2011 TZA), and three papers contained multiple individual studies which we again managed separately (de Clarke 1976a ZWE; de Clarke 1976b ZWE; de Clarke 1976c ZWE; de Clarke 1976d ZWE; Katz 1979a BRA; Katz 1979b BRA; Gryseels 1989a BDI; Gryseels 1989b BDI; Gryseels 1989c BDI).
Of the 52 trials we identified, 19 evaluated praziquantel, 17 evaluated oxamniquine and 12 directly compared praziquantel with oxamniquine. In addition, two compared myrrh (mirazid) with praziquantel, and two compared different brands of praziquantel.
For the two primary outcomes, 47 trials reported cure rate or failure rate, 34 trials reported egg reduction rate and 33 trials reported both outcomes. Only Sukwa 1993 ZMB reported reinfection rate.
For secondary outcomes, five trials (Rugemalila 1984 TZA; Gryseels 1989a BDI; Gryseels 1989b BDI; Gryseels 1989c BDI; Sukwa 1993 ZMB) reported clinical improvement or functional indices, but we could not include Rugemalila 1984 TZA and Sukwa 1993 ZMB in the meta-analysis because of insufficient information. Thirty-three trials reported adverse events.
In the study by de Jonge 1990 SDN, we excluded the two arms that received metrifonate and placebo respectively from the analysis. Also, we excluded one arm of the study by Ibrahim 1980 SDN involving participants who did not have S. mansoni infection and also one arm each of the trials by Rugemalila 1984 TZA and Taylor 1988 ZWE that did not receive treatment from the analysis.
The trial by Tweyongyere 2009 UGA assessing the effects of praziquantel was a nested cohort study within a larger mother and baby cohort study in which pregnant women found to be infected with S. mansoni were randomized to receive praziquantel or placebo. We obtained data on parasitological failure rate and clinical improvement from figures (Gryseels 1989a BDI; Gryseels 1989b BDI; Gryseels 1989c BDI), but it was not possible to extract egg count data.
Trial setting and participants
The trials were conducted in Africa (n = 36), South America (n = 15; all in Brazil) and the Middle East (n = 1). Eight trials were conducted in the late 1970s, 28 in the 1980s, seven in the 1990s and only nine since the year 2000.
Eighteen trials involved children, 12 trials recruited adults, and 22 recruited whole populations comprising children, adolescents and adults.
Seventeen trials recruited participants from the outpatient clinics, six did not specify the setting whilst one trial (Omer 1981 SDN) consisted of both participants identified in a field survey and those attending the hospital; two trials (Katz 1979a BRA; Katz 1979b BRA) involved military officers in a Barracks who became exposed to the infection during training and another trial (Ibrahim 1980 SDN) recruited university students on campus. The remaining 25 trials recruited participants through community surveys.
Risk of bias in included studies
|Figure 1. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.|
|Figure 2. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
We considered 16 trials as low risk of bias with regard to the generation of the randomization sequence (Figure 2). In the remaining 36 trials, the methods used to generate the sequence of allocation were not described and therefore the risk of bias is unclear.
Fourteen trials adequately described allocation concealment and had a low risk of bias. One trial did not conceal allocation (Fernandes 1986 BRA); and the methods were unclear in the remaining 37 trials (Figure 2).
Twenty-seven trials employed blinding and stated who was blinded. However, none described the methods of blinding. Nevertheless, the studies were considered to be at low risk of bias. One trial did not employ blinding (Fernandes 1986 BRA) and we therefore classed it at high risk of bias; whereas in 25 trials blinding was unclear (Figure 2).
Incomplete outcome data
We considered the risk of bias for incomplete outcome data to be low in 17 trials (Figure 2). We deemed the risk of bias to be high in 19 trials, and in the remaining 16 trials as unclear.
All 52 trials had low risk of selective outcome reporting (Figure 2).
Other potential sources of bias
Overall, 42 trials were considered to be free from other biases and the level of bias was unclear in the remaining 10 trials (Figure 2).
Effects of interventions
See: Summary of findings for the main comparison Praziquantel 40 mg/kg for treating S. mansoni infection; Summary of findings 2 Oxamniquine 40 mg/kg for treating S. mansoni infection; Summary of findings 3 Oxamniquine 40 mg/kg versus praziquantel 40 mg/kg; Summary of findings 4 Artesunate (12 mg/kg) plus praziquantel (40 mg/kg) versus praziquantel (40 mg/kg) alone
Section 1. Monotherapies
Nineteen trials, conducted in Africa, Brazil and the Arabian Penunsula, evaluated praziquantel. Four studies compared praziquantel with placebo, and 17 trials directly compared different dosing schedules of praziquantel with the standard dose of 40 mg/kg.
Analysis 1: Praziquantel versus placebo
Two trials from Kenya and Uganda used the WHO recommended dose of 40 mg/kg. Praziquantel 40 mg/kg achieved parasitological cure in 57% and 82% of the patients respectively, compared to placebo where almost all continued to excrete eggs at one to two months (RR 3.13, 95% CI 1.03 to 9.53, two trials, 414 participants, Analysis 1.1).
In addition, one small trial from Brazil compared three different doses of praziquantel with placebo and presented outcomes at six and 12 months. All patients given 40 mg/kg and 60 mg/kg praziquantel achieved parasitological cure at six months, while two out of five patients given 20 mg/kg and almost all those given placebo continued to excrete eggs (one trial, 40 participants, Analysis 1.2). At 12 months, reinfection was demonstrable in some of those given praziquantel ( Analysis 1.3). One further trial from Brazil gave 60 mg/kg praziquantel each day for three days and achieved 100% parasitological cure at six months compared to almost complete failure with placebo (one trial, 55 participants, Analysis 1.2).
None of these trials reported on percentage egg reduction.
No serious adverse events were recorded in these trials but transient dizziness and abdominal pain appeared to be more commonly reported with praziquantel than placebo (seven trials, 1255 participants, Table 2).
Analyses 2 and 3: Lower doses praziquantel versus 40 mg/kg
Lower doses (20 mg/kg to 30 mg/kg) have been evaluated in Zimbabwe, Burundi, Sudan and Brazil. Compared to 40 mg/kg, parasitological failure at one month was more than double with the 20 mg/kg dose, and 50% higher with the 30 mg/kg dose (20 mg/kg: RR 2.23, 95% CI 1.64 to 3.02, two trials, 341 participants; 30 mg/kg: RR 1.52, 95% CI 1.15 to 2.01, three trials, 521 participants; Analysis 2.1). Follow-up at three months ( Analysis 2.2) and at six to 12 months showed a similar pattern ( Analysis 2.3).
In one trial from Brazil evaluating 30 mg/kg versus 40 mg/kg, geometric mean egg reductions were high in both groups, at six months (92.5% versus 97.7%, statistical significance not reported (one trial, 138 participants, Analysis 2.4)).
One trial compared a lower dose of praziquantel at 20 mg/kg with 40 mg/kg and showed no difference in resolving symptoms at three, six, 12 and 24 months of follow-up: diarrhoea (one trial, 44 participants, Analysis 3.3), blood in stool (one trial, 37 participants, Analysis 3.5), hepatomegaly (one trial, 55 participants, Analysis 3.7) and splenomegaly (one trial, 73 participants, Analysis 3.9), except one study that showed that 40 mg/kg significantly improved abdominal pain at one month (RR 0.59, 95% CI 0.36 to 0.98, one trial, 169 participants, Analysis 3.1).
Two trials compared 30 mg/kg with 40 mg/kg and did not show any difference in resolving symptoms at one, three, six, 12 and 24 months of follow-up: abdominal pain (two trials, 318 participants, Analysis 3.2), diarrhoea (two trials, 48 participants, Analysis 3.4), blood in stool (two trials, 82 participants, Analysis 3.6), hepatomegaly (two trials, 109 participants, Analysis 3.8) and splenomegaly (two trials, 122 participants, Analysis 3.10).
In the three trials reporting adverse events, consistent differences in frequency or severity between 20, 30 and 40 mg/kg doses have not been shown (three trials, 319 participants, Table 3).
Analysis 4: Higher doses praziquantel versus 40 mg/kg
Higher doses (50 mg/kg to 60 mg/kg) have been evaluated in Brazil (three trials), Mauritania, Senegal and Tanzania. Compared to 40 mg/kg, parasitological failure has not been shown to be improved with higher doses at one month (five trials, 783 participants, Analysis 4.1).
Among participants still excreting eggs, percentage egg reductions were similar in both groups at one month (four trials, 786 participants, Analysis 4.4).
One multi-country trial reported adverse events and recorded one serious event (a seizure) with the higher dose. At the trial site in Brazil, non-severe adverse events appeared to be more common with the higher dose but this was not seen consistently at the trial sites in Mauritania or Tanzania (one trial, 653 participants, see Table 4).
Analysis 5: Split dose praziquantel versus 40 mg/kg in a single dose
Splitting 40 mg/kg into divided doses given on the same day was evaluated in the 1980s in three trials in Sudan.
At one month, two trials did not demonstrate a statistically significant benefit with the split dose regimen compared to a single 40 mg/kg dose (two trials, 525 participants, Analysis 5.1), but showed benefit at three months (RR 0.31, 95% CI 0.18 to 0.53, two trials, 516 participants, Analysis 5.2).
One further small trial, only reported the outcome at six months and found no difference (one trial, 64 participants, Analysis 5.3).
In the only trial reporting egg count, the mean percent reduction at one month was higher with the divided dose but statistical significance was not reported (divided dose 93.2% versus single dose 86.5%, one trial, 350 participants, Analysis 5.4).
No serious adverse events were reported in these trials. Only one trial reported the frequency of adverse events in each treatment group (Kardaman 1983 SDN). Mild abdominal pain and diarrhoea were less common when the dose was given in divided doses but vomiting was more common (one trial, 350 participants, Table 5).
Analysis 6: Other praziquantel dosing regimens
Several trials from Brazil have evaluated higher praziquantel dosing regimens with 30 mg/kg to 60 mg/kg given for up to six days (see Analysis 6.1). It is difficult to draw conclusions from these studies as the comparator dose is also a non-standard regimen, but one trial did demonstrate improved parasitological cure rates with prolonged courses given over three to six days compared to courses lasting one day.
No serious adverse events were reported in these trials, events were mainly transient dizziness and nausea (one trial, 79 participants, Table 6).
Seventeen trials evaluated oxamniquine, with the most recent conducted in the 1980s. Oxamniquine has since fallen out of use in favour of praziquantel. Four trials compared oxamniquine with placebo and 12 trials directly compared different dosing schedules of oxamniquine in different geographical locations in Africa and Brazil. The most common comparator dose was 40 mg/kg.
Analysis 7: Oxamniquine versus placebo
In two trials in Brazil, 20 mg/kg was significantly superior to placebo at longer timepoints (RR 3.68, 95% CI 2.53 to 5.36, two trials, 146 participants, Analysis 7.2). In two trials from Ethiopia, oxamniquine achieved parasitological cure rates of > 75% with 30, 40, and 60 mg/kg at three to four months, compared to placebo where almost all participants continued to excrete eggs (30 mg/kg: RR 4.34, 95% CI 2.47 to 7.65, two trials, 82 participants; 40 mg/kg: RR 8.74, 95% CI 3.74 to 20.43, two trials, 82 participants; 60 mg/kg: RR 19.38, 95% CI 5.79 to 64.79, two trials, 89 participants; Analysis 7.1).
Among those still excreting eggs at three to four months, two trials from Ethiopia reported significant reductions in egg numbers in those given oxamniquine (68.1% to 100%), compared to increases of 59 to 80.6% in the placebo groups (two trials, 227 participants, Analysis 7.3).
No serious adverse events were reported in these trials. Dizziness was more commonly reported with oxamniquine than placebo but is described as transient, with most resolving within 24 hours (five trials, 425 participants, Table 7).
Analyses 8 and 9: Lower doses oxamniquine versus 40 mg/kg
Lower doses of oxamniquine (20 to 30 mg/kg) have been compared to 40 mg/kg in Ethiopia (two trials), Sudan (two trials), Zimbabwe (two trials), Burundi and Malawi.
Compared to 40 mg/kg, both 20 mg/kg and 30 mg/kg of oxamniquine resulted in significantly more parasitological failures at one month (20 mg/kg: RR 3.78, 95% CI 2.05 to 6.99, two trials, 190 participants; 30 mg/kg: RR 1.78, 95% CI 1.15 to 2.75, four trials, 268 participants, Analysis 8.1), and at three to four months (20 mg/kg: RR 2.28, 95% CI 1.40 to 3.71, three trials, 209 participants; 30 mg/kg: RR 1.64, 95% CI 1.10 to 2.43, seven trials, 373 participants, Analysis 8.2).
At later time points, no statistically significant differences were shown: six months (20 mg/kg: two trials, 163 participants; 30 mg/kg: three trials, 214 participants, Analysis 8.3) and 12 months (20 mg/kg: two trials, 144 participants; 30 mg/kg: one trial, 77 participants, Analysis 8.4).
Percent egg reduction was evaluated in six of these trials and both lower dose and 40 mg/kg showed a wide range of benefit at one, three and six months: lower dose (57.1% to 99%) and 40 mg/kg (42.7 to 100%) (six trials, 878 participants, Analysis 8.5).
One trial compared a lower dose of 20 mg/kg oxamniquine with 40 mg/kg and did not find any difference between the two doses in resolving symptoms at one, three, six, 12 and 24 months of follow-up: abdominal pain (one trial, 95 participants, Analysis 9.1), diarrhoea (one trial, 16 participants, Analysis 9.3), blood in stool (one trial, 85 participants, Analysis 9.5), hepatomegaly (one trial, 64 participants, Analysis 9.7) and splenomegaly (one trial, 69 participants, Analysis 9.9).
Also, 30 mg/kg did not show any difference statistically compared with 40 mg/kg in resolving symptoms at one, three, six, 12 and 24 months of follow-up: abdominal pain (one trial, 95 participants, Analysis 9.2), diarrhoea (one trial, 15 participants, Analysis 9.4), blood in stool (one trial, 41 participants, Analysis 9.6), hepatomegaly (one trial, 51 participants, Analysis 9.8) and splenomegaly (one trial, 54 participants, Analysis 9.10).
Six trials from Ethiopia (two trials), and one trial each from Malawi, Sudan, Zambia and Zimbabwe assessed adverse events and reported no serious events. Dizziness was most commonly reported, but the event rate and severity did not differ between doses (six trials, 508 participants, Table 8).
Analysis 10: Higher doses oxamniquine versus 40 mg/kg
Higher doses of oxamniquine (50 mg/kg to 60 mg/kg) have been compared to 40 mg/kg in six trials from three countries; Sudan (three trials), Ethiopia (two trials) and Zambia (one trial).
Higher doses of oxamniquine have not shown consistent statistically significant benefits over 40 mg/kg at one month (five trials, 349 participants, Analysis 10.1), at three to four months (six trials, 397 participants, Analysis 10.2), or six months (two trials, 177 participants, Analysis 10.3).
Losses to follow-up were high in the trial investigating 50 mg/kg, reaching 76.9% at three months, and heterogeneity between the trials was significant (I
Seven trials evaluated egg count and reported a wide range of percent mean reductions among those not cured at one month (86% to 100% versus 56% to 99.1%, four trials, 561 participants, Analysis 10.4), three to four months (82% to 100% versus 42% to 100%, six trials, 791 participants, Analysis 10.4) and six months (62.% to 100% versus 75% to 100%, four trials, 561 participants, Analysis 10.4).
In five trials reporting adverse events, no serious events were recorded. Dizziness and nausea were most commonly reported, but these were transient and did not require additional interventions (one trial, 482 participants, Table 9).
Analyses 11 and 12: Other oxamniquine dosing regimes
Nine additional trials compared 30 mg/kg oxamniquine with higher and lower doses in Ethiopia (three trials), Zimbabwe (two trials), Burundi (one trial), Nigeria (one trial), Sudan (one trial) and Zambia (one trial).
Lower doses versus 30 mg/kg
Compared to 30 mg/kg, parasitological failure was higher with 15 mg/kg to 20 mg/kg oxamniquine at one month (RR 1.77, 95% CI 1.14 to 2.74, two trials, 230 participants), and at three to four months (RR 2.16, 95% CI 1.40 to 3.32, four trials, 249 participants, Analysis 11.1).
At later follow-up times, no statistically significant difference were demonstrated (six months: two trials, 179 participants; and 12 months: one trial, 95 participants, Analysis 11.1).
Higher doses versus 30 mg/kg
Compared to 30 mg/kg, 60 mg/kg oxamniquine resulted in significantly fewer parasitological failures at one month (RR 0.04, 95% CI 0.01 to 0.26, two trials, 175 participants, Analysis 12.1), at three to four months (RR 0.17, 95% CI 0.07 to 0.39, four trials, 265 participants, Analysis 12.2) and at six months (RR 0.17, 95% CI 0.06 to 0.50, two trials, 157 participants, Analysis 12.3).
No statistically significant differences were seen between 50 mg/kg and 30 mg/kg at one month (one trial, 36 participants, Analysis 12.1) or at three to four months (two trials, 53 participants, Analysis 12.2).
Analysis 13: Praziquantel (40 mg/kg) versus oxamniquine
Eleven trials from different geographical locations directly compared various doses of oxamniquine with praziquantel 40 mg/kg. Dosing schedules commonly applied across different locations are reported in Table 10. The most recent trial, from Sudan, was published in 1990.
We did not identify statistically significant differences between oxamniquine (at doses from 10 mg/kg to 60 mg/kg) and praziquantel 40 mg/kg at one month (see Analysis 13.1). No difference was demonstrable at three months between 25 to 30 mg/kg (three trials, 319 participants), 40 mg/kg (one trial, 18 participants) or 50 to 60 mg/kg (one trial, 14 participants, Analysis 13.2). However, 10 to 20 mg/kg of oxamniquine did result in significantly more failures (RR 3.42, 95% CI 1.10 to 10.61, two trials, 135 participants, Analysis 13.2).
In addition, there were no differences between oxamniquine (lower or higher dose) and praziquantel (40 mg/kg) at six months (nine trials, 1167 participants, Analysis 13.3) or 12 months (one trial, 52 participants, Analysis 13.4).
Three trials from Brazil, Ethiopia and Malawi compared oxamniquine 15, 20, 30, 40, and 50 mg/kg with praziquantel 40 mg/kg and measured high percent egg reduction at one month (82.9% to 100% for oxamniquine versus 90% to 92.8% for praziquantel, two trials, 391 participants), three months (70.2% to 99.5% for oxamniquine versus 70% to 100% for praziquantel, three trials, 440 participants), six months (32.5% to 97% for oxamniquine versus 33.6% to 96.8% for praziquantel, three trials, 291 participants), and 12 months (94% for oxamniquine versus 96% for praziquantel, one trial, 91 participants, Analysis 13.5).
In five trials reporting from Brazil, Ethiopia, Malawi, Saudi Arabia and Tanzania that assessed adverse events, only two serious adverse events were recorded (both with oxamniquine) in two trials: one from a moderate endemicity setting in Ethiopia that used 30 mg/kg in a split dose given the same day; and one trial from Saudi Arabia that used a single dose of 25 mg/kg. No further differences were observed in the number and type of adverse events between oxamniquine and praziquantel although dizziness was recorded in excess with oxamniquine and abdominal pain with praziquantel ( Table 11).
Analysis 14: Myrrh (Mirazid) versus praziquantel
Myrhh (Mirazid) was tested in two trials at a single daily dose of 300 mg for three days, and almost all failed treatment at three to six weeks (RR 4.08, 95% CI 2.87 to 5.78, 236 participants, Analysis 14.1). Consequently, further investigation of this compound was abandoned.
Egg reduction rate
There were only small reductions in reported percent geometric mean egg reduction in these two studies, but they were not clinically important ( Analysis 14.2).
No trial reports adverse events.
Section 2. Combination therapies
Analysis 15: Praziquantel plus artesunate versus praziquantel alone
One trial conducted from 1999 to 2000 in a high endemicity setting in Senegal evaluated artesunate plus praziquantel versus praziquantel alone.
In this setting, parasitological failure at one month occurred in 50% of participants given praziquantel 40 mg/kg alone. The addition of artesunate 12 mg/kg given in a divided dose of 2.5 mg/kg daily for five days resulted in a lower failure rate at one month but this did not reach statistical significance (one trial, 75 participants, Analysis 15.1). At three and six months no additional benefit with artesunate plus praziquantel was seen.
Geometric mean egg reductions appear lower with combination treatment but tests of statistical significance were not reported, and the clinical relevance of this finding are unclear (one trial, 75 participants, Analysis 15.4).
Adverse events were not reported.
Analysis 16: Praziquantel plus oxamniquine versus praziquantel alone
Only one trial in a high endemicity setting in Brazil published in 1987 has evaluated oxamniquine plus praziquantel versus praziquantel alone.
Compared to praziquantel alone (40 mg/kg in two divided doses on one day), a combination of oxamniquine (7.5 mg/kg) plus praziquantel (20 mg/kg) did not demonstrate any statistically significant benefits at three, six or 12 months follow-up (one trial, 52 participants, Analysis 16.1).
The combination treatment was associated with lower geometric mean egg reductions at three, six and 12 months but tests of statistical significance were not reported (one trial, 52 participants, Analysis 16.4).
These were not reported.
Analysis 17: Praziquantel (8 mg/kg) plus oxamniquine (4 mg/kg) versus praziquantel (20 mg/kg) plus oxamniquine (10 mg/kg)
One small trial of schoolchildren from a high endemicity setting co-endemic for S. mansoni and S. haematobium in Zimbabwe investigated different oxamniquine and praziquantel dose combinations.
Children aged seven to 16 years and excreting more than 100 eggs per gram of stool were included in this trial. Statistically fewer failures were seen with the higher dose-combination at one month (RR 6.30, 95% CI 1.60 to 24.75, one trial, 28 participants, Analysis 17.1), but not at three months (one trial, 29 participants, Analysis 17.2) or six months (one trial, 20 participants, Analysis 17.3).
The percentage egg reduction also appeared to be lower in those receiving the higher dose combination but tests of statistical significance were not reported (one trial, 59 participants, Analysis 17.4).
No serious adverse events were recorded and the incidence of non-severe events did not differ between combinations. About 70% of children reported abdominal discomfort but these were transient and had resolved by the following day ( Table 12).
Analysis 18: Praziquantel (15 mg/kg) plus oxamniquine (7.5 mg/kg) versus praziquantel (20 mg/kg) plus oxamniquine (10 mg/kg)
One trial in Zimbabwe investigated slightly higher oxamniquine and praziquantel dose combinations. The included children had to excrete more than 100 eggs per gram of stool.
Egg reduction rate
Percent egg reductions were high at one, three and six months (82% to 96.1% versus 66.3% to 96.6%, one trial, 59 participants, Analysis 18.4).
No serious adverse events were recorded apart from one child who reported dizziness immediately after treatment but required no further treatment ( Table 12).
Section 3. Do failure rates vary in children and adults?
A subgroup analysis conducted in two studies from Burundi raised concern that parasitological failure following 40 mg/kg may be higher in children than in adults. The frequency of parasitological treatment failure was consistently higher in children than adults at one, three, six, and 12 months, and this was also observed for doses of 20 mg/kg and 30 mg/kg (see Table 13).
Similarly, a subgroup analysis of two studies from Burundi and Sudan administering oxamniquine has shown a consistent pattern of higher parasitological treatment failure in children than adults at one to 12 months (see Table 14).
Subgroup analysis of treatment arms receiving 40 mg/kg in the other included studies was not possible given the available data.
Summary of main results
Compared to placebo, praziquantel 40 mg/kg substantially reduced parasitological treatment failure at one month post-treatment (moderate quality evidence). Compared to this standard dose, lower doses of 20 mg/kg and 30 mg/kg were inferior (low quality evidence); and higher doses, up to 60 mg/kg, have not shown any advantage (moderate quality evidence).
Compared to placebo, oxamniquine 40 mg/kg substantially reduced parasitological treatment failure at three months (moderate quality evidence). Lower doses than 40 mg/kg were inferior at one month (low quality evidence), and higher doses such as 60 mg/kg have not shown a consistent benefit (low quality evidence).
Ten trials compared oxamniquine at 20, 30 and 60 mg/kg with praziquantel 40 mg/kg and did not show any convincing differences in failure rate and percent egg reduction. Only one small study directly compared praziquantel 40 mg/kg with oxamniquine 40 mg/kg and did not demonstrate a statistically significant difference in parasitological failure (very low quality evidence).
Combining praziquantel with artesunate has not been shown to have benefits in terms of failure rate compared to praziquantel alone at one month, three or six months (one trial, 75 participants, very low quality evidence). Two trials have also compared combinations of praziquantel and oxamniquine in different doses but did not find statistically significant differences in failure rate.
Compared to 40 mg/kg, no dose effect was demonstrable for clinical improvement with lower doses (20 and 30 mg/kg) of praziquantel or oxamniquine in resolving abdominal pain, diarrhoea, blood in stool, hepatomegaly, and splenomegaly at one, three, six, and 12 months, or up to two years of follow-up. Adverse events were not well reported but were mostly described as minor and transient.
Overall completeness and applicability of evidence
For praziquantel, the evidence presented is generally supportive of the current WHO recommended dose of 40 mg/kg to treat S. mansoni infection (WHO 2002). Parasitological cure as low as 57% has been reported in Kenya in the 1990s (Olds 1999 KEN), and 52% in Senegal in 1993 (Guisse 1997 SEN). However, higher efficacy has been seen in more recent trials; Tanzania (81%), Mauritania (95%) and Brazil (92%) in 2006/2007 (Olliaro 2011 BRA; Olliaro 2011 MRT; Olliaro 2011 TZA), and Uganda (87%) in 2003/2005 (Tweyongyere 2009 UGA). The lower cure rates from the earlier studies could be expected from the high endemicities where pre-treatment intensity of infection were very high (prevalence > 80%) compared to the recent studies (prevalence < 30%). In such situations, even at 95% efficacy, a sufficient number of surviving schistosomes would remain, causing sustained egg excretion in most of the treated participants (Danso-Appiah 2002). Furthermore, as a result of intense transmission, most treated participants might have acquired large numbers of new infections just before treatment and as immature worms are less sensitive to praziquantel most would have escaped drug action and developed into egg-laying adult worms shortly after treatment to present as failures. The high diagnostic sensitivity (mostly duplicate slides from two or more consecutive stool specimens) and lower dose of praziquantel applied in the earlier studies (except Guisse 1997 SEN) would have also contributed to the observed lower cure rates.
The results in this review appear to be generalizable elsewhere but it should be noted that these trials excluded preschool children under five years and concerns remain that this dose may be less effective in this group. This is because praziquantel works in synergy with host immune status (Sabah 1986) and this is not yet fully developed in very young children. A subgroup analysis conducted in two studies from Burundi with praziquantel at 40 mg/kg and another two studies from Burundi and Sudan with oxamniquine 40 mg/kg raises concern as parasitological failure was consistently higher in children than in adults at one to 12 months of follow-up. This trend was also observed for doses of 20 mg/kg and 30 mg/kg for both treatments, and a higher dose (60 mg/kg) for oxamniquine.
Higher doses than 40 mg/kg have been national policy in Brazil since 1995: 60 mg/kg for children and 50 mg/kg for adolescents and adults. We found little direct evidence from randomized controlled trials to support or refute this as a policy. Only a single trial from Brazil reported outcomes at one month and this failed to show a statistically significant advantage with 60 mg/kg compared to 40 mg/kg, and excluded children aged less than 10 years (Olliaro 2011 BRA). Several further trials from Brazil have evaluated higher doses and longer regimens but these only reported outcomes at six months or beyond. These do offer some limited evidence that increasing the dose of praziquantel might have parasitological benefits.
There is no justification for using lower doses, even if potentially effective in morbidity control, as sub-curative doses may eventually select for drug resistant parasites (Doenhoff 1998; Doenhoff 2008).
Praziquantel is known to be less effective on immature schistosomes than adult worms (Sabah 1986), and combination therapy (with drugs with unrelated mechanisms of action and targeting the different developmental stages of the schistosomes), has potential as a future control strategy. Potential partner drugs include oxamniquine and the artemisinin derivatives. Of these, the artemisinin derivatives have been shown to be effective against immature schistosomes in laboratory studies (Utzinger 2001; Utzinger 2002; Utzinger 2003; Utzinger 2007), and there is some indirect evidence for efficacy from non-randomized studies in urinary schistosomiasis (De Clercq 2002; Inyang-Etoh 2004; Boulanger 2007; Inyang-Etoh 2009), and from people with malaria co-infected S. haematobium (Boulanger 2007). However, to date only a single trial has directly evaluated praziquantel plus artesunate and no additional benefit was observed compared to praziquantel alone (De Clercq 2000 SEN).
For oxamniquine, there is no current consensus on the optimal dosing regimen and it has largely fallen out of use in favour of praziquantel. Although the presented data are now more than 20 years old, and suffers some methodological problems, there is sufficient evidence of its efficacy against S. mansoni to suggest that it could be reinstated as an alternate treatment to decrease the pressure on praziquantel. However, a limitation of oxamniquine is that its effect is restricted to S. mansoni as this is the only species possessing the enzyme which converts oxamniquine to its active metabolite (Cioli 1995). It is therefore unsuitable for use in areas where co-infection with S. haematobium is common.
The optimal dose of oxamniquine may also be 40 mg/kg but further studies are required to confirm this, preferably in direct comparison with praziquantel, and trials should include and evaluate the efficacy of this dose in young children.
Safety was under reported and inconsistently assessed in most of these clinical trials. Furthermore, only the few studies comparing the intervention versus placebo allow identification of potentially drug-related events. From these few studies it is therefore not possible to provide a reliable account of treatment tolerability.
Quality of the evidence
The quality of evidence was assessed using the GRADE methodology and displayed in summary of findings (SOF) tables for the main comparisons. The level of quality is judged on a 4-point scale. High quality evidence implies that level of confidence in the effect estimate is high and that further research is unnecessary. Moderate quality evidence implies lower confidence in the result and further research may have an important impact on the result. Low and very low quality evidence reflect increasing uncertainty in the result and a greater need for further research.
The evidence presented is generally considered to be of moderate or low quality due to concerns related to three key factors: i) the age of the trials, with the majority more than 20 years old, ii) the poor methodological reporting of many of these older trials, and iii) the number and size of the trials being small and often underpowered to reliably detect statistically significant differences. The specific reasons for downgrading the quality of the evidence are given in the footnotes to the SOF tables.
Potential biases in the review process
A few minor difficulties in extracting the data from the available papers should be noted but these are unlikely to have introduced major bias into this review. For three trials (Gryseels 1989a BDI; Gryseels 1989b BDI; Gryseels 1989c BDI), data on parasitological failure were obtained from figures and might not be the exact estimates. One trial (Sukwa 1993 ZMB) actually reported reinfection rate but this is included in this review because this outcome is similar to failure rate. The trial by Tweyongyere 2009 UGA was a nested cohort study within a larger mother and baby cohort study in which pregnant women found to be infected with S. mansoni were randomized to receive praziquantel or placebo. Despite representing a special population, this is not likely to affect the validity of the results.
Agreements and disagreements with other studies or reviews
A non-Cochrane review compared praziquantel with placebo in two studies in Brazil and showed slightly higher cure rate with praziquantel (Liu 2011). The reliability of the evidence in this review cannot be established given that the two studies that assessed this outcome involved only 25 participants.
The effects of praziquantel and artesunate in urinary schistosomiasis due to S. haematobium have been evaluated in a separate Cochrane review last published in 2008. Praziquantel was found to be effective against S. haematobium with few adverse events, and similarly to this review there was insufficient evidence for the use of artesunate monotherapy or combination therapy (Danso-Appiah 2008).
Limitations in the design and methodology in schistosomiasis trials identified during the earlier Cochrane review, and consequent future research needs have also been reported elsewhere (Danso-Appiah 2009).
Implications for practice
The available evidence supports single dose praziquantel at 40 mg/kg as the standard treatment for S. mansoni infection as recommended by the WHO.
Oxamniquine, a largely discarded alternative, appears efficacious and production and distribution should continue to ease selective pressure on praziquantel. However, its use should be limited to areas without S. haematobium co-endemicity.
Implications for research
Further research is necessary to find the optimal dosing regimen of praziquantel and oxamniquine in children under five years, given the observational evidence that failure rates with 40 mg/kg may be higher in this age-group.
Combination therapy, ideally with drugs with unrelated mechanisms of action and targeting the different developmental stages of the schistosomes in the human host should be pursued as an area for future research; for example; praziquantel plus oxamniquine, praziquantel plus mefloquine, and praziquantel plus an artemisinin derivative.
We thank Dr. Vittoria Lujte (Information Specialist) for developing the search strategy and doing the search for studies and Christianne Esparza for retrieving hard copies of published studies. We are grateful to Dr. Harriet MacLehose (former Deputy Editor), Anne-Marie Stephani (Managing Editor), Phil Hinds (Editorial Assistant & Administrator), Reive Robbs (former CIDG Co-ordinator) and the entire International Health Group for their support during the preparation of this review. Our sincere thanks go to Dr Otavio Pieri and Prof Martin Meremikwu for critically reading this review and for providing useful comments, and to Dr Deirdre Walshe (Associate Editor) for drafting the plain language summary.
This document is funded by the UK Department for International Development (DFID) for the benefit of developing countries. . The views expressed are not necessarily those of DFID. PLO is a staff member of the WHO; the author alone is responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy, or views of WHO.
H. Saconato and A. Atallah prepared the original version of this review (Issue 3, 1999).
Data and analyses
- Top of page
- Summary of findings [Explanations]
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Last assessed as up-to-date: 16 October 2012.
Contributions of authors
ADA, JU and PLO developed the protocol. ADA selected studies, extracted data, assessed risk of bias in the included studies, analysed the data and drafted the review. JU independently verified study selection, data extraction, risk of bias assessment, results of the analysis and edited the draft review. PLO verified study selection, risk of bias assessment and edited the draft review. SD provided statistical advice and edited the methods section. DS helped restructure the review, verified risk of bias assessment and prepared the SOF tables, which were checked by ADA. ADA, JU and PLO interpreted the data, and all authors helped with revisions following the referees' comments.
Declarations of interest
PLO was the lead author in three of the included trials (Olliaro 2011 BRA; Olliaro 2011 MRT; Olliaro 2011 TZA) and helped secure additional financial support from WHO. The rest of the authors have no known conflict of interest.
Sources of support
- Liverpool School of Tropical Medicine, UK.
- Department for International Development, UK.
- The World Health Organization, Switzerland.
- Swiss National Science Foundation (Project no PPOOB-102883, PPOOB-119129), Switzerland.
Differences between protocol and review
Medical Subject Headings (MeSH)
Oxamniquine [*therapeutic use]; Plant Extracts [therapeutic use]; Praziquantel [*therapeutic use]; Randomized Controlled Trials as Topic; Schistosomiasis mansoni [*drug therapy]; Schistosomicides [*therapeutic use]
MeSH check words
Adolescent; Adult; Child; Humans
* Indicates the major publication for the study