Rotavirus vaccine for preventing diarrhoea

Karla Soares‐Weiser; Elad Goldberg; Ghandi Tamimi; Leonard Leibovici; Femi Pitan

doi:10.1002/14651858.CD002848.pub2

. 2004 Jan 26;2004(1):CD002848. doi: 10.1002/14651858.CD002848.pub2

Rotavirus vaccine for preventing diarrhoea

Karla Soares‐Weiser ^1,^✉, Elad Goldberg ², Ghandi Tamimi ³, Leonard Leibovici ², Femi Pitan ⁴

Editor: Cochrane Infectious Diseases Group

PMCID: PMC6532746 PMID: 14973994

Abstract

Background

Rotaviruses cause viral gastroenteritis and result in more deaths from diarrhoea in children under 5 years of age than any other single agent, particularly in low‐ and middle‐income countries.

Objectives

To assess rotavirus vaccines in relation to preventing rotavirus diarrhoea, death, and adverse events.

Search methods

We searched the Cochrane Infectious Diseases Group's trial register (October 2003), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 3, 2003), MEDLINE (1966 to October 2003), EMBASE (January 1980 to October 2003), LILACS (1982 to October 2003), Biological Abstracts (January 1982 to October 2003), reference lists of articles, and contacted researchers and rotavirus vaccine manufacturers.

Selection criteria

Randomized controlled trials comparing rotavirus vaccines to placebo, no intervention, or other rotavirus vaccines in children and adults.

Data collection and analysis

Two reviewers independently extracted data and assessed trial methodological quality, and contacted trial authors for additional information.

Main results

Sixty‐four trials provided information on efficacy and safety of three main types of rotavirus vaccine (bovine, human, and rhesus) for 21,070 children. Different levels of efficacy were demonstrated with different vaccines varying from 22 to 89% to prevent one episode of rotavirus diarrhoea, 11 to 44% to prevent one episode of all‐cause diarrhoea, and 43 to 90% to prevent one episode of severe rotavirus diarrhoea. Rhesus vaccine demonstrated a similar efficacy against one episode of rotavirus diarrhoea (37 and 44% respectively), and one episode of all‐cause diarrhoea (around 15%) for trials performed in high and middle‐income countries. Results on mortality and safety of the vaccines were scarce and incomplete. We noticed important heterogeneity among the pooled studies and were unable to discard a biased estimation of effect.

Authors' conclusions

Current evidence shows that rhesus rotavirus vaccines (particularly RRV‐TV) and the human rotavirus vaccine 89‐12 are efficacious in preventing diarrhoea caused by rotavirus and all‐cause diarrhoea. Evidence about safety, and about mortality or prevention of severe outcomes, is scarce and inconclusive. Bovine rotavirus vaccines were also efficacious, but safety data are not available. Trials of new rotavirus vaccines will hopefully improve the evidence base. Randomized controlled trials should be performed simultaneously in high‐, middle‐, and low‐income countries.

2010 Editor's Note: Several of the vaccines investigated in this review are no longer in routine clinical use (for example, live attenuated rhesus‐human reassortant tetravalent vaccine) and further new vaccines have been tested and approved for use since this review was written in 2004. For an up‐to‐date assessment of rotavirus vaccines currently approved for use, please see : Soares‐Weiser K, MacLehose H, Ben‐Aharon I, Goldberg E, Pitan F, Cunliffe N. Vaccines for preventing rotavirus diarrhoea: vaccines in use. Cochrane Database of Systematic Reviews 2010, Issue 5. Art. No.: CD008521. DOI: 10.1002/14651858.CD008521.

23 April 2019

No update planned

Review superseded

This Cochrane Review has been superseded by Soares‐Weiser 2019 https://doi.org/10.1002/14651858.CD008521.pub4

Plain language summary

Rotavirus vaccines can prevent diarrhoea caused by rotavirus, but we are still not clear about safety and whether they prevent deaths

Rotavirus diarrhoea causes illness and death in young children. The benefits of the vaccine were different depending on the type of vaccine. The reviewers are unable to make conclusive recommendations regarding the use of rotavirus vaccines.

Background

Rotaviruses cause viral gastroenteritis, and result in more deaths from diarrhoea than any other single agent (Vesikari 1997). In middle‐ and low‐income countries, rotavirus kills an estimated 1600 to 2400 children every day (600,000 to 870,000 per year), and can cause up to 6% of all mortality in children under 5 years of age. It is estimated that rotavirus is responsible for 20 to 25% of all deaths due to diarrhoeal disease worldwide (Bresee 1999; Glass 1999).

Biology

People become infected with rotavirus by direct faecal‐oral spread, and symptoms develop 1 to 4 days after exposure. Babies in the first 3 months tend to have infections without symptoms because of the protection of antibodies transferred to the baby via the placenta and breastfeeding. In older infants, the disease can be without symptoms, or with mild or severe symptoms, which are characterized by vomiting, fever, watery diarrhoea, and dehydration (AAP 1998). Infants and children develop immunity to rotavirus infection during the first 2 or 3 years of life, which is thought to protect from subsequent episodes of severe disease (Vesikari 1997).

Rotaviruses have two proteins that induce neutralizing antibodies. These proteins have two different specific antibody‐inducing sections, called the P and G types (P is short for the VP4 protease‐cleaved protein, and G for the glycoprotein VP7) (Matson 1990). Children that are infected for the first time usually develop specific antibodies to the infecting rotavirus type, but with subsequent infections develop a broader immunity to other types.

Control

Efforts to prevent disease by improving sanitary conditions have not reduced exposure (Glass 1999; Peter 2002). More than 90% of children, in any region of the world, have been exposed and have measurable rotavirus antibody titres by the ages of 3 to 5 years. Hence, the high priority placed on the development of a safe and effective vaccine (Peter 2002).

Vaccines

Several rotavirus vaccines have been developed (Appendix 1) (Vesikari 1997; Bresee 1999). Early vaccine candidates were developed from an animal virus strain, but the efficacy of such vaccines was highly variable due to a predominant homotypic response of these so called monovalent vaccines. In an effort to induce a heterotypic response (tetravalent vaccines), combinations of animal or human‐animal reassortant strains or attenuated human strains of rotaviruses have been included in the second generation of vaccines (Henchal 1996).

The first vaccine tested in the early 1980s was a bovine, monovalent, live‐oral rotavirus vaccine for children (NCDV ‐ RIT 4237) (Appendix 1). The first trials were conducted in Finland (Vesikari 1997), and the vaccine seemed effective both against mild and severe rotavirus infection (Vesikari 1997). After this initial success, the monovalent, live‐oral bovine vaccines (RIT 4237 and WC3) were tested in low‐ and middle‐income countries, and much lower efficacy was reported. The hypothesis generated for such lower efficacy was that in these countries the rotavirus diarrhoea was caused by a different strain, not available in the vaccine. Both bovine rotavirus vaccines were therefore withdrawn and have not been used since then (Vesikari 1997).

Monovalent rhesus (RRV) and human attenuated (M37) vaccines were tested next. These vaccines appeared to be promising, but efficacy varied greatly between different countries. This variability was suggested to occur due to serotype‐specific protection produced by the monovalent vaccine. This finding was demonstrated in controlled trials conducted in Venezuela in the early 1990s (Perez‐Schael 1990b), and led to the development of a polyvalent reassortant vaccine protecting against the four major serotypes of rotavirus that were in common circulation worldwide.

To broaden the antigenic spectrum of the vaccine, rhesus‐human reassortant rotaviruses were developed. These polyvalent vaccines provide serotype‐specific immunity against all four predominant human rotavirus serotypes (G1, G2, G3, and G4), and were described as efficacious in reducing the rate of severe rotavirus diarrhoea by 80% ‐ a rate similar to the protection conferred by natural infection (Vesikari 1997).

In August 1998, the rhesus rotavirus tetravalent vaccine (RRV‐TV, RotaShield, Wyeth‐Lederle Vaccines and Pediatrics, Philadelphia, PA) was licensed by the US Food and Drug Administration for oral administration to infants at 2, 4, and 6 months of age. This vaccine is a live‐attenuated, orally administered product derived from four group A rotavirus. Three of the rotaviruses are single gene reassortants of the VP7 gene of human origin (types G1, G2, and G4), and the fourth strain is rhesus rotavirus (type G3), which is antigenically similar to human G3. No data are available to indicate whether this rotavirus vaccine protects against diarrhoea attributable to rotavirus strains not contained in the vaccine (Vesikari 1997).

Several studies in the USA, Finland, and Venezuela have found that RRV‐TV is very effective in rotavirus disease (Perez‐Schael 1997; Vesikari 1997; AAP 1998). However, other studies conducted in Peru and Brazil produced conflicting results that showed a much lower efficacy (Lanata 1996a; Linhares 1996).

Since the introduction of RRV‐TV, several cases of bowel obstruction and intussusception following administration of the vaccine have been reported, which have lead to the vaccine being withdrawn. In July 1999, the US Centers for Disease Control and Prevention advised doctors to temporarily stop giving children the vaccine against rotavirus after counting at least 20 infants who developed intussusception in the weeks after vaccination with RRV‐TV (CDC 1999; CDC 1999a; MMWR 1999a; MMWR 1999b). In October 1999, the vaccine's manufacturer decided to voluntarily withdraw the vaccine from the market because reports of intussusception possibly linked to the product have reached about 100 (MMWR 1999b).

RRV‐TV was also thought to cause febrile reactions 3 to 4 days after vaccination due to the viral antigens. These reactions seem to be more frequent and severe in older children who lack maternally acquired antibodies. Therefore the first dose of the vaccine was recommended to be given between age 2 and 6 months in the USA.

Using the same strategy as RRV‐TV, a quadrivalent, human‐bovine, reassortant vaccine has been tested. This vaccine incorporates human VP7 with bovine genes (G1 to G3) and VP4 reassortants. Results of efficacy trials with this vaccine are due to be published (Dr Penny Heaton, Vaccine Infectious Diseases ‐ Merck, personal communication).

Other vaccines have been developed (Bresee 1999; Merck 1999; Jones 2001). In Japan, an inactivated parenteral vaccine (BIRVI) with strains of human origin has been tested. In China, a live oral vaccine that uses a lamb strain appeared to be safe and immunogenic in phase II trials, and efficacy trials are now being planned. Vaccine‐like particles for use as an inactivated vaccine are being developed in Korea (Cunliffe 2002).

Sources of heterogeneity

The wide variation in protection observed in randomized controlled trials may be related to the response of the immune system to different strains of the rotavirus or of rotavirus vaccine, the study population, or the study design.

Rotavirus

Cross‐immunization is induced between some, but not all, human and animal strains. The differences in efficacy found in the various studies can be attributed in part to differences in serotypes of circulating strains and the failure of animal strains to elicit heterotypic protection.

Vaccines

Different vaccine formulations may have different efficacy. Premature conclusions about effectiveness were drawn from trials of bovine rotavirus vaccine in low‐ and middle‐income countries, and the same should not be repeated for other vaccines.

Study population

One possible explanation for low efficacy of rotavirus vaccine in low‐ and middle‐income countries may be the interference by other enteric viruses, which are a common cause of diarrhoea in children in such countries (Vesikari 1997). The epidemiology of rotavirus differs in low‐ and middle‐income compared to high‐income countries. In low‐ and middle‐income countries, the disease occurs all the year round (compared with seasonal peaks in high‐income countries), children are infected at an early age (6 to 9 months), immunization efforts require greater levels of coverage to be effective, and children are often subject to infections with strains that are not included in the current vaccines (Bresee 1999).

As a result, it has been speculated that rotavirus vaccine could be more efficacious if given at an earlier age, not associated with other vaccines (particularly 'diphtheria‐pertussis‐tetanus' (DPT) and oral polio vaccine (OPV)), or breastfeeding. Additionally, it has been suggested that higher titre and multiple doses of the vaccine are more efficacious (Vesikari 1997; Bresee 1999).

Study design

It is also plausible that some methodological issues, such as allocation concealment (adequate or unclear), exclusions after randomization (reported or not reported), sample size (< 1000, ≥ 1000), and length of follow up (one rotavirus season or more than one rotavirus season) could provide different levels of efficacy.

Significance of rotavirus vaccine

Finding that a rotavirus vaccine is efficacious might encourage immunization in the countries with the highest burden (Tucker 1998). This has to be weighed against the available resources (Griffiths 1995). Hence, to become a worthwhile intervention, vaccines against rotavirus will not only have to be proven to be efficacious and not related to major adverse events, but also affordable and cost‐effective in comparison with other potential interventions (Glass 1999).

Efficacy levels of as low as 50 to 60% against severe rotavirus disease could prevent a large number of deaths (Josefson 1997). Because rotavirus disease causes up to 6% of all mortality among children less than 5 years of age in low‐ and middle‐income countries, universal immunization with rotavirus vaccine may prevent a significant number of deaths, and the trade off against rare severe adverse effects may well be different from the situation in high‐income countries.

Objectives

To assess rotavirus vaccines in relation to preventing rotavirus diarrhoea, death, and adverse events.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled trials.

Types of participants

Children and adults.

Types of interventions

We included the following interventions and controls regardless of doses and/or schedules used in the different trials.

Intervention

Live‐attenuated bovine rotavirus vaccine (RIT‐4237, WC3, W179‐9, QHBRV).
Rhesus rotavirus vaccine (RRV, RRV‐TV).
Human‐attenuated rotavirus vaccine (M37, 89‐12).
Other rotavirus vaccines (BIRVI, Lamb RV).

Control

Placebo or no vaccination.
Any of the interventions mentioned above.

Types of outcome measures

Primary

Rotavirus diarrhoea

Episodes.
Severe episodes (defined as an episode that last for more than 24 hour of duration, with more than six stools in 24 hours, presence of vomiting, fever, and dehydration (WHO 1995)).
Episodes requiring hospitalization.
Episodes of more than four days duration.

Secondary

All‐cause diarrhoea

Episodes.
Episodes during first week after vaccine.
Severe episodes (defined as an episode that last for more than 24 hour of duration, with more than six stools in 24 hours, presence of vomiting, fever, and dehydration (WHO 1995)).
Episodes requiring re‐hydration.
Episodes requiring hospitalization.
Death: all‐cause.
Death: from rotavirus infection.
Malnutrition at follow up.
Drop outs before the end of the study.

Adverse events

Bowel obstruction and/or intussusception related to the use of the rotavirus vaccine.
Serious adverse events that are fatal, life threatening, or require hospitalization.
Systemic reaction, such as fever, related to the use of the rotavirus vaccine.
Adverse events that require discontinuation of vaccination schedule.

We excluded studies collecting only immunological data.

Search methods for identification of studies

We attempted to identify all relevant studies regardless of language or publication status (published, unpublished, in press, and in progress).

We searched the Cochrane Infectious Diseases Group's trials register for relevant trials up to October 2003 using the search terms: rotavirus; diarrhoea; diarrhea; gastroenteritis; and vaccine. Full details of the Cochrane Infectious Diseases Group methods and the journals handsearched are published in The Cochrane Library in the section on Collaborative Review Groups.

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library (Issue 3, 2003), using the search terms: rotavirus; diarrhoea; diarrhea; gastroenteritis; and vaccine. We also searched the following electronic databases using the search terms in combination with the search strategy developed by The Cochrane Collaboration (Clarke 2003); MEDLINE (1966 to October 2003) using the search terms: rotavirus; rotavirus‐infections; diarrhoea; diarrhea; gastroenteritis; vaccines‐attenuated; and viral‐vaccines; EMBASE (January 1980 to October 2003) using the search terms: rotavirus; diarrhoea; diarrhea; gastroenteritis; vaccines‐attenuated; and viral‐vaccines; and LILACS (www.bireme.br; 1982 to October 2003) using the search terms: rotavirus; diarrhoea; diarrhea; gastroenteritis; and vaccine. Biological Abstracts (January 1982 to October 2003) using the search terms: rotavirus vaccine; diarrhoea; and diarrhea.

We checked the reference lists of all trials identified by the above methods. We also contacted the first or corresponding author of each included study as well as researchers active in the field, and a rotavirus vaccine manufacturer (Merck Sharp & Dohme) for information regarding unpublished trials or complementary information on their own trial.

Data collection and analysis

Selection of studies

Two reviewers (Karla Soares‐Weiser (KSW), Elad Goldberg (EG)) independently inspected the abstract of each reference identified by the search and determined the potential relevance of each article. For potentially relevant articles, or in cases of disagreement, we obtained the full article and independently inspected it and applied the inclusion criteria. Where we were unable to resolve the disagreement through discussion, we added the article to those 'awaiting assessment' and contacted the authors of the study for clarification. In an event of no reply from the authors within 6 months, a third reviewer (Leonard Leibovici (LL)) checked the article to solve disagreements. We also documented our justification for excluding studies from the review.

Data extraction and management

Two reviewers (KSW, EG) independently extracted data from the included trials. In case of any disagreement between the two reviewers, a third reviewer (LL) extracted the data. We discussed the data extraction, documented decisions, and where necessary, contacted the study authors for clarification.

We identified trials by the name of the first author and year in which the trial was first published. We extracted, checked, and recorded the following data.

1. Characteristics of trials.

Date location and setting of trial.
Publication status.
Case definitions used (clinical, serological, virological).
Sponsor of trial (specified, known or unknown).

2. Characteristics of participants.

Number of participants in each group.
Age, gender, nationality, ethnic group, and risk category.
Previous immunization status (if known).
Use of breast‐bottle feed during the vaccination.
Presence of any immunodeficiency.

3. Characteristics of interventions.

Type of vaccine, type of control, dose, and immunization schedule.

4. Characteristics of outcome measures.

Number of people who died in the vaccine and control group.
Number of people who developed rotavirus diarrhoea.
Number of people who developed any diarrhoea.
Number of hospitalizations for diarrhoea, dehydration, or malnutrition.
Systemic adverse events, specifically bowel obstruction.
Any adverse events.
Length of follow up (in months).
Loss to follow up (drop outs) before end of study.

We individually extracted adverse event data for each adverse event wherever possible. For trials reporting adverse events for more than one dose, we recorded the average number of people reporting each adverse event for each dose was recorded.

Assessment of risk of bias in included studies

Two reviewers (KWS, EG) assessed the included trials for methodological quality. We extracted information about the method of treatment allocation (random, quasi‐random, sequential, not stated), concealment of allocation (adequate (A), unclear (B), inadequate (C) according to Juni 2001), blinding (double, single, or not blind), sample size (< 1000, ≥ 1000), and exclusions after randomization (reported, not reported), and the follow‐up period (one rotavirus season after immunization, two or more rotavirus seasons).

Data synthesis

We analysed dichotomous data by calculating the risk ratio (RR) for each trial with the uncertainty in each result being expressed using 95% confidence intervals (CI). Additionally, we tested variability according to differences in type of vaccine (bovine, human, or rhesus).

Using the method of Newcombe‐Wilson hybrid score (not continuity corrected (Newcombe 1998)) and the correspondent 95% CI, we estimated the number of children needed to take rotavirus vaccine to avoid one case of rotavirus diarrhoea (NNT) for trials that tested each specific vaccine.

Heterogeneity and publication bias

We initially assessed heterogeneity on the results of the trials by inspecting of graphical presentations and by calculating a test of heterogeneity (chi‐squared test). However, we were aware of the fact that the chi‐squared test has a poor ability to detect statistically significant heterogeneity among studies. Therefore, we have also quantified the impact of heterogeneity in the meta‐analysis using a measure of the degree of inconsistency in the studies' results (Higgins 2003). This measure (I²) describes the percentage of total variation across studies that are due to heterogeneity rather than the play of chance (Higgins 2003). The values of I² lies between 0 and 100%, and a simplified categorization of heterogeneity could be low, moderate, and high to I² values of 25, 50, and 75% (Higgins 2003).

We anticipated between‐trial variation in estimation of vaccine efficacy for trials that used different rotavirus strains and were conducted in different geographic locations. We performed subgroup analyses in order to assess the impact of these possible sources of heterogeneity. In addition, we performed sensitivity analyses for the main outcome (rotavirus diarrhoea) in order to assess the robustness of the findings to different aspects of the trials methodology: concealment of allocation (adequate or unclear); exclusions after randomization (reported or not reported); sample size (< 1000, ≥1000); and length of follow up (one rotavirus season or more than one rotavirus season). We carried out further exploratory sensitivity analyses to test the impact of different types of vaccines, concomitant use of diphtheria‐tetanus‐pertussis (DPT) and/or oral polio vaccine (OPV), breastfeeding during vaccination, and number of doses of vaccine (single dose, two to three doses) in the pooled efficacy.

We used a fixed effect model throughout the review, except in the event of statistically significant heterogeneity between the trials (P < 0.10), when we chose the random effects model. And because we demonstrated heterogeneity in most of the outcomes when we performed the analyses using odds ratio or RR either on a fixed effect or random effects model (see Appendix 2 for an example), we chose to present the results using the RR with the correspondent 95% CI and a random effects model.

We observed moderate to high statistically heterogeneity among most of the outcomes evaluated on this review. Even when we analysed the vaccines separately, we observed heterogeneity in the combined trials for the following vaccines and outcomes: RRV‐TV (episodes of rotavirus diarrhoea, episodes of severe rotavirus diarrhoea), RRV (all‐cause diarrhoea), RIT‐4237 (episodes of rotavirus diarrhoea, episodes of severe rotavirus diarrhoea, all‐cause diarrhoea), and WC3 (episodes of rotavirus diarrhoea) vaccines.

We examined a funnel plot estimating the precision of trials (plots of logarithm of the RR for efficacy against the sample size) in order to estimate potential asymmetry.

Personal communication with authors and pharmaceutical industry

We contacted the corresponding authors of the 64 included trials, of which 25 replied and 20 provided additional data. We initially sought data on study design, death, bowel obstruction and/or intussusception, and dehydration. Because of the low response rate in the first round, we attempted to contact the authors for a second time using only e‐mail and requesting information only for deaths and adverse events. The results of the personal communication with the authors are summarized in Appendix 3. These data were previously published for only three of the trials (Joensuu 1997; Santosham 1997; Bresee 2001).

We also contacted Merck Sharp & Dohme − the pharmaceutical industry organizing the field trials of the new bovine quadrivalent vaccine − for further details on their trials. Dr Penny Heaton (Vaccine Infectious Diseases, Clinical Research) replied and informed us that most of the trials will be published, and they would prefer not to release any further information until after publication (e‐mail on 11 March 2002).

Results

Description of studies

Eligibility

We identified 138 potentially relevant references through our search strategy, and 1 trial through contact with authors, who also provided unpublished data (Perez‐Schael 2002). Of these, we included 64 trials (90 references), which are described below and in the 'Characteristics of included studies', and excluded 31 studies (33 references) and 16 reviews of the literature.

We have described the 31 excluded studies in the 'Characteristics of excluded studies'. Our main reasons for excluding these studies were because they were not randomized or had an inadequate allocation concealment (7), did not measure any clinical outcome (21), and because all participants received the vaccine and were randomized to different dietary components (3).

Two randomized controlled trials are still awaiting assessment because currently only an abstract without the required information has been published (Seth 2000; Vesikari 2002).

Participants

All but one study tested vaccines or placebo in children, age ranging from newborns to 12 years old; Barnes 1997 is a small safety study included a group of 10 adults (21 to 30 years old). Studies included only healthy children; in a single study, we found a description that children were included if they did not have any debilitating disease (Perez‐Schael 1990c).

Interventions

Details of the six different types of vaccines assessed in the included studies can be found in Appendix 4. Only one trial did not use placebo in the control group (Vesikari 1986a). In another trial (Santosham 1991a and Santosham 1991b), children were allocated to three groups: RIT 4237 (106 children); MMU18006 (108 children); placebo (107 children); the placebo group was arbitrarily divided into two arms, and we report each arm as a separate study.

Outcomes

Of the 64 included trials, 49 trials provided information on the safety of rotavirus vaccine in 16,602 children, and 40 trials provided information on the effectiveness of the vaccine to prevent rotavirus diarrhoea in 21,070 children. See the 'Characteristics of included studies for the complete list of outcomes provided by each study; no study reported on malnourishment.

Rotavirus was detected in all studies using serology and stool analysis by the ELISA method. An episode of diarrhoea was defined in the majority of the studies as at least two liquid or semi‐liquid stools in 24 hours. The definition of severity of diarrhoea varied according to the criteria used in the trial. Thirty‐seven studies used only a clinical definition of severe diarrhoeal diseases. The other studies used either the Clark 1988 scale (4 studies), Flores 1987 scale (7 studies), Rennels 1996 scale (3 studies), Ruuska 1990 scale (6 studies) or the World Health Organization (WHO) criteria (WHO 1995; 7 studies) for diarrhoea to measure the severity of diarrhoeal diseases.

Twenty‐three trials had an unbalanced randomization with more than one experimental group, which may affect the calculation of the NNT: 10 trials tested different strains of the same vaccine in two to four experimental groups; 10 trials tested different dosages of the same vaccine; and 3 trials tested different vaccines in two experimental groups. Of these three trials two provided data only on safety of the vaccine (children were followed up for up to 10 days after vaccination) (Flores 1990; Perez‐Schael 1994); we used data on the placebo and rhesus vaccines groups and discarded data on the M37 vaccine. The other trial provided data on the efficacy of two rotavirus vaccines (RIT 4237, MMU18006) against placebo (Santosham 1991a; Santosham 1991b).

Study location

We divided the trials according to the classification of high‐, middle‐, and low‐income countries of the World Bank (World Bank 2003). Trials were conducted in high‐income countries: Australia (2), Austria (1), Finland (12), Israel (1), Spain (1), Sweden (1), the United Kingdom (1), the United States of America (29); middle‐income countries: Brazil (1), Peru (3), and Venezuela (9); and low‐income countries: Bangladesh (1), Central African Republic (1), Gambia (1), Rwanda (1).

Risk of bias in included studies

Generation of allocation sequence

Seven trials reported information about generation of allocation sequence in the published version. They used random code assignment or computer‐generated randomization. The code was kept by the pharmaceutical industry or academic institutions (including World Health Organization and National Institute of Health, USA) and not broken until the end of the study. In eight trials, randomization was generated by a central computer, and pre‐coded vessels were administered sequentially to the children. It was not possible to determine the method used in the remaining 49 trials.

Concealment of allocation

Fifteen trials had adequate allocation concealment (information was available in the publication in seven trials and sought from authors in eight trials), and scored A. In the remained 49 trials it was not possible to determine the method used to conceal the allocation and these trials scored B (unclear).

Blinding

All trials used placebo in the control arm. Sixty‐one trials were reported to be double blind, and three studies provided no information. Only 29 out of the 61 studies reported a procedure to guarantee double blinding. The vaccine and placebo were of identical appearance in 28 trials. In another trial (Vesikari 1986a), parents and paediatricians were unaware of the code given to vaccine and placebo.

Exclusions after randomization

Exclusion of participants from the trials after randomization varied from 0 to 23%, except in Hanlon 1987, where 43% of the randomized children were excluded because they were absent during the epidemic period or too young to be vaccinated.

From the 40 included trials that provided data on efficacy: 6 did not provide any information about drop outs before end of study; in the other trials exclusion of participants from the trials after randomization varied from 0 to 23%, except in Hanlon 1987, where 43% of the randomized children were excluded because they were absent during the epidemic period or were too young to be vaccinated. Nineteen trials (n = 12,493) provided data for each group (vaccine or placebo) on the number of children dropped out before the end of the study. There was no statistically significant difference in the number of people who dropped out between the vaccine and placebo arms of the rhesus rotavirus vaccine trials (RR 0.95, 95% CI 0.85 to 1.06; Analysis 1.1).

1.1 — Comparison 1 Drop outs before end of study, Outcome 1 Drop outs before end of study.

Follow‐up period

The follow‐up period was less than a month in 21 trials, up to one year (one season) in 20 trials, and of one to four years (two or more seasons) in 23 trials.

We have provided additional information about the methodological quality of the trials in the Characteristics of included studies'.

Effects of interventions

Rotavirus diarrhoea

Episodes

Forty trials provided information for 21,070 children regarding diarrhoeal disease caused by rotaviruses. Rotavirus vaccine prevents rotavirus diarrhoea in all three groups of vaccines: rhesus (RR 0.59, 95% CI 0.50 to 0.70); bovine (RR 0.59, 95% CI 0.45 to 0.76); and human (RR 0.42, 95% CI 0.21 to 0.85) (Analysis 2.1).

2.1 — Comparison 2 Rotavirus diarrhoea, Outcome 1 Episodes.

Converting these RR values to vaccine efficacy gives the following estimates for preventing on episode of rotavirus diarrhoea: 41% (95% CI 30 to 50%) for rhesus vaccines; 41% (95% CI 24 to 55%) for bovine vaccines; and 58% (95% CI 15 to 79%) for human vaccines.

Severe episodes

Rotavirus vaccines also showed a highly statistically significant benefit in preventing severe episodes of rotavirus infection (Analysis 2.2): rhesus (RR 0.42, 95% CI 0.31 to 0.57); bovine (RR 0.38, 95% CI 0.24 to 0.60); and human (RR 0.21, 95% CI 0.13 to 0.35).

2.2 — Comparison 2 Rotavirus diarrhoea, Outcome 2 Severe episodes.

Converting these RR values to vaccine efficacy gives the following estimates to prevent one episode of severe rotavirus diarrhoea: 58% (95% CI 43 to 69%) for rhesus vaccines; 62% (95% CI 40 to 76%) for bovine vaccines; and 79% (95% CI 65 to 87%) for human vaccines.

Episodes requiring hospitalization

Rotavirus vaccines also showed a highly statistically significant benefit in preventing severe rotavirus infection requiring hospitalization (Analysis 2.3): rhesus (RR 0.48, 95% CI 0.27 to 0.86); bovine (RR 0.37, 95% CI 0.18 to 0.74); and human (RR 0.21, 95% CI 0.09 to 0.48).

2.3 — Comparison 2 Rotavirus diarrhoea, Outcome 3 Episodes requiring hospitalization.

Converting these RR values to vaccine efficacy gives the following estimates to prevent one episode of rotavirus diarrhoea requiring hospitalization: 52% (95% CI 14 to 73%) for rhesus vaccines; 63% (95% CI 26 to 82%) for bovine vaccines; and 79% (95% CI 52 to 91%) for human vaccines.

Episodes of more than four days duration

Only the rhesus vaccine significantly reduced the number of cases of rotavirus diarrhoea with more than four days of duration (rhesus: RR 0.61, 95% CI 0.42 to 0.87; bovine: RR 1.01, 95% CI 0.51 to 1.99; Analysis 2.4).

2.4 — Comparison 2 Rotavirus diarrhoea, Outcome 4 Episodes of more than four days.

All‐cause diarrhoea

Episodes

Twenty‐three trials provided information about all‐cause diarrhoea for 11,296 children. A further nine trials provided information on episodes of diarrhoeal disease per person‐year in 3944 children. We could not combine the latter in a meta‐analysis and have summarized them in Analysis 3.1. Meta‐analysis was also not possible for human vaccines, as only one trial provided data on this outcome.

3.1. Analysis.

Comparison 3 All‐cause diarrhoea, Outcome 1 Episodes (number of episodes/number of person‐years).

Episodes (number of episodes/number of person‐years)
Study	Episodes (vaccine)	Person‐years vaccine	Episodes (placebo)	Person‐years placebo
Rhesus rotavirus vaccines
Gothefors 1989	75	53	79	51
Lanata 1996a	9104	1032	3000	339
Lanata 1996b	6518	758	3200	386
Linhares 1996	2579	361	2803	363
Perez‐Schael 1990a	294	141	290	130
Santosham 1991a	397	130	383	125
Santosham 1997	2559	1391	1331	680
Bovine rotavirus vaccines
Bernstein 1990	153	102 (patients)	179	103 (patients)
Lanata 1989	3432	358	1343	127
Santosham 1991b	404	125	383	125
Human rotavirus vaccines
Barnes 1997	62	40 (patients)	23	20 (patients)
Georges‐Courbot 1991	503	232 (patients)	455	230 (patients)

Date	Event	Description
11 August 2009	Amended	Text added to abstract to highlight current actions: The authors are aware that further trials have been conducted for some of the vaccines included in this review. Authors are preparing new reviews that will update the evidence for vaccines approved for use and vaccines in development.
22 October 2008	Amended	Converted to new review format with minor editing.

Vaccine	Years on trial	Serotype
Live attenuated bovine rotavirus (NCDV ‐ RIT 4237)	1982 to 1987	G6, P6
Live attenuated rhesus monovalent rotavirus (RRV ‐ MMU18006)	1984 to 1988	G3, P3
Live attenuated bovine rotavirus (WC3)	1986 to 1990	G6, P5
Live attenuated human rotavirus (M37)	1990 to 1994	G1
Live attenuated rhesus‐human reassortant rotavirus tetravalent (RRV‐TV: RRV‐D + RRV‐DS1 + RRV + RRV‐ST3)	1989 to present	G1‐4, P3
Live attenuated bovine‐human ressortant multivalent (QHBRV)	1994 to present	G1‐3, P5, P8
Live attenuated human rotavirus (HRV 89‐12)	1996 to present	G1
Live attenuated human‐bovine rotavirus (W179‐9)	1988 to present	G1, P8

Type of vaccine	Random‐effects model		Fixed‐effect model		Heterogeneity test
Type of vaccine	Risk ratio (95% CI)	Z statistics	Risk ratio (95% CI)	Z statistics	Heterogeneity test
Rhesus tetravalent vaccine (RRV‐TV)	0.58 (0.47 to 0.72)	4.96, P < 0.00001	0.59 (0.54 to 0.64)	11.96, P < 0.00001	Chi squared 51.64, df 10, P < 0.00001, I squared 80.6%
Rhesus monovalent vaccine (RRV, RRV‐D, RRV‐DS1, or RRV‐ST3)	0.60 (0.46 to 0.78)	3.85, P 0.0001	0.59 (0.47 to 0.75)	4.42, P < 0.00001	Chi squared 9.80, df 8, P 0.28, I squared 18.4%
Bovine vaccine (RIT‐4237)	0.65 (0.48 to 0.86)	2.96, P 0.003	0.68 (0.58 to 0.79)	4.84, P < 0.00001	Chi squared 19.77, df 9, P 0.02, I squared 54.5%
Bovine vaccine (WC3)	0.77 (0.46 to 1.28)	1.00, P 0.32	0.85 (0.66 to 1.10)	1.22, P 0.22	Chi squared 5.09, df 2, P 0.08, I squared 60.7%
Bovine‐human vaccine (W179‐9)	0.33 (0.19 to 0.56)	4.00, P 0.00001	0.29 (0.17 to 0.51)	4.36, P < 0.0001	Chi squared 1.58, df 2, P 0.45, I squared 0%
Human vaccine (89‐12)	0.32 (0.23 to 0.45)	6.52, P < 0.00001	0.31 (0.22 to 0.44)	6.60, P < 0.00001	Chi squared 0.79, df 1, P 0.37, I squared 0%

Trial	Type of vaccine	Intussusceptions	Deaths	Cause of deaths
Gothefors 1989	RRV‐MV	No cases	No information	No information
Perez‐Schael 1990a	RRV‐MV	No cases	1 case, group allocation unknown	Described by authors as "not related to vaccine"
Wright 1987	RRV‐MV	No cases	No cases	No cases
Wright 1991	RRV‐MV	No cases	No cases	No cases
Bresee 2001	RRV‐TV	No cases	2 cases, 1 case placebo group and 1 case vaccine group	Pneumonia and sudden death syndrome
Ing 1991	RRV‐TV	No cases	No cases	No cases
Dennehy 1996	RRV‐TV	No cases	No cases	No cases
Joensuu 1997	RRV‐TV	Vaccine: 1 case, 6 days after thirrd dose; Placebo: 1 case, 44 days after second dose	1 case, placebo group	Suffocation due to foreign body in the bronchia, 11 months old boy
Lanata 1996a	RRV‐TV	No cases	13 cases, group allocation unknown	Described by authors as "not related to vaccine"
Lanata 1996b	RRV‐TV	No cases	6 cases, group allocation unknown	Described by authors as "not related to vaccine"
Linhares 1996	RRV‐TV	Placebo: 1 case, during follow up	7 cases, group allocation unknown	Gastrointestinal reflux + asphyxia, measles + sepsis, poisoning, gastroenteritis + pneumonia, acute respiratory distress
Perez‐Schael 1997	RRV‐TV	No cases	3 cases, placebo group	Bronchial aspiration + diarrhoea + dehydration, pneumonia + sepsis, accidental intoxication
Santosham 1997	RRV‐TV	Placebo: 1 case	4 cases, vaccine group, all after one month of receiving the vaccine	HiB (hemophilus influenza B vaccine), sudden death syndrome, asphyxia, unknown
Lanata 1989	RIT‐4237	No cases	8 cases	No information
Maldonado 1986	RIT‐4237	No cases	No cases	No cases
Senturia 1987	RIT‐4237	No cases	No cases	No cases
Mutz 1989	RIT‐4237	No cases	No cases	No cases
Christy 1993	W179‐9	No cases	No cases	No cases
Bernstein 1998	HRV89‐12	No cases	No cases	No cases
Bernstein 1999	HRV89‐12	No cases	1 case	Pneumococcal sepsis
Georges‐Courbot 1991	WC3	No cases	11 cases, group allocation unknown	Described by authors as "not related to vaccine"
Ho 1989	WC3	Vaccine: 2 cases (1 week and 1 month after)	No cases	No cases

Rotavirus vaccine	Trials	Dosage
RIT‐4237: live attenuated bovine	Hanlon 1987, Lanata 1989, Maldonado 1986 Ruuska 1990, Santosham 1991a, Santosham 1991b, Senturia 1987 Vesikari 1984, Vesikari 1985a, Vesikari 1986a, Vesikari 1987b Vesikari 1991a	10^7.8 to 10^8.3 TCID50 or 10^6.5 to 10^8 PFU
WC3: live attenuated bovine	Bernstein 1990, Clark 1988	10^7 to 10^7.5 PFU
W179‐9: live attenuated human‐bovine	Clark 1990, Christy 1993, Treanor 1995	10^7 to 10^7.5 PFU
RRV‐MV (RRV or RRV‐D or RRV‐DS1 or RRV‐ST3): live attenuated rhesus monovalent	Anderson 1986, Barnes 1997, Christy 1988, Flores 1988, Flores 1989, Gothefors 1989, Ho 1989, Madore 1992, Perez‐Schael 1990a, Pichichero 1990, Rennels 1987, Santosham 1991a, Santosham 1991b Vesikari 1990, Wright 1987, Wright 1991	10^3 to 10^5 PFU
RRV‐TV: live attenuated rhesus tetravalent	Bernstein 1995, Dagan 1992 Dennehy 1996 Escobar 1988,Flores 1990 Flores 1993, Ing 1991, Joensuu 1997, Lanata 1996a Lanata 1996b Linhares 1996, Perez‐Schael 1997, Pichichero 1993, Rennels 1995, Rennels 1996, Santosham 1997, Vesikari 1999	4 x 10^4 to 4 x 10^6 PFU
HRV 89‐12: live attenuated human	Bernstein 1998 Bernstein 1999	10^4 to 10^5 PFU
QHBRV: live attenuated human‐bovine quadrivalent or pentavalent	Clark 1995	4 x 10^7 PFU

Outcome	Vaccine	No. trials	No. participants	RR (95% CI)	NNT (95% CI)	Analysis
Rotavirus diarrhoea: episodes	Rhesus tetravalent (RRV‐TV)	11	11,590	0.58 (0.47 to 0.72)	19 (15 to 25)	Analysis 10.1
	Rhesus monovalent (RRV, RRV‐D, RRV‐DS1, or RRV‐ST3)	9	1715	0.60 (0.46 to 0.78)	13 (9 to 22)	Analysis 10.1
	Bovine (RIT‐4237)	10	3663	0.65 (0.48 to 0.86)	—	Analysis 10.1
	Bovine (WC3)	3	782	0.77 (0.46 to 1.28)	—	Analysis 10.1
	Bovine‐human (W179‐9)	3	433	0.33 (0.19 to 0.56)	7 (5 to 12)	Analysis 10.1
	Bovine‐human (WC3‐QV)	1	405	0.33 (0.19 to 0.58)	7 (5 to 13)	Analysis 10.1
	Human (89‐12)	2	2201	0.32 (0.23 to 0.45)	11 (8 to 15)	Analysis 10.1
	Human (M37)	1	281	1.01 (0.45 to 2.25)	—	Analysis 10.1
Rotavirus diarrhoea: severe episodes	Rhesus tetravalent (RRV‐TV)	7	9569	0.40 (0.28 to 0.57)	18 (15 to 22)	Analysis 10.2
	Rhesus monovalent (RRV‐MMU 18006)	4	1037	0.52 (0.27 to 0.97)	23 (13 to 72)	Analysis 10.2
	Bovine (RIT‐4237)	5	2131	0.38 (0.19 to 0.79)	24 (16 to 53)	Analysis 10.2
	Bovine (WC3)	3	782	0.57 (0.29 to 1.12)	14 (8 to 37)	Analysis 10.2
	Bovine‐human (W179‐9)	1	325	0.14 (0.05 to 0.42)	—	Analysis 10.2
	Bovine‐human (QHBRV)	1	405	0.31 (0.16 to 0.62)	—	Analysis 10.2
	Human (89‐12)	2	2201	0.21 (0.13 to 0.35)	14 (10 to 20)	Analysis 10.2
All‐cause diarrhoea: episodes	Rhesus tetravalent (RRV‐TV)	5	6561	0.89 (0.84 to 0.93)	12 (10 to 17)	Analysis 10.3
	Rhesus monovalent (RRV or RRV‐D or RRV‐DS1 or RRV‐ST3)	6	1145	0.82 (0.69 to 0.98)	8 (5 to 13)	Analysis 10.3
	Bovine (RIT‐4237)	7	2772	0.82 (0.67 to 1.01)	16 (10 to 35)	Analysis 10.3
	Bovine (WC3)	1	104	0.48 (0.28 to 84)	—	Analysis 10.3
	Bovine‐human (W179‐9)	3	433	0.60 (0.50 to 0.72)	4 (3 to 7)	Analysis 10.3
	Human (M37)	1	281	0.91 (0.57 to 1.44)	—	Analysis 10.3

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Drop outs before end of study	19		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	14	11128	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.85, 1.06]
1.2 Bovine rotavirus vaccines	5	1365	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.56, 1.22]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Episodes	40		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	20	13305	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.50, 0.70]
1.2 Bovine rotavirus vaccines	17	5283	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.45, 0.76]
1.3 Human rotavirus vaccines	3	2482	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.21, 0.85]
2 Severe episodes	23		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	11	10606	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.31, 0.57]
2.2 Bovine rotavirus vaccines	10	3643	Risk Ratio (M‐H, Random, 95% CI)	0.38 [0.24, 0.60]
2.3 Human rotavirus vaccines	2	2201	Risk Ratio (M‐H, Random, 95% CI)	0.21 [0.13, 0.35]
3 Episodes requiring hospitalization	13		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	7	7803	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.27, 0.86]
3.2 Bovine rotavirus vaccines	4	1693	Risk Ratio (M‐H, Random, 95% CI)	0.37 [0.18, 0.74]
3.3 Human rotavirus vaccines	2	2201	Risk Ratio (M‐H, Random, 95% CI)	0.21 [0.09, 0.48]
4 Episodes of more than four days	8		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1 Rhesus rotavirus vaccines	5	631	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.42, 0.87]
4.2 Bovine rotavirus vaccines	3	252	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.51, 1.99]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Episodes (number of episodes/number of person‐years)			Other data	No numeric data
1.1 Rhesus rotavirus vaccines			Other data	No numeric data
1.2 Bovine rotavirus vaccines			Other data	No numeric data
1.3 Human rotavirus vaccines			Other data	No numeric data
2 Episodes (number of cases/number of participants)	23		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	11	7706	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.80, 0.92]
2.2 Bovine rotavirus vaccines	11	3309	Risk Ratio (M‐H, Random, 95% CI)	0.73 [0.60, 0.89]
2.3 Human rotavirus vaccines	1	281	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.57, 1.44]
3 Episodes during first week after vaccine (one or multiple doses)	40		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	29	12673	Risk Ratio (M‐H, Fixed, 95% CI)	1.06 [0.98, 1.14]
3.2 Bovine rotavirus vaccines	10	2125	Risk Ratio (M‐H, Fixed, 95% CI)	1.02 [0.89, 1.17]
3.3 Human rotavirus vaccines	3	376	Risk Ratio (M‐H, Fixed, 95% CI)	1.94 [1.03, 3.65]
4 Severe episodes (clinical or measured by any scale)	6		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1 Rhesus rotavirus vaccines	3	1663	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.72, 1.00]
4.2 Bovine rotavirus vaccines	3	714	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.21, 1.26]
5 Episodes requiring re‐hydration	11		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Rhesus rotavirus vaccines	5	5559	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.30, 0.88]
5.2 Bovine rotavirus vaccines	5	2148	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.26, 0.67]
5.3 Human rotavirus vaccines	1	215	Risk Ratio (M‐H, Random, 95% CI)	0.14 [0.01, 2.71]
6 Episodes requiring hospitalization	7		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.1 Rhesus rotavirus vaccines	4	5023	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.66, 0.99]
6.2 Bovine rotavirus vaccines	3	799	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.16, 1.91]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 All‐cause death	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	4	6029	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.16, 3.12]

Methods	Randomization: no details Allocation concealment: no details Blinding: double (no further details) Data collection: no information Intention to treat: no Interim analysis: none Exclusion from analysis: 5/182 were excluded for exceeding the age limit Follow‐up period: 1 week after each vaccination
Participants	Age: 1.5 to 60 months Health status: no information Breastfeeding: no information Immunization status: no vaccination allowed within 2 weeks of rotavirus vaccination
Interventions	1. Human‐bovine D x UK, 10^4.8 PFU, single dose (n = 17) 2. Human‐bovine D x UK, 10^5.8 PFU, single dose (n = 20) 3. Human‐bovine DS‐1 x UK, 10^4.3 PFU, single dose (n = 18) 4. Human‐bovine DS‐1 x UK, 10^5.3 PFU, single dose (n = 11) 5. Human‐bovine P x UK, 10^4.3 PFU, single dose (n = 10) 6. Human‐bovine P x UK, 10^5.3 PFU, single dose (n = 21) 7. Human‐bovine ST3 x UK, 10^4.8 PFU, single dose (n = 8) 8. Human‐bovine ST3 x UK, 10^5.8 PFU, single dose (n = 14). 9. Placebo (buffered solution), single dose (n = 63) 30 ml of formula mixed with 0.4 g of sodium bicarbonate given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of each vaccination
Notes	Study location: USA 1 h fasting before vaccination Clinical symptoms: clinical evaluation; diarrhoea was defined as 3 or more unformed stools within 48 h; fever was defined as a rectal temperature > 38.1 ºC in paediatric participants, confirmed within 20 minutes Laboratory studies: stool analysis by ELISA and PCR; serology by plaque reduction neutralization and ELISA

Methods	Randomization: no details Allocation concealment: no details Blinding: double (no further details) Data collection: no information Intention to treat: yes Interim analysis: none Exclusion from analysis: none Follow‐up period: 1 week after each vaccination
Participants	Age: 1.5 to 60 months Health status: no information Breastfeeding: no information Immunization status: no vaccination allowed within 2 weeks of rotavirus vaccination
Interventions	1. Human‐bovine Wa x UK, 10^6.3 PFU, single dose (n = 21) 2. Human‐bovine Wa x UK, 10^7.3 PFU, single dose (n = 21) 3. Human‐bovine Wa x (DS‐1xUK), 10^6.4 PFU, single dose (n = 14) 4. Human‐bovine Wa x (DS‐1xUK), 10^7.4 PFU, single dose (n = 22) 5. Placebo (buffered soy‐based formula), single dose (n = 37) 30 ml of formula mixed with 0.4 g of sodium bicarbonate given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of each vaccination
Notes	Study location: USA 1 h fasting before vaccination Clinical symptoms: clinical evaluation; diarrhoea was defined as 3 or more unformed stools within 48 h; fever was defined as a rectal temperature > 38.1 ºC in paediatric participants, confirmed within 20 minutes Laboratory studies: stool analysis by ELISA and PCR; serology by plaque reduction neutralization and ELISA

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Adequate allocation concealment (A)	12		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	9	8963	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.45, 0.76]
1.2 Bovine rotavirus vaccines	1	239	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.33, 1.56]
1.3 Human vaccine rotavirus	2	496	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.12, 1.99]
2 Unclear allocation concealment (B)	28		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	11	4423	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.48, 0.67]
2.2 Bovine rotavirus vaccines	16	5040	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.43, 0.76]
2.3 Human vaccine rotavirus	1	1986	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.24, 0.51]
3 Exclusions after randomization reported	20		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	15	12009	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.48, 0.71]
3.2 Bovine rotavirus vaccines	5	1361	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.26, 0.96]
4 Exclusions after randomization not reported	20		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1 Rhesus rotavirus vaccines	5	1377	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.44, 0.76]
4.2 Bovine rotavirus vaccines	12	3918	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.46, 0.82]
4.3 Human vaccine rotavirus	3	2482	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.21, 0.85]
5 Sample size of less than 1000 participants	33		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Rhesus rotavirus vaccines	15	4975	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.56, 0.79]
5.2 Bovine rotavirus vaccines	16	4279	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.41, 0.74]
5.3 Human vaccine rotavirus	2	496	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.12, 1.99]
6 Sample size of at least 1000 participants	7		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.1 Rhesus rotavirus vaccines	5	8411	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.39, 0.65]
6.2 Bovine rotavirus vaccines	1	1000	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.61, 1.41]
6.3 Human vaccine rotavirus	1	1986	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.24, 0.51]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Multiple doses (two or more) of rotavirus vaccine	15		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	8	9698	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.44, 0.69]
1.2 Bovine rotavirus vaccines	5	2148	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.39, 0.86]
1.3 Human rotavirus vaccines	2	2201	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.23, 0.45]
2 Single dose of rotavirus vaccine	25		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	12	3607	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.50, 0.82]
2.2 Bovine rotavirus vaccines	12	3135	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.40, 0.83]
2.3 Human rotavirus vaccines	1	281	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.45, 2.25]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Concomitant use of DPT or OPV allowed	12		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	8	10052	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.44, 0.73]
1.2 Bovine rotavirus vaccines	3	1071	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.44, 1.18]
1.3 Human rotavirus vaccines	1	1986	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.24, 0.51]
2 Concomitant use of DPT or OPV not allowed	15		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	6	1367	Risk Ratio (M‐H, Random, 95% CI)	0.71 [0.52, 0.97]
2.2 Bovine rotavirus vaccines	7	2110	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.40, 1.00]
2.3 Human rotavirus vaccines	2	496	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.12, 1.99]
3 No information	13		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	6	1886	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.43, 0.71]
3.2 Bovine rotavirus vaccines	7	2102	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.28, 0.79]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Breastfeeding allowed during vaccination	13		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	4	3652	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.32, 0.86]
1.2 Bovine rotavirus vaccines	9	3452	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.48, 0.85]
2 Breastfeeding not allowed for at least one hour before and after vaccination	23		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	15	8468	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.51, 0.73]
2.2 Bovine rotavirus vaccines	6	954	Risk Ratio (M‐H, Random, 95% CI)	0.45 [0.21, 0.94]
2.3 Human rotavirus vaccines	2	496	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.12, 1.99]
3 No information	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	1	1185	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.56, 0.87]
3.2 Bovine rotavirus vaccines	2	877	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.20, 1.71]
3.3 Human rotavirus vaccines	1	1986	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.24, 0.51]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 One season	15		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	8	2885	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.42, 0.78]
1.2 Bovine rotavirus vaccines	7	1798	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.32, 0.83]
1.3 Human rotavirus vaccines	1	1986	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.24, 0.51]
2 Two or more seasons	24		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	12	10172	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.48, 0.74]
2.2 Bovine rotavirus vaccines	10	3726	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.43, 0.84]
2.3 Human rotavirus vaccines	2	496	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.12, 1.99]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Rotavirus diarrhoea: episodes	40		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus tetravalent vaccine (RRV‐TV)	11	11590	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.47, 0.72]
1.2 Rhesus monovalent vaccine (RRV, RRV‐D, RRV‐DS1, or RRV‐ST3)	9	1715	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.46, 0.78]
1.3 Bovine vaccine (RIT‐4237)	10	3663	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.48, 0.86]
1.4 Bovine vaccine (WC3)	3	782	Risk Ratio (M‐H, Random, 95% CI)	0.77 [0.46, 1.28]
1.5 Bovine‐human vaccine (W179‐9)	3	433	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.19, 0.56]
1.6 Bovine‐human vaccine (WC3‐QV)	1	405	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.19, 0.58]
1.7 Human vaccine (89‐12)	2	2201	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.23, 0.45]
1.8 Human vaccine (M37)	1	281	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.45, 2.25]
2 Rotavirus diarrhoea: severe episodes	23		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus tetravalent vaccine (RRV‐TV)	7	9569	Risk Ratio (M‐H, Random, 95% CI)	0.40 [0.28, 0.57]
2.2 Rhesus monovalent vaccine (RRV‐MMU 18006)	4	1037	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.27, 0.97]
2.3 Bovine vaccine (RIT‐4237)	5	2131	Risk Ratio (M‐H, Random, 95% CI)	0.38 [0.19, 0.79]
2.4 Bovine vaccine (WC3)	3	782	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.29, 1.12]
2.5 Bovine‐human vaccine (W179‐9)	1	325	Risk Ratio (M‐H, Random, 95% CI)	0.14 [0.05, 0.42]
2.6 Bovine‐human vaccine (QHBRV)	1	405	Risk Ratio (M‐H, Random, 95% CI)	0.31 [0.16, 0.62]
2.7 Human vaccine (89‐12)	2	2201	Risk Ratio (M‐H, Random, 95% CI)	0.21 [0.13, 0.35]
3 All‐cause diarrhoea: episodes (number of cases/number of participants)	23		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Rhesus tetravalent vaccine (RRV‐TV)	5	6561	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.84, 0.93]
3.2 Rhesus monovalent vaccine (RRV, RRV‐D, RRV‐DS1, or RRV‐ST3)	6	1145	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.69, 0.98]
3.3 Bovine vaccine (RIT‐4237)	7	2772	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.67, 1.01]
3.4 Bovine vaccine (WC3)	1	104	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.28, 0.84]
3.5 Bovine‐human vaccine (W179‐9)	3	433	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.50, 0.72]
3.6 Human vaccine (M37)	1	281	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.57, 1.44]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Bowel obstruction and/or intussusception	3		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Rhesus rotavirus vaccines	3	4123	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.07, 2.42]
2 Fever (one or multiple doses of vaccine)	47		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Rhesus rotavirus vaccines	32	13718	Risk Ratio (M‐H, Random, 95% CI)	2.00 [1.51, 2.64]
2.2 Bovine rotavirus vaccines	12	2168	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.73, 1.23]
2.3 Human rotavirus vaccines	5	716	Risk Ratio (M‐H, Random, 95% CI)	1.75 [0.85, 3.64]
3 Vomiting (one or multiple doses of vaccine)	28		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
3.1 Rhesus rotavirus vaccines	16	10289	Risk Ratio (M‐H, Fixed, 95% CI)	0.96 [0.90, 1.03]
3.2 Bovine rotavirus vaccines	10	2016	Risk Ratio (M‐H, Fixed, 95% CI)	1.05 [0.90, 1.22]
3.3 Human rotavirus vaccines	2	331	Risk Ratio (M‐H, Fixed, 95% CI)	1.94 [1.00, 3.75]
4 Irritability (one or multiple doses of vaccine)	9		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
4.1 Rhesus rotavirus vaccines	5	807	Risk Ratio (M‐H, Fixed, 95% CI)	1.01 [0.85, 1.20]
4.2 Bovine rotavirus vaccines	3	512	Risk Ratio (M‐H, Fixed, 95% CI)	1.08 [0.86, 1.36]
4.3 Human rotavirus vaccines	1	215	Risk Ratio (M‐H, Fixed, 95% CI)	0.70 [0.51, 0.98]

Methods	Randomization: block randomization in groups of 6 independently done by the University of Melbourne Statistical Consulting Centre Allocation concealment: no details Blinding: double (vaccine and placebo ampoules identical) Data collection: no details Intention to treat: no Interim analysis: none Exclusion from analysis: 1/60 withdrawn from follow up by parents; 5/59 excluded from analysis due to wild‐type infection before or during trial Follow‐up period: 7 days after vaccination; 2 rotavirus seasons after completion of all vaccinations
Participants	Age: 3 months Health status: healthy Immunization status: vaccine administered not closer than 14 days to routine immunizations Breastfeeding: > 75%
Interventions	1. Human vaccine (RV3), 6.5 x 10^5 PFU with soy‐based formula (n = 20) or without soy‐based formula (n = 20), 3 doses 2. Placebo (uninfected monkey kidney culture fluids) with soy‐based formula, 3 doses (n = 20) 1 h fasting and then 10 ml soy formula given before vaccine (only in 1 of 2 vaccine groups) or placebo 400 mg of sodium bicarbonate given to all groups before vaccination
Outcomes	1. Safety: clinical symptoms within 7 days of vaccination. 2. Efficacy: diarrhoea within 2 rotavirus seasons of beginning of trial
Notes	Study location: Australia Clinical symptoms: clinical evaluation, severity score measured by Flores 1987 scale Laboratory studies: serology by ELISA, stool analysis by ELISA and PCR

Methods	Randomization: randomization code provided by Institut Merieux Allocation concealment: no details Blinding: double (no further details) Data collection: 7 September 1987 to 30 June 1989 Intention to treat: no Interim analysis: none Exclusion from analysis: 1\206 not evaluated; 42\205 not followed over the second year because of no consent Follow‐up period: 7 days after vaccination; 2 rotavirus seasons after beginning of trial
Participants	Age: 2 to 12 months Health status: healthy Breastfeeding: < 50% Immunization status: DTP and OPV vaccination not allowed within 2 weeks of WC3 vaccination
Interventions	1. Bovine vaccine (WC3), 10^7 PFU, single dose (n = 103) 2. Placebo (diluent), 2 ml single dose (n = 103) 30 ml formula or 1 ml/kg antacid given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of vaccination 2. Efficacy: diarrhoea within 2 rotavirus seasons of beginning of trial
Notes	Study location: USA Feeding was withheld from infants for at least 1 h before vaccine. Breastfeeding also withheld for 1 h after vaccine Clinical symptoms: clinical evaluation (diarrhoea defined as ≥ 2 watery stools in 24 h); severity score measured by Clark 1988 scale Laboratory studies: serology by ELISA and fluorescent focus reduction assay; stool analysis by plaque assay and ELISA

Methods	Randomization: random code prepared by the pharmaceutical industry Allocation concealment: no details Blinding: double (identical appearing vials) Data collection: 10 October 1989 to 1 July 1991 Intention to treat: no Interim analysis: performed after first season; no changes made on the study protocol as a result of the interim analysis Exclusion from analysis: 17\1006 not vaccinated 91\989 not included in first year follow up; 34\898 not included in second year follow up Follow‐up period: 5 days after vaccination; 2 rotavirus seasons after beginning of trial
Participants	Age: 1 to 6.5 months Health status: healthy Breastfeeding: < 50% Immunization status: OPV vaccination given within 2 weeks of RRV vaccination in 120 participants
Interventions	1. Rhesus + human tetravalent vaccine (RRV‐TV), 4 x 10^4 PFU, single dose (n = 332). 2. Monovalent vaccine (D x RRV), 4 x 10^4 PFU, single dose (n = 327) 3. Placebo (uninfected tissue culture fluids), single dose (n = 330) 400 mg sodium bicarbonate in 30 ml cow milk or soy formula given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 5 days of vaccination 2. Efficacy: gastrointestinal illness within 2 rotavirus seasons of beginning of trial
Notes	Country: USA Clinical symptoms: clinical evaluation (diarrhoea defined as ≥ 3 watery or loose stools in 24 h); severity score measured by Flores 1987 scale Laboratory studies: serology by ELISA and focus reduction assay; stool analysis by ELISA All other vaccines given as scheduled

PERMALINK

Rotavirus vaccine for preventing diarrhoea

Karla Soares‐Weiser

Elad Goldberg

Ghandi Tamimi

Leonard Leibovici

Femi Pitan

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Biology

Control

Vaccines

Sources of heterogeneity

Rotavirus

Vaccines

Study population

Study design

Significance of rotavirus vaccine

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Intervention

Control

Types of outcome measures

Primary

Rotavirus diarrhoea

Secondary

All‐cause diarrhoea

Adverse events

Search methods for identification of studies

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Data synthesis

Heterogeneity and publication bias

Personal communication with authors and pharmaceutical industry

Results

Description of studies

Eligibility

Participants

Interventions

Outcomes

Study location

Risk of bias in included studies

Generation of allocation sequence

Concealment of allocation

Blinding

Exclusions after randomization

1.1. Analysis.

Follow‐up period

Effects of interventions

Rotavirus diarrhoea

Episodes

2.1. Analysis.

Severe episodes

2.2. Analysis.

Episodes requiring hospitalization

2.3. Analysis.

Episodes of more than four days duration

2.4. Analysis.

All‐cause diarrhoea

Episodes

3.1. Analysis.

3.2. Analysis.

Episodes during first week after vaccine (one or multiple doses)

3.3. Analysis.

Severe episodes

3.4. Analysis.

Methods	Randomization: computer generated, blocks of 10, provided by sponsor Allocation concealment: code was not available until end of study Blinding: double (no further details) Data collection: August 1997 to May 1999 Intention to treat: no Interim analysis: none Exclusion from analysis: 1\215 not vaccinated because of persistent fever; 1\214 not re‐vaccinated because of congenital malformation; 1\213 died of pneumococcal sepsis; 28\212 were not available for second season follow up, no further details Follow‐up period: 1 week after each vaccination; 2 rotavirus season after beginning of trial
Participants	Age: 10 to 16 weeks Health status: healthy Breastfeeding: 50 to 75% Immunization status: no other vaccine allowed within 2 weeks of vaccination
Interventions	1. Human (89‐12), 10^5 PFU, two doses (n = 108) 2. Placebo (tissue culture medium), 1 ml, 2 doses (n = 107) 2 ml antacid given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of each vaccination. 2. Efficacy: gastroenteritis (rotavirus or other) within the study period (18 months)
Notes	Study location: USA Feeding withheld for at least 1 h before and after vaccine Clinical symptoms: clinical evaluation (gastroenteritis defined as vomiting, ≥ 3 loose stools), severity score measured by Flores 1987 scale Laboratory studies: serology by ELISA and antigen reduction assay; stool analysis by ELISA All other immunizations separated from RV by at least 2 weeks

Methods	Randomization: sequential enrolment and vaccination according to table of random numbers Allocation concealment: no details Blinding: double (similar appearance to vaccine and placebo vials) Data collection: October 1986 to July 1986 Intention to treat: no Interim analysis: none Exclusion from analysis: 1\176 for "reasons unrelated to trial" Follow‐up period: 10 days after each vaccination; 10 months after beginning of trial
Participants	Age: 2 to 4 months Health status: healthy Breastfeeding: < 50% Immunization status: DPT and polio vaccinations not allowed within 4 weeks before and 3 weeks after vaccination
Interventions	1. Rhesus (MMU 18006), 10^4 PFU, single dose (n = 88) 2. Placebo (soy formula), 1 ml single dose (n = 88) 400 mg sodium bicarbonate in 30 ml soy formula given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 10 days of each vaccination. 2. Efficacy: gastroenteritis (rotavirus or other) within 10 months of beginning of trial
Notes	Study location: USA All feedings withheld for 1 h before and after vaccine Clinical symptoms: clinical evaluation, severity measured by WHO criteria (fever (≥ 38) vomiting, dehydration, > 6 stools in 24 h, and illness duration ≥ 24 h) Laboratory studies: serology by micro neutralization assay and ELISA; stool analysis by ELISA

Methods	Randomization: block randomization scheme Allocation concealment: no details Blinding: double (no further details) Data collection: 21 November 1991 to 1 June 1992 Intention to treat: no Interim analysis: none Exclusion from analysis: 4/31 excluded for having diarrhoea from wild‐type rotavirus; 1/27 did not receive second dose "for reasons unrelated to vaccination" Follow‐up period: 10 days after each vaccination; 6 months after beginning of trial
Participants	Age: 2 to 6 months Health status: no information Breastfeeding: no information Immunization status: no information
Interventions	1. Bovine + human reassortant (WI79‐9),10^7.5 PFU, 2 doses (n = 16) 2. Placebo (cell culture medium with cherry syrup), 2.5 ml, 2 doses (n = 15) 400 mg sodium bicarbonate in 30 ml soy formula given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 10 days of each vaccination 2. Efficacy: gastroenteritis within 6 months of beginning of trial
Notes	Study location: USA Breastfeeding withheld for 1 h before vaccine Clinical symptoms: clinical evaluation (significant diarrhoea defined as ≥ 2 watery stools in 24 h) Laboratory studies: stool analysis by ELISA; serology by micro‐neutralization

Methods	Randomization: serial coding according to table of random numbers Allocation concealment: code unbroken until end of trial, no further details Blinding: double (numbered identical vials) Data collection: September 1985 to June 1986 Intention to treat: unknown (only number of infants completing the trial mentioned) Interim analysis: none Exclusion from analysis: unknown Follow‐up period: 1 week after each vaccination; 10 months after beginning of trial
Participants	Age: 3 to 12 months Health status: healthy Breastfeeding: < 50% Immunization status: DPT vaccination not allowed within 1 week before vaccination
Interventions	1. Bovine (WC3), 10^7.5 PFU, single dose (n = 49) 2. Placebo (cell culture medium with 0.5 ml cherry syrup), 2.5 ml single dose (n = 55) 30 ml formula or 85 mg antacid given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of each vaccination 2. Efficacy: gastroenteritis (rotavirus or other) within 10 months of beginning of trial
Notes	Study location: USA No feeding 1 h before vaccine, and no breastfeeding 1 h after vaccine Clinical symptoms: clinical evaluation (significant diarrhoea defined as ≥ 2 watery stools in 24 h); severity score measured by Clark 1988 (modified Duffy 1986) scale Laboratory studies: serology by plaque reduction neutralization; stool analysis by PAGE‐SS analysis and ELISA At least 1 week between DTP and RV

Methods	Randomization: table of random numbers Allocation concealment: no details Blinding: double (numbered vials) Data collection: September 1987 to July 1988 Intention to treat: no details, only number of infants completing the trial mentioned Interim analysis: none Exclusion from analysis: no details Follow‐up period: 1 week after each vaccination; 17 weeks after vaccination
Participants	Age: 2 to 11 months Health status: healthy Breastfeeding: < 50% Immunization status: DPT and polio vaccinations not allowed within 7 days of vaccination
Interventions	1. Bovine + human reassortant (WI79‐9), 2 x 10^7 PFU, 2 doses (n = 38) 2. Placebo (cell culture medium with 20% cherry syrup), 2 doses (n = 39) 30 ml formula or 85 mg antacid given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 7 days of each vaccination 2. Efficacy: gastroenteritis within 17 weeks of vaccination
Notes	Study location: USA Clinical symptoms: clinical evaluation; severity score measured by modified Duffy 1986 scale Laboratory studies: serology by plaque reduction neutralization test; stool analysis by latex agglutination test At least 7 days between DTP and RV

Methods	Randomization: random code assignment to 5 groups Allocation concealment: code prepared in advance, not disclosed to field team, broken at NIH laboratory after follow up ended Blinding: double (same size of ampoules given to each group). Data collection: no details Intention to treat: no Interim analysis: none Exclusion from analysis: 20\116 excluded from study for having diarrhoea, nausea, or vomiting in the week prior to vaccination. Follow‐up period: 1 week after vaccination
Participants	Age: 1 to 5 months Health status: healthy Breastfeeding: no information Immunization status: polio vaccine not given within 2 weeks of vaccination
Interventions	1. Rhesus + human (D x RRV), 10^4 PFU, single dose (n = 24) 2. Rhesus + human (DS1 x RRV), 10^4 PFU, single dose (n = 25) 3. Rhesus (RRV 3), 10^4 PFU, single dose (n = 21) 4. Rhesus + human (D x RRV + RRV), 5 x 10^3 PFU, single dose (n = 23) 5. Placebo (formula alone or with red soda), 1 ml single dose (n = 23) 400 mg citrate‐bicarbonate in 30 ml similac formula given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 1 week of vaccination
Notes	Study location: Venezuela Clinical symptoms: clinical evaluation; diarrhoea defined as ≥ 3 watery or loose stools in 24 h; severity score measured by Flores 1987 scale Laboratory studies: serology by ELISA and plaque reduction neutralization assay; stool analysis by ELISA and cultures

Methods	Randomization: random order assignment to 4 groups Allocation concealment: code not disclosed to field team until end of analysis, broken at NIH Blinding: double (no further details) Data collection: no details Intention to treat: yes Interim analysis: none Exclusion from analysis: none Follow‐up period: 1 week after vaccination
Participants	Age: 10 to 20 weeks Health status: no information Breastfeeding: no information Immunization status: no information
Interventions	1. Rhesus + human quadrivalent vaccine (3 RRV + D x RRV + DS1 x RRV + ST3 x RRV), 10^4 PFU, single dose (n = 23) 2. Rhesus + human quadrivalent vaccine (3 RRV + D x RRV + DS1 x RRV + ST3 x RRV), 5 x 10^4 PFU, single dose (n = 22) 3. Human vaccine (M37), 10^4 PFU, single dose (n = 23) 4. Placebo (formula), 1 ml, single dose (n= 22) 400 mg citrate‐bicarbonate in 30 ml similac formula given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 1 week of vaccination
Notes	Study location: Venezuela Clinical symptoms: clinical evaluation; diarrhoea defined as ≥ 3 watery or loose stools in 24 h Laboratory studies: serology by ELISA and plaque reduction neutralization assay; stool analysis by ELISA and cultures

Methods	Randomization: table of random numbers, administered in numeric order in the first dose Allocation concealment: no details Blinding: double (identical precoded vials) Data collection: no information Intention to treat: no Interim analysis: yes Exclusion from analysis: 69\539 did not receive vaccine (7 deaths), 56/470 did not finish follow up (4 deaths) Follow‐up period: 9 months after first vaccination
Participants	Age: 3 months Health status: healthy Breastfeeding: > 75% Immunization status: DTCP vaccine given with rotavirus vaccine
Interventions	1. Bovine (WC3), 10^7 PFU, 2 doses (n = 237) 2. Placebo (sacharose and lactose substrates), 2 doses (n = 235) Milk given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 1 week of vaccination 2. Efficacy: diarrhoea within 9 months of first vaccination
Notes	Country: Central African Republic Clinical symptoms: clinical evaluation; diarrhoea defined as ≥ 3 watery or loose stools in 24 h; severity score measured by Clark 1988 (modified Duffy 1986) scale Laboratory studies: stool analysis by latex agglutination and ELISA; serology by plaque reduction neutralization assay

Methods	Randomization: prepared randomization schedule Allocation concealment: no details Blinding: none Data collection: January 1989 to November 1989 Intention to treat: yes Interim analysis: none Exclusion from analysis: none Follow‐up period: 5 days after vaccination; 2 months after vaccination
Participants	Age: 1.5 to 4 months Health status: healthy
Interventions	1. Rhesus + human tetravalent vaccine (RRV‐TV), 4 x 10^5 PFU, with OPV/DTP vaccination, single dose (n = 135). 2. OPV/DTP alone (n = 39) 400 mg sodium bicarbonate in 25 ml soybean formula given before vaccine
Outcomes	1. Safety: clinical symptoms within 5 days of vaccination 2. Illness or diarrhoea within 2 months of vaccination
Notes	Study location: USA Clinical symptoms: clinical evaluation (diarrhoea defined as ≥ 3 loose to watery stools in 24 h) Laboratory studies: serology by ELISA

Methods	Randomization: computer‐generated randomization schedule, blocks of four, provided by the pharmaceutical industry Allocation concealment: code sent to investigator in sealed envelopes and returned to pharmaceutical industry unbroken at the end of the study Blinding: double (except 4th year of follow up), identically appearing vaccine and placebo Data collection: September 1993 to June 1997 Intention to treat: yes Interim analysis: none Exclusion from analysis: 38\2398 excluded, no further details Follow‐up period: 5 days after vaccination; 4 rotavirus seasons after vaccination
Participants	Age: 1.5 to 4.5 months Health status: healthy Breastfeeding: no information Immunization status: all other vaccinations given according to schedule without delays
Interventions	1. Rhesus + human tetravalent vaccine (RRV‐TV), 4 x 10^5 PFU, 3 doses (n = 1191) 2. Placebo (tissue culture fluid), 3 ml, 3 doses (n = 1207) 3 ml sodium bicarbonate and citric acid buffer given before vaccine or placebo
Outcomes	1. Safety: clinical symptoms within 5 days of vaccination. 2. Efficacy: gastroenteritis (rotavirus or other) within 4 rotavirus seasons
Notes	Study location: Finland Clinical symptoms: clinical evaluation (gastroenteritis defined as ≥ 3 looser than normal stools or ≥ 3 loose stools and one vomiting episode in 24 h); severity score measured by Ruuska 1990 scale Laboratory studies: stool analysis by ELISA and PCR; serology by ELISA