Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Vaccines for preventing rotavirus diarrhoea: vaccines in use

Authors' declarations of interest

Version published: 28 October 2019 Version history

https://doi.org/10.1002/14651858.CD008521.pub5

This is not the most recent version

view the current version 17 November 2021
Collapse all Expand all

Background

Rotavirus results in more diarrhoea‐related deaths in children under five years than any other single agent in countries with high childhood mortality. It is also a common cause of diarrhoea‐related hospital admissions in countries with low childhood mortality. Rotavirus vaccines that have been prequalified by the World Health Organization (WHO) include a monovalent vaccine (RV1; Rotarix, GlaxoSmithKline), a pentavalent vaccine (RV5; RotaTeq, Merck), and, more recently, another monovalent vaccine (Rotavac, Bharat Biotech).

Objectives

To evaluate rotavirus vaccines prequalified by the WHO (RV1, RV5, and Rotavac) for their efficacy and safety in children.

Search methods

On 4 April 2018 we searched MEDLINE (via PubMed), the Cochrane Infectious Diseases Group Specialized Register, CENTRAL (published in the Cochrane Library), Embase, LILACS, and BIOSIS. We also searched the WHO ICTRP, ClinicalTrials.gov, clinical trial reports from manufacturers' websites, and reference lists of included studies and relevant systematic reviews.

Selection criteria

We selected randomized controlled trials (RCTs) in children comparing rotavirus vaccines prequalified for use by the WHO versus placebo or no intervention.

Data collection and analysis

Two review authors independently assessed trial eligibility and assessed risks of bias. One review author extracted data and a second author cross‐checked them. We combined dichotomous data using the risk ratio (RR) and 95% confidence interval (CI). We stratified the analysis by country mortality rate and used GRADE to evaluate evidence certainty.

Main results

Fifty‐five trials met the inclusion criteria and enrolled a total of 216,480 participants. Thirty‐six trials (119,114 participants) assessed RV1, 15 trials (88,934 participants) RV5, and four trials (8432 participants) Rotavac.

RV1

Children vaccinated and followed up the first year of life

In low‐mortality countries, RV1 prevents 84% of severe rotavirus diarrhoea cases (RR 0.16, 95% CI 0.09 to 0.26; 43,779 participants, 7 trials; high‐certainty evidence), and probably prevents 41% of cases of severe all‐cause diarrhoea (RR 0.59, 95% CI 0.47 to 0.74; 28,051 participants, 3 trials; moderate‐certainty evidence). In high‐mortality countries, RV1 prevents 63% of severe rotavirus diarrhoea cases (RR 0.37, 95% CI 0.23 to 0.60; 6114 participants, 3 trials; high‐certainty evidence), and 27% of severe all‐cause diarrhoea cases (RR 0.73, 95% CI 0.56 to 0.95; 5639 participants, 2 trials; high‐certainty evidence).

Children vaccinated and followed up for two years

In low‐mortality countries, RV1 prevents 82% of severe rotavirus diarrhoea cases (RR 0.18, 95% CI 0.14 to 0.23; 36,002 participants, 9 trials; high‐certainty evidence), and probably prevents 37% of severe all‐cause diarrhoea episodes (rate ratio 0.63, 95% CI 0.56 to 0.71; 39,091 participants, 2 trials; moderate‐certainty evidence). In high‐mortality countries RV1 probably prevents 35% of severe rotavirus diarrhoea cases (RR 0.65, 95% CI 0.51 to 0.83; 13,768 participants, 2 trials; high‐certainty evidence), and 17% of severe all‐cause diarrhoea cases (RR 0.83, 95% CI 0.72 to 0.96; 2764 participants, 1 trial; moderate‐certainty evidence).

No increased risk of serious adverse events (SAE) was detected (RR 0.88 95% CI 0.83 to 0.93; high‐certainty evidence). There were 30 cases of intussusception reported in 53,032 children after RV1 vaccination and 28 cases in 44,214 children after placebo or no intervention (RR 0.70, 95% CI 0.46 to 1.05; low‐certainty evidence).

RV5

Children vaccinated and followed up the first year of life

In low‐mortality countries, RV5 probably prevents 92% of severe rotavirus diarrhoea cases (RR 0.08, 95% CI 0.03 to 0.22; 4132 participants, 5 trials; moderate‐certainty evidence). We did not identify studies reporting on severe all‐cause diarrhoea in low‐mortality countries. In high‐mortality countries, RV5 prevents 57% of severe rotavirus diarrhoea (RR 0.43, 95% CI 0.29 to 0.62; 5916 participants, 2 trials; high‐certainty evidence), but there is probably little or no difference between vaccine and placebo for severe all‐cause diarrhoea (RR 0.80, 95% CI 0.58 to 1.11; 1 trial, 4085 participants; moderate‐certainty evidence).

Children vaccinated and followed up for two years

In low‐mortality countries, RV5 prevents 82% of severe rotavirus diarrhoea cases (RR 0.18, 95% CI 0.08 to 0.39; 7318 participants, 4 trials; moderate‐certainty evidence). We did not identify studies reporting on severe all‐cause diarrhoea in low‐mortality countries. In high‐mortality countries, RV5 prevents 41% of severe rotavirus diarrhoea cases (RR 0.59, 95% CI 0.43 to 0.82; 5885 participants, 2 trials; high‐certainty evidence), and 15% of severe all‐cause diarrhoea cases (RR 0.85, 95% CI 0.75 to 0.98; 5977 participants, 2 trials; high‐certainty evidence).

No increased risk of serious adverse events (SAE) was detected (RR 0.93 95% CI 0.86 to 1.01; moderate to high‐certainty evidence). There were 16 cases of intussusception in 43,629 children after RV5 vaccination and 20 cases in 41,866 children after placebo (RR 0.77, 95% CI 0.41 to 1.45; low‐certainty evidence).

Rotavac

Children vaccinated and followed up the first year of life

Rotavac has not been assessed in any RCT in countries with low child mortality. In India, a high‐mortality country, Rotavac probably prevents 57% of severe rotavirus diarrhoea cases (RR 0.43, 95% CI 0.30 to 0.60; 6799 participants, moderate‐certainty evidence); the trial did not report on severe all‐cause diarrhoea at one‐year follow‐up.

Children vaccinated and followed up for two years

Rotavac probably prevents 54% of severe rotavirus diarrhoea cases in India (RR 0.46, 95% CI 0.35 to 0.60; 6541 participants, 1 trial; moderate‐certainty evidence), and 16% of severe all‐cause diarrhoea cases (RR 0.84, 95% CI 0.71 to 0.98; 6799 participants, 1 trial; moderate‐certainty evidence).

No increased risk of serious adverse events (SAE) was detected (RR 0.93 95% CI 0.85 to 1.02; moderate‐certainty evidence). There were eight cases of intussusception in 5764 children after Rotavac vaccination and three cases in 2818 children after placebo (RR 1.33, 95% CI 0.35 to 5.02; very low‐certainty evidence).

There was insufficient evidence of an effect on mortality from any rotavirus vaccine (198,381 participants, 44 trials; low‐ to very low‐certainty evidence), as the trials were not powered to detect an effect at this endpoint.

Authors' conclusions

RV1, RV5, and Rotavac prevent episodes of rotavirus diarrhoea. Whilst the relative effect estimate is smaller in high‐mortality than in low‐mortality countries, there is a greater number of episodes prevented in these settings as the baseline risk is much higher. We found no increased risk of serious adverse events.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Vaccines for preventing rotavirus diarrhoea: vaccines in use

What is the aim of this review?

The aim of this Cochrane Review was to find out if rotavirus vaccines are effective in preventing diarrhoea and deaths in infants and young children. We also aimed to find out if the rotavirus vaccines are safe. We collected and analyzed all relevant studies to answer these questions, and found 55 studies.

Key messages

RV1, RV5, and Rotavac prevent episodes of rotavirus diarrhoea (moderate‐ to high‐certainty evidence). We found no increased risk of serious adverse events (moderate‐ to high‐certainty evidence) including intussusception (where the bowel telescopes on itself, and can cause obstruction) (very low to low‐certainty evidence).

What was studied in the review?

Rotavirus infection is a common cause of diarrhoea in infants and young children, and can cause mild illness, hospitalization, and death. Since 2009, the World Health Organization (WHO) has recommended that a rotavirus vaccine be included in all national infant and child immunization programmes, and 95 countries have so far followed this recommendation. In the years before infants and children started receiving rotavirus vaccine, rotavirus infection resulted in about half a million deaths a year in children aged under five years, mainly in low‐ and middle‐income countries.

In this review we included randomized controlled trials in infants and young children that evaluated a monovalent rotavirus vaccine (RV1; Rotarix, GlaxoSmithKline) or a pentavalent rotavirus vaccine (RV5; RotaTeq, Merck). These vaccines have been evaluated in several large trials and are approved for use in many countries. We also included trials that evaluated another monovalent rotavirus vaccine (Rotavac; Bharat Biotech), which is used in India only. The rotavirus vaccines were compared with placebo or with no vaccine. The included studies did not allow comparisons between the vaccines.

What are the main results of the review?

We found 55 relevant studies with 216,480 participants. The trials took place in several locations worldwide. These studies compared a rotavirus vaccine versus placebo or versus no vaccine for infants and young children. The vaccines tested were RV1 (36 trials with 119,114 participants), RV5 (15 trials with 88,934 participants), and Rotavac (four trials with 8432 participants). Fifty‐one studies were funded or co‐funded by vaccine manufacturers, while four were independent of manufacturer funding.

In the first two years of life, RV1:

●prevents more than 80% of severe cases of rotavirus diarrhoea in countries with low death rates (high‐certainty evidence)
●prevents 35% to 63% of severe rotavirus diarrhoea in countries with high death rates (high‐certainty evidence)
●probably prevents 37% to 41% of severe cases of diarrhoea from all causes (such as any viral infection, bacterial infection, or parasitic infection) in countries with low death rates (moderate‐certainty evidence)
●probably prevents 18% to 27% of severe cases of diarrhoea from all causes in countries with high death rates (moderate‐ to high‐certainty evidence).

In the first two years of life, RV5:

●probably prevents 82% to 92% of severe cases of rotavirus diarrhoea in countries with low death rates (moderate‐certainty evidence)
●prevents 41% to 57% of severe cases of rotavirus diarrhoea in countries with high death rates (high‐certainty evidence)
●probably prevents 15% of severe cases of diarrhoea from all causes in countries with high death rates (moderate‐ to high‐certainty evidence); we did not identify any studies that reported on diarrhoea from all causes in countries with low death rates.

In the first two years of life, Rotavac:

●probably prevents more than 50% of severe cases of rotavirus diarrhoea in India, a country with high death rates (moderate‐certainty evidence)
●probably prevents 18% of severe cases of diarrhoea from all causes in India (moderate‐certainty evidence). Rotavac has not been evaluated in a randomized controlled trial in a country with low death rates.

We found little or no difference in the number of serious adverse events (moderate‐ to high‐certainty evidence), or intussusception cases (low‐ to very low‐certainty evidence), between those receiving RV1, RV5, or Rotavac compared with placebo or no intervention.

How up‐to‐date is this review?

We searched for studies that had been published up to 4 April 2018.

Authors' conclusions

Implications for practice

  • RV1, RV5, and Rotavac are efficacious vaccines in preventing rotavirus diarrhoea. RV1 and RV5 have comparable safety and efficacy profiles and have been evaluated in different settings worldwide. The evidence on Rotavac is currently limited to studies from India. The systematic review data support the global WHO rotavirus vaccine recommendation (SAGE 2009; SAGE 2012).

  • The data from the included RCTs exclude a risk of intussusception with RV1, RV5, and Rotavac of the magnitude observed with the first licensed vaccine (RRV‐TV, RotaShield). However, since the data cannot exclude a smaller risk of intussusception or other rare serious adverse events, routine vaccine introduction should be accompanied by safety surveillance (Buttery 2011; Patel 2011; Shui 2012; Weintraub 2014).

Implications for research

Placebo‐controlled efficacy trials of RV1 and RV5 have been undertaken in representative populations of low‐ and high‐mortality countries and do not require repetition; efficacy or effectiveness trials of Rotavac outside of India should be considered if Rotavac is introduced globally. Further research would be valuable in the following areas:

  • Continued post‐introduction studies to examine the impact and effectiveness of rotavirus vaccination, particularly in high‐mortality countries.

  • A greater understanding of the lower vaccine efficacy observed in high‐mortality countries compared to low‐mortality countries in Africa and Asia in the first and second years of life.

  • Studies to assess the potential benefit of alternative dosage schedules of rotavirus vaccine, especially in high‐mortality countries (e.g. neonatal dosing, additional dosing).

  • Continued post‐introduction studies in representative countries should examine vaccine safety with particular respect to intussusception and should analyze the risk/benefit of rotavirus vaccination (Patel 2011). Post‐introduction safety studies of Rotavac are currently lacking (Dutta 2017). Given the rarity of the event, data from different countries may need to be pooled (Escolano 2011; Escolano 2015), or self‐controlled case series analyses may need to be carried out (Carlin 2013; Stowe 2016; Tate 2018; Yih 2014).

Summary of findings

Open in table viewer
Summary of findings for the main comparison. RV1 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries

Patient or population: children
Setting: low‐mortality countries (WHO strata A and B)
Intervention: RV1
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV1

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

13 per 1000

2 per 1000
(1 to 3)

RR 0.16
(0.09 to 0.26)

43,779
(7 studies)

⊕⊕⊕⊕
higha

RV1 reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

One study (RV1 Vesikari 2007a‐EU) reported higher efficacy compared to the pooled data. When we excluded this study from the analysis, there was no heterogeneity observed in the pooled data

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

24 per 1000

4 per 1000
(3 to 5)

RR 0.18
(0.14 to 0.23)

36,002
(9 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

41 per 1000

24 per 1000
(19 to 30)

RR 0.59
(0.47 to 0.74)

28,051
(3 studies)

⊕⊕⊕⊝
moderateb

due to reporting bias

RV1 probably reduces severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

Severe episodes of all‐cause diarrhoea
Follow‐up: up to 2 years

39 per 1000

24 per 1000
(22 to 28)

Rate Ratio 0.63
(0.56 to 0.71)

39,091
(2 studies)

⊕⊕⊕⊝
moderatec

due to reporting bias

RV1 probably reduces severe all‐cause diarrhoea compared to placebo at up to two years follow‐up.

Three additional studies reported on cases of children with severe all‐cause diarrhoea (RR 0.60, 95% CI 0.36 to 1.02; 9417 participants); these data could not be pooled with the studies reporting on number of episodes

All‐cause death
Follow‐up: 2 months to 2 years

1 per 1000

2 per 1000
(1 to 2)

RR 1.22
(0.87 to 1.71)

97,597
(22 studies)

⊕⊕⊝⊝
lowd

due to imprecision

RV1 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

45 per 1000

40 per 1000
(37 to 42)

RR 0.88
(0.83 to 0.93)

96,233
(24 studies)

⊕⊕⊕⊕
high

RV1 slightly reduces serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

1 per 1000

1 per 1000
(0 to 1)

RR 0.69
(0.45 to 1.04)

96,513
(17 studies)

⊕⊕⊝⊝
lowe

due to imprecision

RV1 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aWe observed heterogeneity (I2 statistic = 61%) in the pooled data, but given the strength of the evidence, and that estimates were all in the same direction, we did not downgrade the outcome.
bDowngraded by one for risk of selective reporting bias. Only three of the seven studies reporting on severe rotavirus diarrhoea provided data for this outcome.
cDowngraded by one for risk of selective reporting bias. Only five of the nine studies reporting on severe rotavirus diarrhoea provided data for this outcome.
dDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
eDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Open in table viewer
Summary of findings 2. RV1 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children
Settings: high‐mortality countries (WHO strata D and E)
Intervention: RV1
Comparison: placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo or no intervention

RV1

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

60 per 1000

22 per 1000
(14 to 36)

RR 0.37
(0.23 to 0.60)

6114
(3 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo or no intervention at up to one year follow‐up.

We did not downgrade for inconsistency as the heterogeneity observed in the pooled data (I2 statistic = 57%) was due to within‐study heterogeneity (RV1 Madhi 2010‐AF results split by country)

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

43 per 1000

28 per 1000
(22 to 35)

RR 0.65
(0.51 to 0.83)

13,768**
(2 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo or no intervention at up to two years follow‐up.

Sensitivity analysis excluding the cluster‐RCT (RV1 Zaman 2017‐BGD) that contributed data to this outcome showed no significant change in effect estimate or 95% CI (RR 0.58, 95% CI 0.42 to 0.79, n = 2764, 1 RCT)

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

176 per 1000

129 per 1000
(99 to 167)

RR 0.73
(0.56 to 0.95)

5639
(2 studies)

⊕⊕⊕⊕
high

RV1 reduces severe all‐cause diarrhoea compared to placebo or no intervention at up to one year follow‐up.

We did not downgrade for inconsistency as the heterogeneity observed in the pooled data (I2 statistic = 75%) was due to within‐study heterogeneity (RV1 Madhi 2010‐AF results split by country)

Severe cases of all‐cause diarrhoea
Follow‐up: up to 2 years

233 per 1000

191 per 1000
(166 to 222)

RR 0.82
(0.71 to 0.95)

2764
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

RV1 probably slightly reduces severe all‐cause diarrhoea compared to placebo or no intervention at up to two years follow‐up.

All‐cause death
Follow‐up: 2 months to 2 years

24 per 1000

21 per 1000
(16 to 30)

RR 0.88
(0.64 to 1.22)

8181
(8 studies)

⊕⊕⊝⊝
lowb

due to imprecision

RV1 may make little or no difference to all‐cause death compared to placebo or no intervention.

All serious adverse events
Follow‐up: 2 months to 2 years

95 per 1000

84 per 1000
(72 to 99)

RR 0.89
(0.76 to 1.04)

7481
(7 studies)

⊕⊕⊕⊕
high

RV1 makes little or no difference to serious adverse events compared to placebo or no intervention.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

0 per 100,000

0 per 100,000
(0 to 0)

RR 1.49
(0.06 to 36.63)

17,492**
(4 studies)

⊕⊕⊝⊝
lowc

due to imprecision

RV1 may make little or no difference to intussusception compared to placebo or no intervention.

Sensitivity analysis excluding the cluster‐RCT (RV1 Zaman 2017‐BGD) that contributed data to this outcome showed no change in effect estimate or 95% CI

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**Number of participants in this table shows the true number of participants for this outcome; the number of events and the number of participants in the analysis has been adjusted for the included cluster trial RV1 Zaman 2017‐BGD using a design effect of 2.53.
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for indirectness. Trials were conducted in Malawi and South Africa, so generalization to any high‐mortality country is difficult.
bDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
cDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Open in table viewer
Summary of findings 3. RV5 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries

Patient or population: children
Settings: low‐mortality countries (WHO strata A and B)
Intervention: RV5
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV5

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

17 per 1000

1 per 1000
(1 to 5)

RR 0.08
(0.03 to 0.22)

4132
(5 studies)

⊕⊕⊕⊝
moderatea

due to imprecision

RV5 probably reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

25 per 1000

4 per 1000
(2 to 10)

RR 0.18
(0.08 to 0.39)

7318
(4 studies)

⊕⊕⊕⊝
moderateb

due to inconsistency

RV5 probably reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe all‐cause diarrhoea
Follow‐up: up to 1 year

‐

‐

‐

‐

‐

We found no studies that reported on this outcome in this setting

Severe all‐cause diarrhoea
Follow‐up: up to 2 years

‐

‐

‐

‐

‐

We found no studies that reported on this outcome in this setting

All‐cause death
Follow‐up: 2 months to 2 years

1 per 1000

1 per 1000
(0 to 1)

RR 1.13
(0.65 to 1.96)

77,642
(9 studies)

⊕⊕⊝⊝
lowc

due to imprecision

RV5 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

27 per 1000

25 per 1000
(23 to 28)

RR 0.93
(0.86 to 1.02)

75,672
(8 studies)

⊕⊕⊕⊕
high

RV5 makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

1 per 1000

0 per 1000
(0 to 1)

RR 0.77
(0.41 to 1.45)

78,907
(12 studies)

⊕⊕⊝⊝
lowd

due to imprecision

RV5 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for imprecision. The total number of events was very low.
bDowngraded by one for inconsistency. We found substantial heterogeneity (I2 statistic = 44%). Consistency was restored when removing the one study carried out only in a very low‐mortality (stratum A) country, with results then showing a slightly smaller effect (RR 0.22, 95% CI 0.13 to 0.36, 6291 participants, 3 studies).
cDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Open in table viewer
Summary of findings 4. RV5 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children
Settings: high‐mortality countries (WHO strata D and E)
Intervention: RV5
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV5

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

30 per 1000

13 per 1000
(9 to 19)

RR 0.43
(0.29 to 0.62)

5916
(2 studies)

⊕⊕⊕⊕
high

RV5 reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

63 per 1000

37 per 1000
(27 to 51)

RR 0.59
(0.43 to 0.82)

5885
(2 studies)

⊕⊕⊕⊕
high

RV5 reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

77 per 1000

62 per 1000
(45 to 85)

RR 0.8
(0.58 to 1.11)

4085
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

RV5 probably makes little or no difference to severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 2 years

130 per 1000

110 per 1000
(97 to 127)

RR 0.85
(0.75 to 0.98)

5977
(2 studies)

⊕⊕⊕⊕
high

RV5 slightly reduces severe all‐cause diarrhoea compared to placebo at up to two years follow‐up.

All‐cause death
Follow‐up: 2 months to 2 years

26 per 1000

23 per 1000
(17 to 32)

RR 0.92
(0.68 to 1.24)

6806
(3 studies)

⊕⊕⊝⊝
lowb

due to imprecision

RV5 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

21 per 1000

19 per 1000
(14 to 27)

RR 0.92
(0.66 to 1.28)

6830
(4 studies)

⊕⊕⊕⊝
moderatec

due to imprecision

RV5 probably makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

See comment

See comment

Not estimable

6588
(2 studies)

⊕⊕⊝⊝
lowd

due to imprecision

No events were reported. RV5 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for indirectness. Single trial conducted in three African countries (Mali, Ghana, and Kenya), so generalization to any high‐mortality country is difficult.
bDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
cDowngraded by one for imprecision. The 95% CI includes both no effect and appreciable harm.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Open in table viewer
Summary of findings 5. Rotavac compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children

Settings: one high‐mortality country (India) (WHO stratum D)

Intervention: Rotavac

Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of Participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

Rotavac

Severe cases of rotavirus diarrhoea

follow‐up: up to 1 year

31 per 1000

13 per 1000
(9 to 19)

RR 0.43
(0.30 to 0.60)

6799
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea follow‐up: up to 2 years

47 per 1000

21 per 1000
(16 to 28)

RR 0.46
(0.35 to 0.60)

6541
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea

follow‐up: up to 2 years

93 per 1000

78 per 1000
(66 to 91)

RR 0.84
(0.71 to 0.98)

6799
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably slightly reduces severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

All‐cause death

follow‐up: up to 2 years

7 per 1000

6 per 1000
(4 to 11)

RR 0.92
(0.52 to 1.62)

8155
(2 studies)

⊕⊝⊝⊝
very lowb,c

due to indirectness and imprecision

We are uncertain whether Rotavac reduced all‐cause death as the certainty of the evidence is very low.

All serious adverse events

follow‐up: up to 2 years

204 per 1000

189 per 1000
(173 to 208)

RR 0.93
(0.85 to 1.02)

8210
(3 studies)

⊕⊕⊕⊝
moderateb

due to indirectness

Rotavac probably makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception

follow‐up: up to 2 years

1 per 1000

1 per 1000
(0 to 5)

RR 1.33
(0.35 to 5.02)

8582
(4 studies)

⊕⊝⊝⊝
very lowb,d

due to indirectness and imprecision

No events were reported in three of the four studies. We are uncertain whether Rotavac has an effect on intussusception as the certainty of the evidence is very low.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk Ratio

GRADE Working Group grades of evidence
High certainty: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate certainty: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low certainty: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low certainty: We are very uncertain about the estimate.

aDowngraded by one for indirectness. Single trial conducted in India, so generalization to any high‐mortality country is difficult.
bDowngraded by one for indirectness. All trials were conducted in India, so generalization to any high‐mortality country is difficult.
cDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), therefore, these trials were not powered to detect an association between Rotavac and intussusception.

Background

Description of the condition

The global impact of rotavirus infection

Rotavirus is the leading known cause of severe gastroenteritis in infants and young children worldwide (Parashar 2006a; Vesikari 1997; WHO 2013). While nearly every child experiences at least one rotavirus infection in early childhood regardless of setting, the vast majority of rotavirus‐associated deaths occur in children in low‐ and middle‐income countries, particularly in sub‐Saharan Africa and in the Indian subcontinent. Prior to the rollout of rotavirus vaccination, rotavirus caused 37% of diarrhoeal deaths (˜ 450,000 deaths worldwide in 2008) in children younger than five years. Five countries accounted for more than half of all deaths, and 22% of deaths attributable to rotavirus infection occurred in India (Tate 2012). In high‐income countries, where deaths due to rotavirus are rare, rotavirus accounted for 40% to 50% of hospital admissions due to diarrhoeal disease in the pre‐rotavirus vaccine period (Linhares 2008; Parashar 2006a; Tate 2012).

Epidemiology of rotavirus infection

Rotavirus is transmitted primarily via the faecal‐oral route, with symptoms typically developing one to two days following infection. Rotavirus infection occurs throughout life, and successive rotavirus infections occur during infancy and early childhood. The first rotavirus infection typically results in the most severe disease outcome; subsequent rotavirus infections are associated with milder disease or may be asymptomatic. However, differences in the age of first infection and number of infections required to acquire protection from symptomatic disease vary from one population to another. Rotavirus diarrhoea is particularly associated with severe outcomes between the ages of three and 35 months (Parashar 2006b), with a peak incidence of all episodes occurring between six and 24 months (CDC‐ASIP 1999; Linhares 2008). The peak incidence of severe rotavirus disease occurs earlier in high‐mortality countries than in low‐mortality countries; an estimated 43% of all rotavirus hospitalizations in children aged under five occur by eight months of age in Africa compared with 27% in Europe (Crawford 2017; Sanderson 2011). Typically, infants in low‐income countries experience a greater number of symptomatic episodes (Gladstone 2011; Velázquez 1996). In temperate countries rotavirus infections display marked seasonality, with distinct peaks during the winter months and few infections identified outside this period, whereas rotavirus infections occur year‐round in most tropical countries.

Rotavirus classification

Rotaviruses are double‐stranded (ds) RNA viruses: genus Rotavirus, family Reoviridae. Each of the 11 dsRNA segments, contained within the core of a triple‐layered viral particle, encodes one or more viral proteins. Rotavirus A, which causes most human disease, is genetically diverse in each of its 11 genome segments (called genotypes), and a nucleotide sequence‐based, complete genome classification system is used. Because of their importance in protective immunity, the outer capsid proteins VP7 and VP4 have been most extensively investigated. Species A rotaviruses are classified into G and P genotypes, based on the sequence diversity of the RNA segments encoding VP7 and VP4, respectively; 32 G genotypes and 47 P genotypes have been described (Crawford 2017) (see Figure 1 for details). Rotavirus vaccines are designed to protect against disease caused by the most prevalent strain types; globally, G1P[8], G2P[4], G3P[8], G4P[8], G9P[8] and G12 in combination with P[6] or P[8] account for over 90% of the genotypes that infect humans (Bányai 2012).

Figure 1
A simplified diagram of the location of rotavirus structural proteins (source: Graham Cohn, Wikipedia (public domain image)): Rotaviruses are segmented, double‐stranded RNA viruses. The mature, triple‐layered virus particle comprises a core (which contains the viral genome), a middle layer (comprised of viral protein (VP)6, and an outer layer (comprised of VP7 and VP4) as shown in the figure. VP6 defines rotavirus group, and most rotaviruses that infect humans are of group A. The two outer capsid proteins independently induce neutralizing antibodies: VP7, a glycoprotein, defines G‐serotype; and the protease‐sensitive VP4 protein defines P‐serotype. G‐serotype determined by serological methods correlates precisely with G‐genotype obtained through molecular assays, whereas there is an imperfect correlation of P‐serotype and P‐genotype; P‐genotype is thus included in square brackets.

A simplified diagram of the location of rotavirus structural proteins (source: Graham Cohn, Wikipedia (public domain image)): Rotaviruses are segmented, double‐stranded RNA viruses. The mature, triple‐layered virus particle comprises a core (which contains the viral genome), a middle layer (comprised of viral protein (VP)6, and an outer layer (comprised of VP7 and VP4) as shown in the figure. VP6 defines rotavirus group, and most rotaviruses that infect humans are of group A. The two outer capsid proteins independently induce neutralizing antibodies: VP7, a glycoprotein, defines G‐serotype; and the protease‐sensitive VP4 protein defines P‐serotype. G‐serotype determined by serological methods correlates precisely with G‐genotype obtained through molecular assays, whereas there is an imperfect correlation of P‐serotype and P‐genotype; P‐genotype is thus included in square brackets.

Description of the intervention

Vaccines approved for use

This review evaluates three vaccines, including a monovalent rotavirus vaccine (RV1; Rotarix, GlaxoSmithKline Biologicals) and a pentavalent rotavirus vaccine (RV5; RotaTeq, Merck & Co., Inc.), which have been evaluated in several large trials and are in routine use in many countries; and a further monovalent vaccine (Rotavac, Bharat Biotech Ltd.), which is currently licensed in India only. All three vaccines are listed as prequalified vaccines by the WHO (Dellepiane 2015; WHO 2018). As of April 2018, 95 countries have introduced rotavirus vaccines into their immunization programmes (ROTA council 2018).

RV1 is an oral, live‐attenuated, human rotavirus vaccine derived from the most common circulating wild‐type strain G1P[8]. RV1 is based on a rotavirus of entirely human origin and is administered to infants in two oral doses with an interval of at least four weeks between doses. The manufacturer states that the "vaccination course should preferably be given before 16 weeks of age, but must be completed by the age of 24 weeks" (EMA 2011). As of May 2016, RV1 had been introduced in national immunization programmes in 63 countries around the world (PATH 2016).

RV5 is an oral, live, human‐bovine, reassortant, multivalent rotavirus vaccine developed from an original Wistar calf 3 (WC3) strain of bovine rotavirus. The vaccine contains five live, human‐bovine reassortant rotavirus strains. Four reassortant rotavirus strains each express one of the common human VP7 (G) types including G1, G2, G3, and G4, and the fifth reassortant expresses the common human VP4 (P) type P[8]. The three‐dose liquid vaccine is intended for infants aged between six and 32 weeks, with the first dose given at six to 12 weeks and subsequent doses administered at four‐ to 10‐week intervals; however, the third dose should not be given after 32 weeks of age (Merck 2008). As of May 2016, RV5 had been introduced in national immunization programmes in 22 countries around the world (PATH 2016).

Rotavac is a live‐attenuated, monovalent vaccine derived from a naturally‐occurring reassortant G9P[11] strain [116E] isolated from a newborn child in India (Yen 2014). This oral vaccine was developed by Bharat Biotech Ltd. in India and was licensed in India in 2014 (VAC Chandola 2017‐IND). Three doses are recommended, to be administered at 6, 10, and 14 weeks of age.

There are a further three rotavirus vaccines that have been licensed and approved for use in individual countries, but are not yet prequalified by the WHO. Lanzhou lamb rotavirus vaccine (LLR; Lanzhou Institute of Biomedical Products) which is licensed and used in China; a bovine rotavirus pentavalent vaccine (BRV‐PV, Rotasiil, Serum Institute of India Ltd.) which is licensed and used in India; and a monovalent vaccine (Rotavin‐M1, POLYVAC) which is licensed and used in Vietnam.

Vaccines no longer in use

Several vaccines, including the first licensed rotavirus vaccine (RRV‐TV; RotaShield, Wyeth Laboratories) were developed, tested in trials, and later abandoned or withdrawn from use. These vaccines are covered in a separate Cochrane Review (Soares‐Weiser 2004). RRV‐TV, a tetravalent rhesus‐human reassortant vaccine, was withdrawn from use in 1999 following reports of intussusception (bowel obstruction which occurs when one segment of bowel becomes enfolded within another segment). Evaluations have since suggested that the risk of intussusception was age‐related, with 80% of intussusception cases occurring in infants who were more than 90 days old when the first vaccine dose was administered (Simonsen 2005). Although it is still currently licensed, this vaccine is no longer in clinical use (Dennehy 2008).

How the intervention might work

Recommendations for rotavirus vaccine use

Vaccination with RV1 and RV5 was first recommended in 2006 in Europe and the Americas, where clinical trials had demonstrated vaccine efficacy of 85% to 100% (RV1 Ruiz‐Palac 06‐LA/EU; RV5 Vesikari 2006b‐INT). In April 2009, following clinical trials of RV1 and RV5 in low‐ and middle‐income countries in Africa and Asia, the WHO Strategic Advisory Group of Experts (SAGE) on Immunization recommended "the inclusion of rotavirus vaccination of infants into all national immunization programmes", with a stronger recommendation for countries where "diarrhoeal deaths account for ≥10% of mortality among children aged <5 years" (SAGE 2009). Due to an age‐related risk of intussusception identified with RRV‐TV (Murphy 2001), SAGE recommended administering the first dose of RV1 or RV5 to infants of six to 15 weeks of age, with the last dose administered before 32 weeks of age (SAGE 2009). In April 2012, SAGE relaxed the age restricted recommendation and advised to vaccinate "as soon as possible after the age of six weeks" because "the current age restrictions for the first dose (< 15 weeks) and last dose (< 32 weeks) are preventing vaccination of many vulnerable children" (Patel 2012; SAGE 2012).

Performance of oral rotavirus vaccines by setting

Many oral vaccines, including rotavirus vaccines, have demonstrated lower immunogenicity and efficacy in low‐ and middle‐income countries in Africa and Asia compared to high‐income countries in North America, South America, and Europe (Levine 2010). A systematic review demonstrated a correlation between lower vaccine efficacy against severe rotavirus diarrhoea and high child mortality rates (Fischer Walker 2011). The reasons for reduced oral vaccine efficacy in countries with higher child mortality rates are unknown; factors may include interference by maternal antibody, co‐administration with oral poliovirus vaccine, histoblood group antigen, diverse rotavirus strain types, micronutrient deficiencies, endemic infections such as malaria, tuberculosis, or HIV, concomitant enteric infections, gut inflammation, and altered gut microbiota (Czerkinsky 2015).

Outcomes of interest

The safety and efficacy of the licensed vaccines for the prevention of rotavirus gastroenteritis in infants have been assessed in several randomized controlled trials (RCTs) worldwide. The goal of this review is to systematically assess these trials and evaluate vaccine efficacy against rotavirus diarrhoea, all‐cause diarrhoea, and diarrhoea‐related medical visits and hospitalization. We also examine the occurrence of deaths and serious adverse events, including intussusception, to provide decision‐makers, clinicians, and caregivers with the relevant information to aid decisions about vaccine use.

Why it is important to do this review

Development of Cochrane systematic rotavirus vaccine reviews

The original Cochrane Review of rotavirus vaccines (Soares‐Weiser 2004) examined vaccines in use and other vaccines, including those that were no longer in use or were in development. Soares‐Weiser 2004 concluded that more trials were needed before routine vaccine use could be recommended. An update in 2009 included a new search, revised inclusion criteria (only vaccines in use in children), updated review methods and new authors. The review was updated again in 2010 with nine new studies (Soares‐Weiser 2010). The 2010 version of the review concluded that RV1 and RV5 are both effective vaccines for the prevention of rotavirus diarrhoea. Another update in February 2012 added a further nine new studies, GRADE ‘Summary of findings' tables and, again, new authors joined the team (Soares‐Weiser 2012a). The November 2012 update included a new search, major restructuring of analyses, including re‐evaluating primary outcomes in consultation with the WHO to reflect the observation that vaccine efficacy profiles are different in countries with different mortality rates (Soares‐Weiser 2012b). This current update adds a further 10 RV1 and RV5 studies to the review and four studies of a new vaccine, Rotavac, that has been prequalified by the WHO since the previous version of the review.

Objectives

To evaluate rotavirus vaccines prequalified by the WHO (RV1, RV5, and Rotavac) for their efficacy and safety in children.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled trials (RCTs).

Types of participants

Children (age as defined in the trials).

Types of interventions

Intervention

Rotavirus vaccines approved by the WHO vaccine prequalification programme (Dellepiane 2015; WHO 2018).

Control

Placebo, no vaccination, or other vaccine.

Types of outcome measures

Primary

We selected our primary outcome measures in consultation with the WHO, and stratified them according to high‐ or low‐mortality rate, based on WHO mortality strata (WHO 1999), and up to one and up to two years follow‐up.

  • Rotavirus diarrhoea: severe (as defined in trial report)

  • All‐cause diarrhoea: severe

  • All‐cause death

  • Serious adverse events (that are fatal, life‐threatening, or result in hospitalization); e.g. Kawasaki disease

  • Intussusception

Secondary

  • Rotavirus diarrhoea: of any severity

  • All‐cause diarrhoea (as defined in trial report)

  • Rotavirus diarrhoea: requiring hospitalization

  • All‐cause diarrhoea: requiring hospitalization

  • Emergency department visit

  • Hospital admission: all‐cause

  • Reactogenicity (capacity to produce an adverse reaction, such as fever, diarrhoea, and vomiting)

  • Adverse events that require discontinuation of vaccination schedule

Other

  • Immunogenicity

    • Vaccine virus shedding in stool

    • Seroconversion: conversion from seronegative to seropositive for anti‐rotavirus IgA antibodies

  • Dropouts

Search methods for identification of studies

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

For this review update, Dr Vittoria Lutje (Information Specialist, Cochrane Infectious Diseases Group) searched the following databases using the search terms and strategy described in Appendix 1.

  • Cochrane Infectious Diseases Group Specialized Register (4 April 2018)

  • Cochrane Central Register of Controlled Trials (CENTRAL), published in the Cochrane Library (2018, Issue 4)

  • MEDLINE (via PubMed; 1966 to April 2018)

  • Embase (1974 to 4 April 2018)

  • LILACS (1982 to 4 April 2018)

  • BIOSIS (1926 to 4 April 2018)

We also searched the WHO International Clinical Trials Registry Platform (ICTRP) and Clinicaltrials.gov Clinical Study Register (www.clinicaltrials.gov) on 4 April 2018, using ‘rotavirus' as the search term.

We searched manufacturers' websites for clinical trial reports. We also checked the reference lists of relevant systematic reviews and included studies.

Data collection and analysis

Selection of studies

For this review update, we uploaded and screened references in DistillerSR online. Two review authors independently screened each title and abstract identified in the search. We retrieved full texts for potentially relevant references and two review authors again screened them independently, resolving disagreements by recourse to a third review author. We tabulated the excluded studies along with the reason for excluding them in the Characteristics of excluded studies tables. We ensured that data from each trial were entered only once in our review. In previous versions of this review we had screened references in an EndNote database.

Data extraction and management

For this review update, we extracted data in DistillerSR online. We created forms for data collection, which were piloted and then revised after the review author team's discussion. For previous versions of this review we had used Microsoft Word or Excel data collection forms.

One review author extracted data and another review author cross‐checked them. All outcomes were dichotomous, and we extracted the total number of participants and the number of participants who experienced the event. We cross‐checked the extracted data to identify errors, resolving disagreements by referring to the trial report or by consulting a third review author. One review author entered data into Review Manager 5 (RevMan 5) (RevMan 2014).

The use and mentioning of trade names in this review represents no endorsement of or advertisement for any product. The use of trade names was unavoidable as no generic names were identified for some of the vaccines evaluated here.

Assessment of risk of bias in included studies

Two review authors independently assessed the risks of bias of each trial, using the Cochrane ‘Risk of bias' tool (Higgins 2017). Based on the guidance of the Cochrane ‘Risk of bias' tool (Higgins 2017), we created a form to make judgements on the risk of bias for the rotavirus diarrhoea outcome measure in six domains: sequence generation; allocation concealment; blinding (of participants, personnel, and outcome assessors); incomplete outcome data; selective outcome reporting; and other potential sources of bias. We categorized these judgements as ‘low', ‘high', or ‘unclear' risk of bias. We resolved disagreements through discussion with a third review author.

For the 2012 published version of this review, we asked for help from Dr Ana Maria Restrepo at the WHO Initiative for Vaccine Research, who contacted the vaccine manufacturers GlaxoSmithKline (RV1) and Merck (RV5), who were involved in designing and funding most of the included trials. We provided them with an Excel spreadsheet with specific details of each trial that would impact on the assessment of risk of bias. We received details from Merck (RV5), (see Characteristics of included studies for details). For this review update, we matched most of the previously‐included RV1 studies to the full clinical trial reports available on the manufacturer's website (www.gsk‐clinicalstudyregister.com). More details were available in these trial reports than in the published studies, that were helpful in assessing the risks of bias for these studies.

Measures of treatment effect

We analyzed dichotomous data of cases by calculating the risk ratio (RR) for each trial (expressed using blue squares in forest plots) with the uncertainty in each result expressed using 95% confidence intervals (CIs). For dichotomous data of events that could occur more than once in one participant, we calculated the rate ratio (expressed using red squares in forest plots) on the logarithmic scale using the generic inverse variance method (see Data synthesis for more details). For outcomes that included cluster‐RCTs we calculated risk ratios (expressed using red squares in forest plots) using the generic inverse variance method (see Unit of analysis issues for more details).

Unit of analysis issues

When trials had multiple treatment arms and we considered it suitable, we grouped the trial arms. We excluded irrelevant trial arms.

We pooled cluster‐RCT data that had been adjusted for clustering with data from trials that randomly assigned individuals (individual‐RCTs). For outcomes that included cluster‐RCTs, we pooled risk ratios on the logarithmic scale with their standard errors using the generic inverse variance method (16.3.3. in Higgins 2011). When the results of a cluster‐RCT had not been adjusted for clustering, we imputed the clustering effect (intracluster correlation coefficient (ICC)) from another study, and performed sensitivity analyses excluding these studies.

Dealing with missing data

We undertook a complete‐case analysis (the number analyzed) and an intention‐to‐treat analysis when data were available.

Assessment of heterogeneity

We initially assessed heterogeneity in the results of the trials by inspecting the graphical presentations and by calculating the Chi2 test of heterogeneity. However, we were aware of the fact that the Chi2 test has a poor ability to detect statistically significant heterogeneity among studies. We therefore 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 (the I2 statistic) describes the percentage of total variation across studies that are due to heterogeneity rather than to the play of chance (Higgins 2003). The I2 statistic values lie between 0% and 100%, and a simplified categorization of heterogeneity could be low, moderate, and high for I2 statistic values of 25%, 50%, and 75% respectively (Higgins 2003).

Assessment of reporting biases

If 10 or more studies were included in an outcome, we examined a funnel plot for the primary outcome (severe rotavirus diarrhoea), estimating the precision of trials (plotting the RR against the standard error (SE) of the log of RR) to estimate potential asymmetry.

Data synthesis

We stratified all analyses by the type of vaccine, RV1, RV5 or Rotavac. Subsequently, we grouped all outcomes in the meta‐analyses according to the time point when the outcome was measured or the number of rotavirus seasons, or both, as follows: less than two months; up to one year (one rotavirus season); up to two years (up to two rotavirus seasons); and up to three years (three rotavirus seasons). If data were available for more than one time point, we used the number of completers for each time point in the trial.

For the current update, we stratified each primary outcome (rotavirus diarrhoea, all‐cause diarrhoea, all‐cause death, all serious adverse events, and intussusception) and selected secondary outcomes (rotavirus diarrhoea and all‐cause diarrhoea of any severity, and all‐cause hospitalization) by country mortality rate according to WHO mortality strata (WHO 1999), as follows:

  • Low‐mortality: countries in WHO strata A and B (very low/low child mortality and low adult mortality)

  • High‐mortality: countries in WHO strata D and E (high child mortality and high/very high adult mortality)

We used a fixed‐effect model, unless we found statistically significant heterogeneity (P < 0.10) for a specific outcome, in which case we used the random‐effects model.

We included separate analyses for cases of diarrhoea (e.g. a child who has diarrhoea regardless of the number of episodes) and episodes (i.e. one child can experience more than one episode), where data permitted. We combined episodes using the rate ratio in the logarithmic scale and SE, with the uncertainty in each result being expressed using a 95% CI (9.4.8. in Higgins 2011).

Certainty of the evidence

We interpreted the findings of this review using the GRADE approach (Schünemann 2017), and we used GRADE profiler (GRADE 2004) to import data from RevMan 5 (RevMan 2014) to create ‘Summary of findings' tables. These tables provide outcome‐specific information concerning the overall certainty of evidence from each included study in the comparison, the magnitude of effect of the interventions examined, and the sum of available data on all outcomes we rated as important to patient care and decision‐making, and is reflected as follows: high certainty ("vaccine prevents...."); moderate certainty ("vaccine probably prevents..."); low certainty ("vaccine may prevent...."); and very low certainty ("we do not know whether or not the vaccine prevents....").

We selected primary outcomes, all stratified by vaccine and high or low country mortality, for inclusion in the ‘Summary of findings' tables: severe rotavirus diarrhoea; severe all‐cause diarrhoea; all‐cause death; serious adverse events; and intussusception.

Subgroup analysis and investigation of heterogeneity

In addition to stratifying the results by country‐based high‐mortality and low‐mortality rates using WHO mortality country strata (WHO 1999), we planned to perform subgroup analyses to assess the impact of the following possible sources of heterogeneity for any of the included vaccines: vaccine protection against specific rotavirus G types; and vaccination of special groups, including immunocompromised (including HIV‐infected) children and children with malnutrition. In previous versions of this review (Soares‐Weiser 2010; Soares‐Weiser 2012a), we also analyzed vaccine effect according to each study's country income, use of other childhood vaccines, number of doses administered, source of funding, and whether infants were born prematurely or were breast‐ or formula‐fed. These subgroup analyses did not show any differences, and are not presented in this updated version; they can be found in Soares‐Weiser 2010 and Soares‐Weiser 2012a.

Sensitivity analysis

We also planned to conduct sensitivity analyses for the primary outcomes according to allocation concealment (high, low, and unclear risk of bias) for outcomes in which data could not be pooled because of significant heterogeneity (I2 statistic > 75%).

Results

Description of studies

Results of the search

The update search in 2017 identified 1247 records and the update search in 2018 identified a further 488 records. After de‐duplication, we screened 1614 records and considered 1500 to be irrelevant. We reviewed the full texts of 114 records. In the previously published version of this review there were 41 included studies. The review now includes 55 independent trials (see Characteristics of included studies), 14 of which are new to this update (RV1 Colgate 2016‐BGD; RV1 Kim 2012‐KOR; RV1 Li 2013a‐CHN; RV1 Li 2013b‐CHN; RV1 Li 2014‐CHN; RV1 NCT00158756‐RUS; RV1 Zaman 2017‐BGD; RV5 Dhingra 2014‐IND; RV5 Levin 2017‐AF; RV5 Mo 2017‐CHN; VAC Bhandari 2006‐IND; VAC Bhandari 2009‐IND; VAC Bhandari 2014‐IND; VAC Chandola 2017‐IND) and we also added another 23 new companion papers to previously included trials with this update. The review also includes 15 ongoing studies (see Characteristics of ongoing studies). We excluded 78 studies for the reasons given in the Characteristics of excluded studies section.

Included studies

The 55 included trials enrolled about 216,480 participants (approximate number, as some trials provided only the number evaluable), and each trial compared a rotavirus vaccine with a placebo. The vaccines tested were RV1 (36 trials reported in 171 publications or reports; 119,114 participants), RV5 (15 trials reported in 60 publications or reports; 88,934 participants), and Rotavac (4 trials reported in 13 publications or reports; 8432 participants).

The trials were conducted in Africa, Asia, Europe, and the Americas, and the location can be identified in the study reference: AF, Africa; AS, Asia; EU, Europe; INT, several international locations; LA, Latin America; NA, North America; or country three‐letter acronym according to ISO 3166‐1 Alpha‐3 (e.g. BGD for Bangladesh) from www.all‐acronyms.com/special/countries_acronyms_and_abbreviations, if the study was conducted in a single country.

1. RV1

The 36 RV1 trials were published between 1998 and 2017. Five of the trials are unpublished and were located on the GlaxoSmithKline website through clinicalstudyresults.org or clinicaltrials.gov. One trial (RV1 Madhi 2010‐AF) provided country‐specific data for efficacy outcomes but not for safety outcomes, and was consequently split into RV1 Madhi 2010‐MWI and RV1 Madhi 2010‐ZAF for the Malawi‐ and South Africa‐specific data. Twenty‐five trials enrolled around 500 participants or fewer, three trials enrolled around 1000 participants, seven trials enrolled between 2155 and 12,318 participants, and one large trial enrolled 63,225 participants. Most children were aged between one and three months at the time of the first vaccination.

Population

Most trials included healthy infants. Two trials included HIV‐infected or ‐exposed infants (RV1 Madhi 2010‐AF; RV1 Steele 2010a‐ZAF), one trial included premature infants (RV1 Omenaca 2012‐EU), and one trial included children aged two to six years (RV1 Li 2013a‐CHN).

Outcome measures

Each trial reported on one or more of the outcome measures specified for this review (see Appendix 2). We included data on participants requiring medical visits, as this was reported in some trials and is a similar outcome measure to participants requiring hospitalization.

Twenty‐three trials were safety studies, reporting mainly safety outcomes (e.g. serious adverse events and reactogenicity), immunogenicity outcomes, or both. Eleven of these trials also reported efficacy outcomes with a follow‐up of up to two months. Eleven trials reported one or more efficacy outcomes (e.g. rotavirus diarrhoea) in addition to safety outcomes; most reported one or more immunogenicity outcomes. Two trials reported on efficacy or effectiveness but not safety or immunogenicity (RV1 Colgate 2016‐BGD; RV1 Zaman 2017‐BGD). The trials varied in the length of follow‐up, but in general the trials that specified efficacy outcome measures had longer follow‐up times (Appendix 2).

As shown in Appendix 3, rotavirus diarrhoea (of any severity) was the most common efficacy outcome reported (by 23 trials); 14 trials reported on severe rotavirus diarrhoea, and 10 reported on rotavirus diarrhoea requiring hospitalization. Data on all‐cause diarrhoea were provided by 17 trials, and severe all‐cause diarrhoea by nine trials. Most reported all‐cause death and dropouts, but other efficacy outcomes were reported by few trials.

For safety outcomes (Appendix 4), 29 trials reported on reactogenicity, all but four trials reported on serious adverse events, and 24 reported on adverse events leading to discontinuation of the intervention.

Most trials reported on one or more immunogenicity outcomes; see Appendix 4.

Location

Early trials were conducted in North America and Europe, but since 2005 trials have also been conducted in Asia (Bangladesh, China, India, Japan, Philippines, South Korea, Singapore, Thailand, Vietnam; 17 trials), Latin America (Argentina, Brazil, Chile, Colombia, Dominican Republic, Honduras, Mexico, Nicaragua, Panama, Peru, Venezuela; six trials), and Africa (South Africa, Malawi; four trials); see Appendix 5. Most trials had multiple sites, often in several countries; RV1 Vesikari 2007a‐EU included 98 sites in six European countries.

Country mortality rate

Most trials were conducted in countries with low mortality rates, corresponding to WHO mortality strata A and B. Eight trials were conducted in countries with high mortality rates (RV1 Colgate 2016‐BGD; RV1 Madhi 2010‐AF; RV1 Narang 2009‐IND; RV1 Steele 2008‐ZAF; RV1 Steele 2010a‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2009‐BGD; RV1 Zaman 2017‐BGD), corresponding to WHO mortality strata D and E; see Appendix 5. For RV1 Madhi 2010‐AF, available data were split between countries into RV1 Madhi 2010‐MWI and RV1 Madhi 2010‐ZAF. Two trials were conducted in several countries with both low and high mortality: RV1 GSK[033] 2007‐LA was conducted in four study centres in a high‐mortality country (Peru), but also in three study centres in two low‐mortality countries (Colombia and Mexico), and was placed in the high‐mortality group; and RV1 Ruiz‐Palac 06‐LA/EU was conducted mainly in low‐mortality countries in Latin America and in Finland, but also in two high‐mortality countries (Nicaragua and Peru), and was placed in the low‐mortality group.

Vaccine schedule

The trials varied in the vaccine dose and schedule (see Appendix 6). Most trials gave two doses of the vaccine with virus concentration of more than 106 plaque‐forming units (PFUs). Older trials, conducted between 1998 and 2005, tended to include slightly lower PFUs or a range of PFUs for comparison.

RV1 was given as two doses in all but five trials: one trial conducted in partnership with GlaxoSmithKline and PATH Rotavirus Vaccine Program tested two and three doses of the vaccine (RV1 Madhi 2010‐AF); another trial conducted by GlaxoSmithKline in which the poliovirus vaccine was co‐administered with RV1, tested two or three vaccine doses to investigate differences in immune response (RV1 Steele 2010b‐ZAF); a third study tested three vaccine doses in HIV‐positive infants (RV1 Steele 2010a‐ZAF); a fourth study tested three vaccine doses in healthy infants (RV1 GSK[021] 2007‐PAN); a fifth study that included children aged two to six years administered one dose only (RV1 Li 2013a‐CHN).

Some trials compared more than one arm: different PFU virus concentrations (RV1 Vesikari 2004a‐FIN; RV1 Dennehy 2005‐NA; RV1 Phua 2005‐SGP; RV1 Salinas 2005‐LA; RV1 Ward 2006‐USA); different formulations (RV1 GSK[021] 2007‐PAN; RV1 GSK[033] 2007‐LA; RV1 GSK[101555] 2008‐PHL; RV1 Kerdpanich 2010‐THA; RV1 Vesikari 2011‐FIN); co‐administration of other vaccine (RV1 Steele 2008‐ZAF; RV1 Zaman 2009‐BGD; RV1 NCT00158756‐RUS; RV1 Li 2014‐CHN); and different intervals between doses (RV1 Anh 2011‐PHL; RV1 Anh 2011‐VNM).

Infant vaccination status

All but four trial reports referred to vaccination with other infant vaccines (see Appendix 6). Most trials co‐administered other routine infant vaccines, such as diphtheria‐tetanus‐acellular pertussis, Haemophilus influenzae type b (HiB), inactivated polio vaccine, and hepatitis B vaccine (HBV). Some trials also co‐administered oral polio vaccine. Other trials imposed a two‐week separation between other infant vaccines and rotavirus vaccine or placebo, or specified other vaccines as not allowed.

Methods for collecting adverse event data

Fifteen of the 36 trials did not provide details of how adverse event data were collected. Out of the trials that did report the method of collecting adverse event data, 13 trials used passive methods (e.g. diary cards), two used an active method ("active surveillance system"), and five used both passive and active methods (e.g. diary card plus regular telephone calls to parents); see Appendix 7.

Source of funding

Most trials were supported by GlaxoSmithKline Biologicals, three of which were in partnership with PATH Rotavirus Vaccine Program (RV1 Li 2014‐CHN; RV1 Madhi 2010‐AF; RV1 Zaman 2009‐BGD), and another two in partnership with RAPID trials and the WHO (RV1 Steele 2008‐ZAF; RV1 Steele 2010a‐ZAF). One trial was funded by The Bill and Melinda Gates Foundation (RV1 Colgate 2016‐BGD) and one by GAVI and PATH (RV1 Zaman 2017‐BGD). Three trials were sponsored by Avant Immunotherapeutics (formerly Virus Research Institute, Inc.) (RV1 Bernstein 1998‐USA; RV1 Bernstein 1999‐USA; RV1 Ward 2006‐USA).

2. RV5

We identified 15 trials of RV5 vaccine. The earliest was reported in 2003 and the most recent in 2017. One of the trials is unpublished and was accessed via clinicalstudyresults.org. Two trials (RV5 Armah 2010‐AF and RV5 Zaman 2010‐AS) provided country‐specific data for some outcomes but not for all outcomes, and were consequently split into RV5 Armah 2010‐GHA; RV5 Armah 2010‐KEN; and RV5 Armah 2010‐MLI for the Ghana‐, Kenya, and Mali‐specific data, and RV5 Zaman 2010‐BGD and RV5 Zaman 2010‐VNM for the Bangladesh‐ and Vietnam‐specific data. Overall, 88,934 participants were included in the trials; the largest trial included 70,301 participants (RV5 Vesikari 2006b‐INT) and the smallest included 48 participants (RV5 Lawrence 2012‐CHN). For the 2012 update of this review, we received new information from Merck (Merck 2012) for some of the trials on the outcomes serious adverse events, intussusception, and deaths. We have incorporated the new information into the analyses and have indicated this in the Characteristics of included studies section.

Population

Most trials included healthy infants. One trial included both healthy and HIV‐infected infants (RV5 Armah 2010‐KEN), another trial included HIV‐exposed but uninfected and HIV‐infected infants (RV5 Levin 2017‐AF), and one trial included prematurely‐born infants as well as those born at normal gestation (RV5 Vesikari 2006b‐INT). All but two trials enrolled children aged between one month and three months; the children in RV5 Vesikari 2006a‐FIN were aged between three months and six months, and there was a child cohort (2‐ to 6‐year‐old children) in addition to an infant cohort in RV5 Lawrence 2012‐CHN.

Outcome measures

Six trials were safety studies (Appendix 2), reporting safety outcomes (e.g. serious adverse events and reactogenicity) and generally immunogenicity outcomes as well. The other nine trials reported one or more efficacy and safety outcomes, and seven out of those nine also reported immunogenicity outcomes (Appendix 2). The trials varied in the length of follow‐up (Appendix 2), but in general the trials that specified efficacy outcome measures had longer follow‐up times (up to three years). Similar to the RV1 trials, we included data on participants requiring medical visits, as this was reported in some trials and is a similar outcome measure to participants requiring hospitalization.

As shown in Appendix 3, rotavirus diarrhoea, severe cases and cases of any severity, were the most common efficacy outcomes reported (by eight trials); only one of these reported rotavirus diarrhoea requiring hospitalization. Three trials provided data on severe cases of all‐cause diarrhoea; two also presented data on cases with any severity. Eleven trials reported all‐cause death, and 13 of the 15 trials reported dropouts.

For safety outcomes, all trials reported on serious adverse events and reactogenicity, and 13 trials reported on adverse events leading to discontinuation of the intervention; see Appendix 4.

Twelve trials reported on an immunogenicity outcome (Appendix 4).

Location

Half of the trials were conducted in low‐mortality countries in North America and Europe. Six trials, including the smallest and the largest trials, were conducted in other regions: RV5 Armah 2010‐AF was conducted in Ghana, Kenya and Mali; RV5 Levin 2017‐AF was conducted in Botswana, Tanzania, Zambia and Zimbabwe, RV5 Dhingra 2014‐IND was conducted in India, RV5 Kim 2008‐KOR was conducted in South Korea; RV5 Iwata 2013‐JPN was conducted in Japan; RV5 Lawrence 2012‐CHN and RV5 Mo 2017‐CHN were conducted in China; RV5 Vesikari 2006b‐INT was conducted in 12 countries in Asia, the Caribbean, Europe, Latin America, North America; and RV5 Zaman 2010‐AS was conducted in Bangladesh and Vietnam. Each trial had multiple sites, ranging from three (RV5 Vesikari 2006a‐FIN) to 356 sites (RV5 Vesikari 2006b‐INT); see Appendix 5.

Country mortality rate

Most trials were conducted in countries with low mortality rates, corresponding to WHO mortality strata A and B; see Appendix 5. One trial was conducted in high‐mortality India (RV5 Dhingra 2014‐IND). Four trials were conducted in several low‐ and high‐mortality countries. RV5 Armah 2010‐AF was conducted in three high‐mortality countries, Ghana, Kenya, and Mali, and when available the data were split into RV5 Armah 2010‐GHA, RV5 Armah 2010‐KEN and RV5 Armah 2010‐MLI. RV5 Levin 2017‐AF was conducted in four high‐mortality countries (Botswana, Tanzania, Zambia and Zimbabwe). RV5 Vesikari 2006b‐INT was conducted mainly in European and Latin American low‐mortality countries, but also in Guatemala, a high‐mortality country, and was placed in the low‐mortality group. RV5 Zaman 2010‐AS was conducted in one high‐mortality country (Bangladesh) with 1136 participants, and in one low‐mortality country (Vietnam) with 900 participants, and was placed in the high‐mortality group, except when data could be split into RV5 Zaman 2010‐BGD and RV5 Zaman 2010‐VNM.

Vaccine schedule

Each trial used three doses of RV5 vaccine, with intervals between doses of four and 10 weeks (see Appendix 6). All but two trials had one vaccine and one placebo arm; RV5 Vesikari 2006a‐FIN included three vaccine arms in which there were different RV5 components (G1‐4, P1A, G1‐4, and P1A), and RV5 Dhingra 2014‐IND included a RV5 arm, a placebo arm, and three arms with different concentrations of BRV‐TV vaccine.

Infant vaccination status

Most trials did not restrict the use of other childhood vaccines (see Appendix 6). Two trials co‐administered hepatitis B, diphtheria‐tetanus‐pertussis, poliovirus, and H influenzae type b vaccines with RV5 (RV5 Ciarlet 2009‐EU; RV5 Dhingra 2014‐IND). One trial randomized participants to either concomitant or staggered administration of other childhood vaccines (OPV, DTaP) with RV5 or placebo (RV5 Mo 2017‐CHN). Three trials allowed the use of oral polio vaccine, in addition to other licensed childhood vaccines (RV5 Armah 2010‐AF; RV5 Mo 2017‐CHN; RV5 Zaman 2010‐AS). Three trials did not allow the use of other vaccines (RV5 Clark 2003‐USA; RV5 Clark 2004‐USA; RV5 Lawrence 2012‐CHN), and one trial did not mention their use (RV5 Iwata 2013‐JPN).

Methods for collecting adverse event data

As shown in Appendix 7, seven trials used a combination of passive methods (e.g. diary cards for parents) and active methods (directly contacting parents) to collect adverse event data. The other trials used passive methods only (diary cards, three trials), active methods only ("active surveillance", three trials), or the information was not provided (two trials).

Source of funding

All but one trial was funded by Merck & Co., Inc. Two of those trials also received funding and were run by PATH (GAVI Alliance grant) (RV5 Armah 2010‐AF; RV5 Zaman 2010‐AS), and one trial also received funding from the International Maternal, Pediatric, and Adolescent AIDS Clinical Trial Network (IMPAACT) through the National Institute of Health (RV5 Levin 2017‐AF). One trial was funded by Shantha Biotechnics Ltd (RV5 Dhingra 2014‐IND).

3. Rotavac

We identified four trials of Rotavac vaccine. The earliest was reported in 2006 and the most recent in 2017. Overall, 8432 participants were included in the trials; the largest trial included 6799 participants (VAC Bhandari 2014‐IND) and the smallest included 90 participants (VAC Bhandari 2006‐IND).

Population

All trials included healthy infants. Trials enrolled infants aged between six weeks and nine weeks.

Outcome measures

Three trials were safety studies (Appendix 2) reporting safety outcomes and immunogenicity outcomes. They reported on follow‐up results for one to 12 months after the last vaccine dose. The other trial (VAC Bhandari 2014‐IND) reported on efficacy, safety, and immunogenicity outcomes until the infants were two years of age.

As shown in Appendix 3, VAC Bhandari 2014‐IND reported on rotavirus diarrhoea (severe cases, cases of any severity, and cases requiring medical attention). The same trial also provided data on severe cases of all‐cause diarrhoea. Two trials reported all‐cause death, and three of the four trials reported dropouts.

For safety outcomes, all trials reported on serious adverse events and two reported on reactogenicity. All trials reported on an immunogenicity outcome (Appendix 4).

Location

All trials were conducted in India, one at three sites in the cities of Delhi, Pune, and Vellore (VAC Bhandari 2014‐IND), and the remaining three studies at one site in Delhi.

Country mortality rate

All trials were conducted in India, a high‐mortality country (WHO mortality stratum D).

Vaccine schedule

Most trials used three doses of Rotavac vaccine, with intervals between doses of four to eight weeks (see Appendix 6). One trial (VAC Bhandari 2006‐IND) administered one dose. One trial had one vaccine and one placebo arm (VAC Bhandari 2014‐IND). VAC Bhandari 2006‐IND included an additional vaccine arm for a rotavirus vaccine candidate (I321) that we did not include for analysis in this review. VAC Bhandari 2009‐IND randomized participants to high‐ (1 x 105 ffu) and low‐dose (1 x 104 ffu) vaccine arms which we combined in this review. VAC Chandola 2017‐IND randomized participants to three vaccine production lots as well as to placebo. We combined the different production lot arms in our analyses.

Infant vaccination status

Two trials separated the use of other routine childhood vaccines from Rotavac administration by at least two weeks (VAC Bhandari 2006‐IND; VAC Bhandari 2009‐IND). Two trials co‐administered other routine childhood vaccines (OPV, DPT, Hep B and Hib) with Rotavac (VAC Bhandari 2014‐IND; VAC Chandola 2017‐IND).

Methods for collecting adverse event data

As shown in Appendix 7, three trials used a combination of passive methods (e.g. diary cards for parents) and active methods (directly contacting parents) to collect adverse event data. The other trial (VAC Chandola 2017‐IND) used active methods only (directly contacting parents).

Source of funding

One trial was funded by Bharat Biotech (VAC Bhandari 2006‐IND), one trial was co‐funded by Bharat Biotech (VAC Bhandari 2009‐IND) and the other two trials were funded by PATH, the Government of India, and other not‐for‐profit organizations (VAC Bhandari 2014‐IND; VAC Chandola 2017‐IND).

Ongoing studies

We identified 15 ongoing trials, three of RV1, one of RV5 and 11 others (RV1 together with RV5; RV3‐BB; Rotasiil; Rotavac; BRV‐TV; Trivalent P2VP8; Bio Farma's rotavirus vaccine) (see Characteristics of ongoing studies). As shown in Appendix 8, the RV1 trials are being conducted in South Africa and Bangladesh. The ongoing RV5 trial is in Bangladesh, and the studies testing other vaccines are located in Australia, Bangladesh, China, India, Indonesia, Malawi, Mexico, South Africa, and the USA.

Excluded studies

There are 78 excluded studies with 100 references (Figure 2). We excluded most studies because they were not RCTs (34 studies). We excluded 27 studies because they reported on comparisons not relevant to this review, three studies because they did not report on RV vaccines, three because they included adult populations, 10 because they reported on unlicensed vaccines in development (OTHER Bines 2015; OTHER Bines 2018; OTHER Cowley 2017; OTHER Groome 2017) or licensed vaccines that have not been prequalified by the WHO (OTHER CTRI/2009/091/000821; OTHER Dang 2012; OTHER Isanaka 2017‐NER; OTHER Kulkarni 2017; OTHER Zade 2014a‐IND; OTHER Zade 2014b‐IND), and one because it reported on a withdrawn vaccine (OTHER Armah 2013).

Figure 2
PRISMA diagram.

PRISMA diagram.

Risk of bias in included studies

We prepared a ‘Risk of bias' assessment for each trial, with a focus on the rotavirus diarrhoea outcome measure. Of the 55 RCTs analyzed in this review, 48 (87%) reported an adequate generation of allocation sequence, while the method of assignment was unclear in the remaining studies. We considered the methods used to conceal allocation to be adequate in 46 trials (84%), and unclear in the remaining studies. Information about blinding of participants, care providers, or outcome assessors was provided and we considered it to be adequate in 42 studies (76%), unclear for nine studies, and at high risk of bias for four studies (RV1 Colgate 2016‐BGD; RV1 Kerdpanich 2010‐THA; RV1 Zaman 2017‐BGD; RV5 Dhingra 2014‐IND). Incomplete outcome data were adequately addressed in 46 studies (84%), unclear in eight studies, and was not addressed adequately in one study. Thirty‐eight (69%) trials were free from selective reporting bias, nine were not, and the remaining eight trials were unclear. No other bias was apparent for 31 trials (56%). An overall pictorial summary of the ‘Risk of bias' assessment is shown in Figure 3 and Figure 4.

Figure 3
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 4
Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

RV1

Since the previous update of this review, detailed clinical study reports of most of the GlaxoSmithKline‐sponsored studies (another five, totaling 27 of the 36 trials) have been published online (gsk‐clinicalstudyregister.com). Full details of blinding, participant selection, and attrition are available from these reports, and we could subsequently update risks of bias for these studies, where previously there was no information available. We rated five trials as at high risk of bias for at least one domain; three trials for blinding (RV1 Colgate 2016‐BGD; RV1 Kerdpanich 2010‐THA; RV1 Zaman 2017‐BGD), and three trials for selective reporting bias (RV1 Ruiz‐Palac 06‐LA/EU; RV1 Salinas 2005‐LA; RV1 Zaman 2017‐BGD).

RV5

Based on unpublished information provided by Merck, many of the trials' risks of bias were upgraded for the previous 2012 version of this review. Details of the new information are indicated in the ‘Risk of bias' tables in the Characteristics of included studies section. We judged 10 of the 15 RV5 trials as having a low risk of bias for sequence generation, allocation concealment, and blinding, and varying risks of bias for attrition, selective reporting and other bias. We rated two of these trials (RV5 Armah 2010‐AF; RV5 Zaman 2010‐AS) at an overall low risk of bias. Seven of the 15 RV5 trials had a high risk of bias for one or more domains, most commonly a high risk of selective reporting.

Rotavac

Peer‐reviewed articles for most Rotavac studies reported clearly on how the trials were conducted. Full details about blinding, participant selection, attrition, and outcome reporting could be obtained from most of these reports. We rated only one of the trials at unclear risk of performance and detection bias, since no details about blinding were provided and unclear risk of attrition bias since not all outcomes were assessed with the full study population and the reason for this was not clear (VAC Bhandari 2009‐IND).

Effects of interventions

See: Summary of findings for the main comparison RV1 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries; Summary of findings 2 RV1 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries; Summary of findings 3 RV5 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries; Summary of findings 4 RV5 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries; Summary of findings 5 Rotavac compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

1. RV1

1.1. Primary outcomes
1.1.1. Rotavirus diarrhoea: severe

Eleven trials provided data on the efficacy of RV1 to prevent severe rotavirus diarrhoea in children; see Analysis 1.1 for up to one‐year follow‐up and Analysis 1.2 for two years follow‐up. Trials were performed in low‐mortality countries (RV1 Bernstein 1999‐USA; RV1 Kawamura 2011‐JPN; RV1 Li 2014‐CHN; RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Salinas 2005‐LA; RV1 Tregnaghi 2011‐LA; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2007a‐EU), and high‐mortality countries (RV1 Colgate 2016‐BGD; RV1 Madhi 2010‐MWI; RV1 Madhi 2010‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2017‐BGD). Data below are grouped accordingly.

Low‐mortality countries (WHO strata A and B)

RV1 reduced severe rotavirus diarrhoea cases by 84% after one year (RR 0.16, 95% CI 0.09 to 0.26; 43,779 participants, 7 trials) and by 82% after two years (RR 0.18, 95% CI 0.14 to 0.23; 36,002 participants, 9 trials; Analysis 1.2). After three years there was no statistically significant difference between RV1 and placebo (RR 0.10, 95% CI 0.01 to 1.52; 12,109 participants, two trials (RV1 Phua 2009‐AS and RV1 Vesikari 2007a‐EU; data not shown)). Pooled results showed statistical heterogeneity at one‐year (I2 statistic = 61%, Analysis 1.1) and three years (I2 statistic = 69%, data not shown) follow‐up.

High‐mortality countries (WHO strata D and E)

RV1 reduced severe rotavirus diarrhoea cases by 63% during the first year of follow‐up (RR 0.37, 95% CI 0.23 to 0.60; 6114 participants, 4 comparisons from 3 trials) and by 35% after two years (RR 0.65, 95% CI 0.51 to 0.83; 7113 participants, 3 comparisons from 2 trials; Analysis 1.2). Pooled results showed statistical heterogeneity at one‐year follow‐up (I2 statistic = 57%, Analysis 1.1).

We noted a funnel plot asymmetry for trials reporting results up to one year (Figure 5).

Figure 5
Funnel plot of comparison: 1 RV1 versus placebo, outcome: 1.1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

Funnel plot of comparison: 1 RV1 versus placebo, outcome: 1.1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

1.1.2. All‐cause diarrhoea: severe

Severe all‐cause diarrhoea was reported as cases in six trials (RV1 Colgate 2016‐BGD; RV1 Li 2014‐CHN; RV1 Madhi 2010‐AF; RV1 Phua 2005‐SGP; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Vesikari 2007a‐EU) and as episodes in two trials (RV1 Phua 2009‐AS; RV1 Ruiz‐Palac 06‐LA/EU). We have reported these data separately. Trials were performed in low‐mortality countries (RV1 Li 2014‐CHN; RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Vesikari 2007a‐EU), and in high‐mortality countries (RV1 Colgate 2016‐BGD; RV1 Madhi 2010‐MWI; RV1 Madhi 2010‐ZAF).

Low‐mortality countries (WHO strata A and B)

RV1 reduced the number of severe cases of all‐cause diarrhoea by 41% at one year (RR 0.59, 95% CI 0.47 to 0.74; 28,051 participants, 3 trials; Analysis 1.3), and by 40% at two years (RR 0.60, 95% CI 0.36 to 1.02; 9417 participants, 3 trials; Analysis 1.4). Pooled results showed statistical heterogeneity at both one year (I2 statistic = 63%) and two years follow‐up (I2 statistic = 90%). RV1 reduced the rate of severe episodes of all‐cause diarrhoea by 40% at one year (rate ratio 0.60, 95% CI 0.50 to 0.72; 17,867 participants, 1 trial; Analysis 1.5), and by 37% at two years (rate ratio 0.63, 95% CI 0.56 to 0.71; 39,091 participants, 2 trials; Analysis 1.6). One trial reported on severe all‐cause diarrhoea after three years follow‐up (RV1 Phua 2009‐AS); RV1 reduced the number of severe cases by 27% (RR 0.73, 95% CI 0.61 to 0.88; 10,519 participants; data not shown).

High‐mortality countries (WHO strata D and E)

RV1 reduced the number of severe cases of all‐cause diarrhoea by 27% at one year follow‐up (RR 0.73, 95% CI 0.56 to 0.95; 5639 participants, 3 comparisons from 2 trials; Analysis 1.3), and by 17% at two years follow‐up (RR 0.83, 95% CI 0.72 to 0.96; 2764 participants, 2 comparisons from 1 trial; Analysis 1.4). Pooled results showed statistical heterogeneity at one‐year follow‐up (I2 statistic = 75%).

1.1.3. All‐cause death

Thirty trials reported on all‐cause death, either as the number of deaths (RV1 Bernstein 1999‐USA; RV1 Kim 2012‐KOR; RV1 Li 2013b‐CHN; RV1 Li 2014‐CHN; RV1 Madhi 2010‐AF; RV1 NCT00158756‐RUS; RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Steele 2010a‐ZAF; RV1 Vesikari 2007a‐EU) or as the number of fatal serious adverse events (RV1 Anh 2011‐PHL; RV1 Anh 2011‐VNM; RV1 GSK[021] 2007‐PAN; RV1 GSK[033] 2007‐LA; RV1 GSK[041] 2007‐KOR; RV1 GSK[101555] 2008‐PHL; RV1 Kawamura 2011‐JPN; RV1 Kerdpanich 2010‐THA; RV1 Narang 2009‐IND; RV1 Omenaca 2012‐EU; RV1 Rivera 2011‐DOM; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Salinas 2005‐LA; RV1 Steele 2008‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Tregnaghi 2011‐LA; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2011‐FIN; RV1 Zaman 2009‐BGD). We pooled the number of deaths and fatal serious adverse events; see Analysis 1.7. We present details of causes of death for each trial in Appendix 9. Most trials were performed in low‐mortality countries, with eight trials in high‐mortality countries (RV1 Colgate 2016‐BGD; RV1 GSK[033] 2007‐LA; RV1 Madhi 2010‐AF; RV1 Narang 2009‐IND; RV1 Steele 2008‐ZAF; RV1 Steele 2010a‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2009‐BGD).

Low‐mortality countries (WHO strata A and B)

There was no statistically significant difference in all‐cause death between the two arms (RR 1.22, 95% CI 0.87 to 1.71; 97,597 participants, 22 trials).

High‐mortality countries (WHO strata D and E)

There was no statistically significant difference in all‐cause death between the two arms (RR 0.88, 95% CI 0.64 to 1.22; 8181 participants, 8 trials).

1.1.4. All serious adverse events

The total number of serious adverse events was reported in 31 trials, performed in low‐mortality countries (RV1 Anh 2011‐PHL; RV1 Anh 2011‐VNM; RV1 Bernstein 1998‐USA; RV1 Dennehy 2005‐NA; RV1 GSK[021] 2007‐PAN; RV1 GSK[041] 2007‐KOR; RV1 GSK[101555] 2008‐PHL; RV1 Kawamura 2011‐JPN; RV1 Kerdpanich 2010‐THA; RV1 Kim 2012‐KOR; RV1 Li 2013a‐CHN; RV1 Li 2014‐CHN; RV1 NCT00158756‐RUS; RV1 Omenaca 2012‐EU; RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Rivera 2011‐DOM; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Salinas 2005‐LA; RV1 Tregnaghi 2011‐LA; RV1 Vesikari 2004a‐FIN; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2007a‐EU; RV1 Vesikari 2011‐FIN), and in high‐mortality countries (RV1 GSK[033] 2007‐LA; RV1 Madhi 2010‐AF; RV1 Narang 2009‐IND; RV1 Steele 2008‐ZAF; RV1 Steele 2010a‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2009‐BGD); see Analysis 1.8.

Low‐mortality countries (WHO strata A and B)

Fewer children allocated to RV1 had serious adverse events compared with placebo (RR 0.88, 95% CI 0.83 to 0.93; 96,233 participants, 24 trials). In addition, in one trial (RV1 Li 2013a‐CHN) that vaccinated 25 older children (aged two to six years) with one‐dose RV1 there were no serious adverse events reported.

High‐mortality countries (WHO strata D and E)

There was no statistically significant difference in the number of serious adverse events between the two arms (RR 0.89, 95% CI 0.76 to 1.04; 7481 participants, 7 trials).

1.1.5. Serious adverse events: intussusception

Twenty‐one trials reported on intussusception, and 11 of these reported that no cases of intussuception had occurred. Trials were performed in low‐mortality countries (RV1 Dennehy 2005‐NA; RV1 GSK[041] 2007‐KOR; RV1 Kawamura 2011‐JPN; RV1 Kim 2012‐KOR; RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Rivera 2011‐DOM; RV1 Ruiz‐Palac 06‐LA/EU; RV1 Salinas 2005‐LA; RV1 Tregnaghi 2011‐LA; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2007a‐EU; RV1 Vesikari 2011‐FIN), and in high‐mortality countries (RV1 Madhi 2010‐AF; RV1 Steele 2008‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2017‐BGD); see Analysis 1.9.

Low‐mortality countries (WHO strata A and B)

Twenty‐nine cases of intussusception were reported in a total of 49,355 children in the RV1 arm compared with 28 cases of intussusception in 42,477 children of the placebo arm. Pooled results showed no increased risk for intussusception in children receiving RV1 when compared to placebo (RR 0.69, 95% CI 0.45 to 1.04; 96,513 participants, 17 trials).

High‐mortality countries (WHO stratum E)

One case of intussusception was reported in a total of 3677 children in the RV1 arm compared with no cases of intussusception in 1737 children in the placebo or no‐intervention arm. Pooled results showed no increased risk for intussusception in children receiving RV1 when compared to placebo (RR 1.49, 95% CI 0.06 to 36.63; 10,460 participants, 4 trials).

1.2. Secondary outcomes
1.2.1 Serious adverse events: Kawasaki disease

Three trials reported four cases of Kawasaki disease among 7701 children allocated to RV1 compared to no cases in 5416 children allocated to placebo (RV1 Phua 2005‐SGP; RV1 Phua 2009‐AS; RV1 Salinas 2005‐LA). We did not observe a statistically significant difference between the intervention and placebo groups (RR 1.79, 95% CI 0.30 to 10.61; 13,117 participants, 3 trials; Analysis 1.10).

1.2.2. Serious adverse events requiring hospitalization

Two trials reported serious adverse events requiring hospitalization (RV1 Ruiz‐Palac 06‐LA/EU; RV1 Steele 2008‐ZAF) and found fewer events in the RV1 group than the placebo group (RR 0.88, 95% CI 0.81 to 0.96; 63,675 participants, 2 trials; Analysis 1.11).

1.2.3 Rotavirus diarrhoea of any severity

Eighteen trials provided data for the efficacy of RV1 to prevent rotavirus diarrhoea in children; see Analysis 1.12 for two‐months safety trial follow‐up, Analysis 1.13 for one‐year follow‐up and Analysis 1.14 for two‐year follow‐up. Trials were performed in low‐mortality countries (RV1 Anh 2011‐PHL; RV1 Anh 2011‐VNM; RV1 Bernstein 1999‐USA; RV1 GSK[041] 2007‐KOR; RV1 GSK[101555] 2008‐PHL; RV1 Kerdpanich 2010‐THA; RV1 Omenaca 2012‐EU; RV1 Phua 2005‐SGP; RV1 Rivera 2011‐DOM; RV1 Salinas 2005‐LA; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2007a‐EU; RV1 Vesikari 2011‐FIN), and in high‐mortality countries (RV1 Madhi 2010‐MWI; RV1 Madhi 2010‐ZAF; RV1 Narang 2009‐IND; RV1 Steele 2010a‐ZAF; RV1 Steele 2010b‐ZAF; RV1 Zaman 2009‐BGD). Data below are grouped accordingly.

Low‐mortality countries (WHO strata A and B)

Safety trials (up to two months follow‐up): RV1 was not superior to placebo in the prevention of rotavirus diarrhoea in the trials assessing outcomes up to two months after vaccination (RR 1.28, 95% CI 0.66 to 2.50; 3537 participants, 9 trials). These trials, although reporting cases of rotavirus diarrhoea, were not designed to measure efficacy.

Efficacy trials (one to three years follow‐up): RV1 reduced rotavirus diarrhoea by 78% at up to one year (RR 0.22, 95% CI 0.13 to 0.40; 9083 participants, 4 trials) and 65% at the second year of follow‐up (RR 0.35, 95% CI 0.25 to 0.48; 10,441 participants, 6 trials). Pooled results, however, showed statistical heterogeneity at one year (I2 statistic = 80%, Analysis 1.13) and two years (I2 statistic = 55%, Analysis 1.14) of follow‐up. At the third year of follow‐up, there were very few reported cases of rotavirus diarrhoea of any severity. Based on a single trial (RV1 Vesikari 2007a‐EU, 1590 participants), there was no difference between RV1 and placebo groups (data not shown).

High‐mortality countries (WHO strata D and E)

Safety trials (up to two months follow‐up): Three trials found no difference in the RV1 group compared to placebo when outcomes were assessed up to two months after vaccination (RR 1.00, 95% CI 0.41 to 2.41; 757 participants, 3 trials).

Efficacy trials (one to two years follow‐up): RV1 reduced rotavirus diarrhoea by 51% during the first year of follow‐up (RR 0.49, 95% CI 0.35 to 0.68; 6114 participants, 4 comparisons from 3 trials), and by 59% during the second year (RR 0.41, 95% CI 0.28 to 0.62; 1251 participants, 1 trial). Pooled results showed statistical heterogeneity at one‐year follow‐up (I2 statistic = 76%, Analysis 1.13).

1.2.4. All‐cause diarrhoea: of any severity

This outcome was reported as cases in 11 trials from low‐mortality countries (RV1 Anh 2011‐PHL; RV1 Anh 2011‐VNM; RV1 Kerdpanich 2010‐THA; RV1 Kim 2012‐KOR; RV1 Li 2014‐CHN; RV1 Omenaca 2012‐EU; RV1 Phua 2005‐SGP; RV1 Rivera 2011‐DOM; RV1 Salinas 2005‐LA; RV1 Vesikari 2004b‐FIN; RV1 Vesikari 2011‐FIN), in two trials from high‐mortality countries (RV1 Colgate 2016‐BGD; RV1 Steele 2010a‐ZAF), and as episodes in three trials from low‐mortality countries (RV1 Rivera 2011‐DOM; RV1 Salinas 2005‐LA; RV1 Vesikari 2004b‐FIN). We have reported these data separately.

Low‐mortality countries (WHO strata A and B)

Safety trials (up to two months follow‐up): RV1 was not better than placebo in reducing the number of cases of all‐cause diarrhoea at two months (RR 0.86, 95% CI 0.67 to 1.09; 3032 participants, 6 trials; Analysis 1.15).

Efficacy trials (one to two years follow‐up): RV1 was not better than placebo in reducing the number of cases of all‐cause diarrhoea at one year follow‐up (RR 0.92, 95% CI 0.82 to 1.03; 2204 participants, 2 trials, Analysis 1.16), or after two years (RR 0.93, 95% CI 0.87 to 1.00; 5937 participants, 3 trials; Analysis 1.17).Two trials reported the number of episodes, with no statistically significant benefit with RV1 when compared to placebo at one year (Rate Ratio 0.98, 95% CI 0.88 to 1.10; 2204 participants, 2 trials; Analysis 1.18) or at two years (Rate Ratio 1.02, 95% CI 0.78 to 1.33; 736 participants, 1 trial; Analysis 1.19).

High‐mortality countries (WHO stratum E)

Safety trials (up to two months follow‐up): RV1 was not better than placebo in reducing the number of cases of all‐cause diarrhoea at two months (RR 1.04, 95% CI 0.69 to 1.58; 100 participants, 1 trial; Analysis 1.15).

Efficacy trials (one‐year follow‐up): RV1 was not better than no intervention in reducing the number of cases of all‐cause diarrhoea at one‐year follow‐up (RR 0.99, 95% CI 0.93 to 1.05; 700 participants, 1 trial; Analysis 1.16)

1.2.5. All‐cause hospitalizations

Two trials (RV1 Phua 2005‐SGP; RV1 Ruiz‐Palac 06‐LA/EU) provided data for the efficacy of RV1 to prevent all‐cause hospitalizations.

Low‐mortality countries (WHO stratum A)

RV1 was not better than placebo in reducing the number of hospitalizations at up to two years of follow‐up (RR 0.63, 95% CI 0.27 to 1.47; 65,646 participants, 2 trials; Analysis 1.20).

1.2.6. Rotavirus diarrhoea: requiring hospitalization or medical attention

Rotavirus‐related hospitalizations were reduced by 82% after one year (RR 0.18, 95% CI 0.09 to 0.33; 48,718 participants, 8 trials), 85% at two years (RR 0.15, 95% CI 0.11 to 0.22; 35,331 participants, 7 trials), and 95% at three years (RR 0.05, 95% CI 0.02 to 0.16; 10,519 participants, 1 trial (RV1 Phua 2009‐AS, data not shown)); pooled results showed statistical heterogeneity at one year of follow‐up (I2 statistic = 55%); see Analysis 1.21.

RV1 reduced rotavirus‐related medical visits by 92% at one year (RR 0.08, 95% CI 0.04 to 0.16; 3874 participants, 1 trial) and 78% at two years (RR 0.22, 95% CI 0.16 to 0.31; 7017 participants, 3 trials); see Analysis 1.22.

1.2.7. All‐cause diarrhoea: requiring hospitalization

There was no significant difference between RV1 and placebo in cases of hospitalization for all‐cause diarrhoea at one‐year follow‐up (RR 0.43, 95% CI 0.17 to 1.11; 14,393 participants, 2 trials; Analysis 1.23). At two years follow‐up, RV1 reduced cases by 48% (RR 0.52, 95% CI 0.27 to 0.99; 14,367 participants, 2 trials; Analysis 1.23). RV1 Phua 2009‐AS reported that for hospitalizations due to all‐cause diarrhoea at three years of follow‐up, RV1 reduced hospitalizations by 28% (RR 0.72, 95% CI 0.59 to 0.86; 10,519 participants, data not shown). Pooled results showed statistical heterogeneity at one year (I2 statistic = 83%) and at two years follow‐up (I2 statistic = 77%).

RV1 Ruiz‐Palac 06‐LA/EU presented data on the number of episodes (Analysis 1.24); RV1 reduced hospitalizations by 42% at one year (rate ratio 0.58, 95% CI 0.47 to 0.71; 17,867 participants, 1 trial) and 47% at two years (rate ratio 0.53, 95% CI 0.46 to 0.61; 14,286 participants, 1 trial).

1.2.8. Reactogenicity

The occurrence of fever (Analysis 1.25), diarrhoea (Analysis 1.26), and vomiting (Analysis 1.27) were evaluated at several time points: after the first dose, after the second dose, after the third dose, and at the end of the follow‐up period. Most trials contributed data to these outcomes. There were similar results for RV1 and placebo for each outcome and time point.

1.2.9. Adverse events that require discontinuation of vaccination schedule

There was no statistically significant difference between RV1 and placebo in the number of adverse events leading to discontinuation of the vaccination schedule (RR 1.03, 95% CI 0.83 to 1.26; 94,980 participants, 26 trials; Analysis 1.28).

1.3. Immunogenicity

Data on immunogenicity was not stratified by WHO strata. RV1 was more immunogenic than placebo when measured by vaccine virus shedding after the final vaccine dose (RR 10.94, 95% CI 4.90 to 24.43; 2638 participants, 16 trials), although the results showed statistical heterogeneity (I2 statistic = 76%, Analysis 1.29). RV1 was also more immunogenic when measured by seroconversion at all time points (Analysis 1.30); although the pooled data showed statistical heterogeneity after one dose (I2 statistic = 57%), after two doses (I2 statistic = 79%), and after three doses (I2 statistic = 51%).

1.4. Dropouts before the end of trial

Twenty‐eight trials reported on the number of participants who dropped out of the trial before it ended. Overall, there was no statistically significant difference between the RV1 and placebo or no‐intervention groups (RR 0.95, 95% CI 0.90 to 1.00; 93,106 participants, 28 trials; Analysis 1.31).

1.5. Subgroup analyses
1.5.1. G type

Rotavirus diarrhoea: of any severity

Six trials reported on rotavirus diarrhoea of any severity by different G types. There were significantly fewer episodes of rotavirus diarrhoea of any severity in the group receiving RV1 when compared to placebo, regardless of G type (G1, G2, G3, G4, or G9); however, the pooled data for G1 (I2 statistic = 81%) and G9 (I2 statistic = 63%) types showed statistical heterogeneity, see Analysis 1.32.

Rotavirus diarrhoea: severe

There were significantly fewer severe episodes of rotavirus diarrhoea in the RV1 groups compared with placebo in episodes attributed to the G1, G2, G3, G9, and G12 types; see Analysis 1.33. Results were not statistically significant for G4 and G8 types. The pooled data for G8 types showed statistical heterogeneity (I2 statistic = 63%).

1.5.2. Malnourished children

Rotavirus diarrhoea: of any severity

One trial provided data separately as the number of cases of rotavirus diarrhoea of any severity in a subgroup of malnourished children (RV1 Salinas 2005‐LA). RV1 was significantly better than placebo in preventing rotavirus diarrhoea for this subgroup at one year of follow‐up (RR 0.39, 95% CI 0.19 to 0.79; 287 participants, 1 trial, Analysis 1.34).

1.5.3. Children infected with HIV

Rotavirus diarrhoea: of any severity

One safety trial included only confirmed HIV‐positive, asymptomatic or mildly symptomatic children (RV1 Steele 2010a‐ZAF). At one‐month follow‐up, no statistically significant difference between the RV1 and placebo arms for rotavirus diarrhoea was reported (RR 1.00, 95% CI 0.26 to 3.78; 100 participants, 1 trial; Analysis 1.35).

One efficacy trial included children who were infected with HIV or children that had been exposed to HIV, as long as they were not clinically immunosuppressed (e.g. AIDS) at the age of vaccination (six weeks) (RV1 Madhi 2010‐AF). HIV tests were performed on approximately 46% of children from Malawi and 23% of children from South Africa. We did not conduct a specific analysis for this population, but the authors stated that demographic characteristics and the proportion of children who were infected with HIV were similar across the study groups.

1.6 Sensitivity analysis
1.6.1 Primary outcomes with high heterogeneity according to allocation concealment

To investigate heterogeneity for primary outcomes with pooled results where I2 statistic > 75%, we planned to pool data only from studies with low risk of bias for allocation concealment in a sensitivity analysis. We rated all trials at low risk of bias for allocation concealment for the two outcomes where heterogeneity was high (I2 statistic > 75%); see Analysis 1.3 (I2 statistic = 75%) and Analysis 1.4 (I2 statistic = 90%).

1.6.2 Cluster‐randomised trials

Two outcomes (serious adverse events: intussusception, and rotavirus severe diarrhoea at two years) included one cluster‐randomised trial carried out in a high‐mortality country (RV1 Zaman 2017‐BGD). When we excluded data from this trial there was a small but non‐significant change to the effect estimate and 95% CI for Rotavirus diarrhoea: severe (up to 2 years follow‐up) (RR 0.58, 95% CI 0.42 to 0.79, 2764 participants, 1 trial; analysis not shown), and there were no changes to effect estimates or 95% CIs for serious adverse events: intussusception.

‘Summary of findings'

Summary of findings of primary outcomes according to country mortality rate (WHO strata A to E) are presented in summary of findings Table for the main comparison (RV1, low‐mortality countries), and in summary of findings Table 2 (RV1, high‐mortality countries).

2. RV5

2.1. Primary outcomes
2.1.1. Rotavirus diarrhoea: severe

Seven trials provided data for the efficacy of RV5 to prevent severe rotavirus diarrhoea in children; see Analysis 2.1 for one‐year follow‐up and Analysis 2.2 for two years follow‐up. Trials were performed in low‐mortality countries (RV5 Clark 2004‐USA; RV5 Vesikari 2006a‐FIN; RV5 Vesikari 2006b‐INT; RV5 Block 2007‐EU/USA; RV5 Iwata 2013‐JPN; RV5 Mo 2017‐CHN), one trial was split between low‐mortality Vietnam in stratum B (RV5 Zaman 2010‐VNM) and high‐mortality Bangladesh in stratum D (RV5 Zaman 2010‐BGD), and another between high‐mortality Ghana and Mali in stratum D (RV5 Armah 2010‐GHA; RV5 Armah 2010‐MLI) and high‐mortality Kenya in stratum E (RV5 Armah 2010‐KEN). Data below are grouped accordingly.

Low‐mortality countries (WHO strata A and B)

RV5 reduced the number of severe rotavirus diarrhoea cases by 92% at one year (RR 0.08, 95% CI 0.03 to 0.22; 4132 participants, 5 trials) and 82% by two years (RR 0.18, 95% CI 0.08 to 0.39; 7318 participants, 4 trials). Pooled results showed statistical heterogeneity at two‐year follow‐up (I2 statistic = 44%); see Analysis 2.2.

High‐mortality countries (WHO strata D and E)

RV5 reduced the number of severe rotavirus diarrhoea cases by 57% at one year (RR 0.43, 95% CI 0.29 to 0.62; 5916 participants, 4 comparisons from 2 trials) and 41% at two years (RR 0.59, 95% CI 0.43 to 0.82; 5885 participants, 4 comparisons from 2 trials). Pooled results showed statistical heterogeneity at two‐year follow‐up (I2 statistic = 43%); see Analysis 2.2.

2.1.2. All‐cause diarrhoea: severe

Only two trials provided data for the efficacy of RV5 to prevent severe all‐cause diarrhoea in children; see Analysis 2.3 for one‐year follow‐up and Analysis 2.4 for two‐year follow‐up. Trials were performed in high‐mortality countries (RV5 Armah 2010‐GHA; RV5 Armah 2010‐KEN; RV5 Armah 2010‐MLI; RV5 Zaman 2010‐AS). We did not identify any trial that reported on this outcome that was performed in a low‐mortality country.

High‐mortality countries (WHO strata D and E)

There was no statistically significant difference between RV5 and placebo for all‐cause severe diarrhoea at one‐year follow‐up (RR 0.80, 95% CI 0.58 to 1.11; 4085 participants, 3 comparisons from 1 trial). At two‐year follow‐up, RV5 reduced severe cases by 15% (RR 0.85, 95% CI 0.75 to 0.98; 5977 participants, 4 comparisons from 2 trials). Pooled results showed statistical heterogeneity at one‐year follow‐up (I2 statistic = 46%); see Analysis 2.3.

2.1.3. All‐cause death

Eleven trials reported on all‐cause death, in most trials as the number of deaths (RV5 Armah 2010‐AF; RV5 Iwata 2013‐JPN; RV5 Lawrence 2012‐CHN; RV5 Levin 2017‐AF; RV5 Merck[009] 2005‐USA; RV5 Mo 2017‐CHN; RV5 Vesikari 2006a‐FIN; RV5 Vesikari 2006b‐INT; RV5 Zaman 2010‐AS), and in two trials as fatal serious adverse events (RV5 Block 2007‐EU/USA; RV5 Ciarlet 2009‐EU). We pooled the number of deaths and fatal serious adverse events; see Analysis 2.5. We present details of causes of death for each trial in Appendix 9. Most trials were performed in low‐mortality countries, with one trial split between low‐mortality Vietnam in stratum B (RV5 Zaman 2010‐VNM) and high‐mortality Bangladesh in stratum D (RV5 Zaman 2010‐BGD), and another between high‐mortality Ghana and Mali in stratum D (RV5 Armah 2010‐GHA; RV5 Armah 2010‐MLI) and high‐mortality Kenya in stratum E (RV5 Armah 2010‐KEN).

Low‐mortality countries (WHO strata A and B)

There was no statistically significant difference in all‐cause death between RV5 and placebo arm (RR 1.13, 95% CI 0.65 to 1.96; 77,642 participants, 9 trials; Analysis 2.5).

High‐mortality countries (WHO strata D and E)

There was no statistically significant difference in all‐cause death between the two arms (RR 0.92, 95% CI 0.68 to 1.24; 6806 participants, 5 comparisons from 3 trials; Analysis 2.5).

2.1.4. All serious adverse events

Serious adverse events were reported in 11 trials, in trials in low‐mortality countries (RV5 Block 2007‐EU/USA; RV5 Ciarlet 2009‐EU; RV5 Iwata 2013‐JPN; RV5 Kim 2008‐KOR; RV5 Lawrence 2012‐CHN; RV5 Mo 2017‐CHN; RV5 Vesikari 2006b‐INT; RV5 Zaman 2010‐VNM), and in high‐mortality countries (RV5 Armah 2010‐GHA; RV5 Armah 2010‐KEN; RV5 Armah 2010‐MLI; RV5 Dhingra 2014‐IND; RV5 Levin 2017‐AF; RV5 Zaman 2010‐BGD); see Analysis 2.6.

Low‐mortality countries (WHO strata A and B)

Pooled results showed no statistically significant difference in the number of serious adverse events in the RV5 group compared with the placebo group (RR 0.93, 95% CI 0.86 to 1.02; 75,672 participants, 8 trials; Analysis 2.6). In addition, in a separate cohort of RV5 Lawrence 2012‐CHN that vaccinated 24 older children (aged two to six years) with one‐dose RV5 there were no serious adverse events reported.

High‐mortality countries (WHO strata D and E)

Pooled results showed no statistically significant difference in the number of serious adverse events in the RV5 group compared with the placebo group (RR 0.92, 95% CI 0.66 to 1.28; 6830 participants, 6 comparisons from 4 trials; Analysis 2.6).

2.1.5. Serious adverse events: intussusception

Thirteen trials reported cases of intussusception. Trials were performed in low‐mortality countries (RV5 Block 2007‐EU/USA; RV5 Ciarlet 2009‐EU; RV5 Clark 2003‐USA; RV5 Clark 2004‐USA; RV5 Iwata 2013‐JPN; RV5 Kim 2008‐KOR; RV5 Lawrence 2012‐CHN; RV5 Merck[009] 2005‐USA; RV5 Mo 2017‐CHN; RV5 Vesikari 2006a‐FIN; RV5 Vesikari 2006b‐INT; RV5 Zaman 2010‐VNM), and in high‐mortality countries (RV5 Armah 2010‐GHA; RV5 Armah 2010‐KEN; RV5 Armah 2010‐MLI; RV5 Zaman 2010‐BGD); see Analysis 2.7.

Low‐mortality countries (WHO strata A and B)

Fourteen cases of intussusception were reported in a total of 38,321 children in the RV5 arm compared with 20 cases of intussusception in 36,553 children in the placebo arm. Pooled results showed no increased risk of intussusception in children receiving RV5 when compared to placebo (RR 0.77, 95% CI 0.41 to 1.45; 78,907 participants, 12 trials; Analysis 2.7).

High‐mortality countries (WHO strata D and E)

There were no reported cases of intussusception in a total of 3294 children in the RV5 arm and 3294 children in the placebo arm (4 comparisons from 2 trials).

2.2. Secondary outcomes
2.2.1. Rotavirus diarrhoea: of any severity

Nine trials provided data for the efficacy of RV5 to prevent rotavirus diarrhoea of any severity in children; see Analysis 2.8 for one‐year follow‐up and Analysis 2.9 for two‐year follow‐up. Trials were performed in low‐mortality countries (RV5 Block 2007‐EU/USA; RV5 Clark 2003‐USA; RV5 Clark 2004‐USA; RV5 Iwata 2013‐JPN; RV5 Mo 2017‐CHN; RV5 Vesikari 2006a‐FIN; RV5 Vesikari 2006b‐INT), and in high‐mortality countries (RV5 Armah 2010‐GHA; RV5 Armah 2010‐KEN; RV5 Armah 2010‐MLI; RV5 Zaman 2010‐AS). Data below are grouped accordingly.

Low‐mortality countries (WHO strata A and B)

RV5 reduced the number of cases of rotavirus diarrhoea by 70% at one year (RR 0.30, 95% CI 0.25 to 0.37; 8644 participants, 5 trials; Analysis 2.8) and by 66% during the second year (RR 0.34, 95% CI 0.26 to 0.43; 6144 participants, 3 trials; Analysis 2.9).

High‐mortality countries (WHO strata D and E)

RV5 reduced the number of cases of rotavirus diarrhoea by 48% at one year (RR 0.52, 95% CI 0.28 to 0.94; 4806 participants, 3 comparisons from 1 trial; Analysis 2.8) and by 39% during the second year (RR 0.61, 95% CI 0.45 to 0.83; 6744 participants, 4 comparisons from 2 trials; Analysis 2.9). Pooled results were significantly heterogenous at one‐year (I2 statistic = 67%; see Analysis 2.8) and at two‐year (I2 statistic = 69%; see Analysis 2.9) follow‐up.

2.2.2. All‐cause diarrhoea: of any severity

One trial performed in high‐mortality Kenya (RV5 Armah 2010‐KEN) provided data for the efficacy of RV5 to prevent all‐cause diarrhoea of any severity; see Analysis 2.10 for one‐year and Analysis 2.11 for two‐year follow‐up.

High‐mortality countries (WHO stratum E)

There was no statistically significant difference between RV5 and placebo for any severity all‐cause diarrhoea at one year (RR 0.82, 95% CI 0.61 to 1.11; 1059 participants, 1 trial; Analysis 2.10) or at two‐year follow‐up (RR 0.89, 95% CI 0.68 to 1.16; 1059 participants, 1 trial; Analysis 2.11).

All‐cause hospitalization

Data on all‐cause hospitalization were provided from one trial carried out in Botswana, Tanzania, Zambia, and Zimbabwe (RV5 Levin 2017‐AF).

There was no statistically significant difference between RV5 and placebo for all‐cause hospitalization at two‐year follow‐up (RR 1.21, 95% CI 0.42 to 3.49; 202 participants, 1 trial; Analysis 2.12).

2.2.3. Rotavirus diarrhoea: requiring hospitalization or medical attention

RV5 reduced hospitalizations due to rotavirus diarrhoea episodes by 96% at one year of follow‐up (RR 0.04, 95% CI 0.02 to 0.10; 57,134 participants, 1 trial; Analysis 2.13).

RV5 reduced the number of children requiring medical attention at one year of follow‐up by 93% compared to placebo (RR 0.07, 95% CI 0.04 to 0.12; 57,134 participants, 1 trial; Analysis 2.14).

Data for medical attention and hospitalization rates due to all‐cause diarrhoea were not estimable.

2.2.4. Reactogenicity

The incidence of fever (Analysis 2.15), diarrhoea (Analysis 2.16), and vomiting (Analysis 2.17) were evaluated after the first dose, second dose, and third dose, and at the end of the follow‐up period. We found no statistically significant differences between the RV5 and placebo groups for any of the reactogenicity outcomes and time points. We noted significant heterogeneity for the pooled post‐first dose data on fever (I2 statistic = 61%).

2.2.5. Adverse events that require discontinuation of vaccination schedule

Ten trials reported the number of adverse events leading to discontinuation of the vaccination schedule, with no statistically significant difference between RV5 and placebo (RR 0.89, 95% CI 0.57 to 1.39; 15,471 participants, 10 trials; Analysis 2.18).

2.3. Immunogenicity

RV5 immunogenicity was measured by rotavirus vaccine virus shedding (5 trials, Analysis 2.19) and seroconversion (10 trials, Analysis 2.20) after the third vaccine dose. We decided not to pool the data, however, because of significant heterogeneity (I2 statistic = 80% and 87%, respectively).

2.4. Dropouts before the end of trial

Similar numbers of children taking RV5 and placebo dropped out from trials before they ended (RR 0.98, 95% CI 0.90 to 1.08; 85,855 participants, 13 trials; Analysis 2.21).

2.5. Subgroup analyses
2.5.1. G type

Rotavirus diarrhoea: of any severity

When the analyses were stratified by the G type (Analysis 2.22), there were fewer episodes of rotavirus diarrhoea in the RV5 group compared to the placebo group for the G1 type (RR 0.26, 95% CI 0.21 to 0.32; 11,022 participants, 4 trials), the G2 type (RR 0.35, 95% CI 0.16 to 0.78; 9907 participants, 3 trials), and the G9 type (RR 0.33, 95% CI 0.20 to 0.54; 9537 participants, 2 trials). The results were not statistically significant for G3 (RR 0.40, 95% CI 0.08 to 2.02; 11,022 participants, 4 trials) or for G4 (RR 0.41, 95% CI 0.13 to 1.33; 9907 participants, 3 trials).

Rotavirus diarrhoea: severe

There were significantly fewer severe episodes of rotavirus diarrhoea in the RV5 groups for G4 (RR 0.12, 95% CI 0.03 to 0.46; 76,606 participants, 3 trials) and G9 (RR 0.13, 95% CI 0.05 to 0.34; 76,606 participants, 3 trials). Pooled results were not significant for G1 (RR 0.23, 95% CI 0.03 to 1.74; 76,606 participants, 3 trials), G2 (RR 0.41, 95% CI 0.13 to 1.37; 76,606 participants, 3 trials), and for G3 (RR 0.38, 95% CI 0.05 to 2.74; 76,606 participants, 3 trials). The pooled data for G1 (I2 statistic = 97%) and G3 (I2 statistic = 64%) types showed statistical heterogeneity.

2.5.2. HIV‐infected children

One trial (RV5 Armah 2010‐AF) performed HIV tests for 89% of participants and reported outcomes for HIV‐infected children (38/1158); another trial (RV5 Levin 2017‐AF) included and reported outcomes for HIV‐exposed but uninfected and HIV‐infected children. We included only HIV‐infected children from this study in this subgroup analysis (Analysis 2.24).

Rotavirus diarrhoea: severe (up to two years of follow‐up)

1/21 children in the vaccine arm, and 0/17 children in the placebo arm had severe rotavirus diarrhoea at two‐year follow‐up; there was no statistically significant difference detected between the two treatment arms (1 trial).

All‐cause diarrhoea: severe (up to two years of follow‐up)

5/21 children in the vaccine arm, and 1/17 children in the placebo arm had severe all‐cause diarrhoea at two‐year follow‐up; there was no statistically significant difference detected between the two treatment arms (1 trial).

All‐cause death

9/58 children in the vaccine arm, and 6/56 children in the placebo arm died; there was no statistically significant difference between the two arms (2 trials).

Serious adverse events (1 ‐ 14 days after any dose)

10/58 children in the vaccine arm, and 6/55 children in the placebo arm had a serious adverse event; there was no statistically significant difference between the two arms (2 trials).

2.6 Sensitivity analysis
2.6.1 Primary outcomes with high heterogeneity according to allocation concealment

There were no primary outcomes with high heterogeneity (I2 statistic > 75%).

‘Summary of findings'

Summary of findings of primary outcomes according to country mortality rate (WHO strata A to E) are presented in summary of findings Table 3 (RV5, low‐mortality countries), and in summary of findings Table 4 (RV5, high‐mortality countries).

3. Rotavac

3.1. Primary outcomes
3.1.1. Rotavirus diarrhoea: severe

High‐mortality countries (WHO stratum D)

One trial conducted in India provided data for the efficacy of Rotavac to prevent severe rotavirus diarrhoea in children. Rotavac reduced severe rotavirus diarrhoea cases by 57% at one year (RR 0.43, 95% CI 0.30 to 0.60; 6799 participants, 1 trial; Analysis 3.1) and by 54% by two years (RR 0.46, 95% CI 0.35 to 0.60; 6541 participants, 1 trial; Analysis 3.2).

3.1.2. All‐cause diarrhoea: severe

High‐mortality countries (WHO stratum D)

One trial conducted in India provided data for the efficacy of Rotavac to prevent severe all‐cause diarrhoea in children. The trial showed a reduction in the number of severe cases of diarrhoea with Rotavac compared to placebo at one year by 16% (RR 0.84, 95% CI 0.71 to 0.98; 6799 participants, 1 trial; Analysis 3.3).

3.1.3. All‐cause death

High‐mortality countries (WHO stratum D)

Two trials conducted in India reported on all‐cause death. There was no statistically significant difference in all‐cause death between Rotavac and placebo (RR 0.92, 95% CI 0.52 to 1.62; 8155 participants Analysis 3.4). We present details of causes of death for each trial in Appendix 9.

3.1.4. All serious adverse events

High‐mortality countries (WHO stratum D)

Serious adverse events were reported in three trials conducted in India. Pooled results showed no statistically significant difference in the number of serious adverse events in the Rotavac group compared with the placebo group (RR 0.93, 95% CI 0.85 to 1.02; 8210 participants, 3 trials; Analysis 3.5).

3.1.5. Serious adverse events: intussusception

High‐mortality countries (WHO stratum D)

Four trials conducted in India reported on cases of intussusception. Eight cases of intussusception were reported in a total of 5764 children in the Rotavac arm compared with three cases of intussusception in 2818 children in the placebo arm. Pooled results showed no increased risk of intussusception in children receiving Rotavac when compared to placebo (RR 1.33, 95% CI 0.35 to 5.02; 8582 participants, 4 trials; Analysis 3.6).

3.2. Secondary outcomes
3.2.1. Rotavirus diarrhoea: of any severity

One trial provided data for the efficacy of Rotavac to prevent rotavirus diarrhoea of any severity in children. Rotavac reduced the number of cases of rotavirus diarrhoea of any severity by 34% at both one‐year (RR 0.66, 95% CI 0.56 to 0.78; 6799 participants, 1 trial; Analysis 3.7) and two‐year follow‐up (RR 0.66, 95% CI 0.57 to 0.76; 6541 participants, 1 trial; Analysis 3.8).

3.2.2. Rotavirus diarrhoea: requiring medical attention

Rotavac reduced the number of children requiring medical attention due to rotavirus diarrhoea at one year of follow‐up by 31% compared to placebo (RR 0.69, 95% CI 0.58 to 0.81; 6799 participants, 1 trial; Analysis 3.9).

3.2.3. Reactogenicity

The incidences of fever (Analysis 3.10), diarrhoea (Analysis 3.11), and vomiting (Analysis 3.12) were evaluated after the first dose in two trials, second dose in one trial, and third dose in one trial. We found no statistically significant differences between the Rotavac and placebo groups for most of the reactogenicity outcomes and time points, except for diarrhoea, which demonstrated an increase with Rotavac compared to placebo after the second dose (RR 1.55, 95% CI 1.00 to 2.41; 356 participants) and third dose (RR 4.09, 95% CI 2.11 to 7.92; 358 participants).

3.2.4. Immunogenicity

Rotavac was more immunogenic than placebo when measured by vaccine virus shedding at the end of follow‐up (RR 9.86, 95% CI 2.58 to 37.63; 427 participants, 2 trials, Analysis 3.13). It was also more immunogenic when measured by seroconversion at all time points (Analysis 3.14): after the first dose (RR 3.58, 95% CI 2.03 to 6.29; 121 participants, 1 trial), after the second dose (RR 2.97, 95% CI 1.78 to 4.98; 117 participants, 1 trial), and after the third dose (RR 2.82, 95% CI 2.26 to 3.51; 1699 participants, 3 trials).

3.2.5. Dropouts before the end of trial

Similar numbers of children taking Rotavac or placebo dropped out from trials before they ended (RR 0.81, 95% CI 0.62 to 1.06; 8215 participants, 3 trials; Analysis 3.15).

3.3. Subgroup analyses
3.3.1. G type

Rotavirus diarrhoea: severe

One trial reported severe cases of rotavirus diarrhoea by G and P type (VAC Bhandari 2014‐IND).

At one‐year follow‐up (Analysis 3.16) there were significantly fewer severe episodes of rotavirus diarrhoea in the Rotavac groups for G2P[4] (RR 0.39, 95% CI 0.22 to 0.69; 6541 participants) and G12P[6] (RR 0.31, 95% CI 0.13 to 0.74; 6541 participants); results were not significantly different between Rotavac and placebo for G1P[8] (RR 0.66, 95% CI 0.36 to 1.20; 6541 participants) and G12P[8] (RR 0.30, 95% CI 0.07 to 1.26; 6541 participants).

At two‐year follow‐up (Analysis 3.17) there were significantly fewer severe episodes of rotavirus diarrhoea in the Rotavac groups for G1P[8] (RR 0.59, 95% CI 0.38 to 0.93; 6541 participants), G2P[4] (RR 0.37, 95% CI 0.23 to 0.62; 6541 participants), G12P[6] (RR 0.31, 95% CI 0.13 to 0.74; 6541 participants), and G12P[8] (RR 0.31, 95% CI 0.10 to 0.96; 6541 participants).

The included Rotavac trials did not report separate data on immunocompromised or malnourished subgroups.

3.4 Sensitivity analyses
3.4.1 Primary outcomes with high heterogeneity according to allocation concealment

There were no primary outcomes with high heterogeneity (I2 statistic > 75%).

‘Summary of findings'

Summary of findings of primary outcomes are presented in summary of findings Table 5 (Rotavac, high‐mortality countries),

Discussion

Rotavirus vaccines have been under development since the 1980s, and to date three have been prequalified by the WHO (RV1, RV5 and Rotavac). Three additional rotavirus vaccines are licensed for use in individual countries (LLR, Rotasiil, and Rotavin, see Appendix 10). RRV‐TV (RotaShield) has not been used since 1999. The three vaccines prequalified by the WHO (RV1, RV5, Rotavac), and currently in use, are the focus of this review.

Summary of main results

We included 55 trials with a total of 216,480 participants, that evaluated RV1 (36 trials), RV5 (15 trials), and Rotavac (4 trials). Our analysis stratified the primary outcomes by WHO mortality strata (high‐mortality countries, with high child mortality; and low‐mortality, with low or very low child mortality; WHO 1999).

The trials were not designed or powered to detect an effect on preventing death or on the occurrence of possible rare serious adverse events, such as intussusception.

1. RV1 in countries with low child mortality (WHO strata A and B)

Fourteen trials were conducted in Asia, six in Europe, four in Latin America, four in North America, and one in Europe and Latin America.

In infants under one year

RV1 prevents 84% of cases of severe rotavirus diarrhoea: RR 0.16, 95% CI 0.09 to 0.26; 43,779 participants, 7 trials; high‐certainty evidence.

RV1 prevents 41% of cases of severe all‐cause diarrhoea: RR 0.59, 95% CI 0.47 to 0.74; 28,053 participants, 3 trials; moderate‐certainty evidence.

In children up to two years

RV1 prevents 82% of cases of severe rotavirus diarrhoea: RR 0.18, 95% CI 0.14 to 0.23; 36,002 participants, 9 trials; high‐certainty evidence.

RV1 prevents 37% of severe all‐cause diarrhoea episodes: Rate ratio 0.63, 95% CI 0.56 to 0.71; 39,091 participants, 2 trials; moderate‐certainty evidence.

For all‐cause death, an effect of the vaccine has not been shown: RR 1.22, 95% CI 0.87 to 1.71; 97,597 participants, 22 trials; low‐certainty evidence.

For serious adverse events, children receiving RV1 had 12% fewer events than those receiving placebo: RR 0.88, 95% CI 0.83 to 0.93; 96,233 participants, 24 trials; high‐certainty evidence.

For intussusception, RV1 was not associated with a higher risk: RR 0.69, 95% CI 0.45 to 1.04; 96,513 participants, 17 trials; low‐certainty evidence.

See summary of findings Table for the main comparison.

2. RV1 in countries with high child mortality (WHO strata D and E)

Two trials were conducted in Bangladesh, one in India, one in Peru, three in South Africa, and one in South Africa and Malawi.

In infants under one year

RV1 prevents 63% of cases of severe rotavirus diarrhoea: RR 0.37, 95% CI 0.23 to 0.60; 6114 participants, 3 trials; high‐certainty evidence.

RV1 prevents 27% of cases of severe all‐cause diarrhoea: RR 0.73, 95% CI 0.56 to 0.95; 5639 participants, 2 trials; high‐certainty evidence.

In children up to two years

RV1 prevents 35% of cases of severe rotavirus diarrhoea: RR 0.65, 95% CI 0.51 to 0.83; 13,768 participants, 2 trials; high‐certainty evidence.

RV1 prevents 17% of cases of severe all‐cause diarrhoea: RR 0.83, 95% CI 0.72 to 0.96; 2764 participants, 1 trial; moderate‐certainty evidence.

For all‐cause death, an effect of the vaccine has not been shown: RR 0.88, 95% CI 0.64 to 1.22; 8181 participants, 8 trials; low‐certainty evidence.

For serious adverse events, an effect of the vaccine has not been shown: RR 0.89, 95% CI 0.76 to 1.04; 7481 participants, 7 trials; high‐certainty evidence.

For intussusception, RV1 was not associated with a higher risk: RR 1.49, 95% CI 0.06 to 36.63; 17,492 participants, 4 trials; low‐certainty evidence.

See summary of findings Table 2.

3. RV5 in countries with low child mortality (WHO strata A and B)

Three trials were conducted in Asia, two in Europe, three in North America, one in Europe and the USA, one in Europe and the Americas.

In infants under one year

RV5 prevents 92% of cases of severe rotavirus diarrhoea: RR 0.08, 95% CI 0.03 to 0.22; 4132 participants, 5 trials; moderate‐certainty evidence.

We found no RV5 trials that reported on severe all‐cause diarrhoea.

In children up to two years

RV5 prevents 82% of cases of severe rotavirus diarrhoea: RR 0.18, 95% CI 0.08 to 0.39; 7318 participants, 4 trials; moderate‐certainty evidence.

We found no RV5 trials that reported on severe all‐cause diarrhoea.

For all‐cause death, an effect of the vaccine has not been shown: RR 1.13, 95% CI 0.65 to 1.96; 77,642 participants, 9 trials; low‐certainty evidence.

For serious adverse events, an effect of the vaccine has not been shown: RR 0.93, 95% CI 0.86 to 1.02; 75,672 participants, 8 trials; high‐certainty evidence.

For intussusception, RV5 was not associated with a higher risk: RR 0.77, 95% CI 0.41 to 1.45; 78,907 participants, 12 trials; low‐certainty evidence.

See summary of findings Table 3.

4. RV5 in countries with high child mortality (WHO strata D and E)

Two trials were conducted in Asia and two in Africa.

In infants under one year

RV5 prevents 57% of cases of severe rotavirus diarrhoea: RR 0.43, 95% CI 0.29 to 0.62; 5916 participants, 2 trials; high‐certainty evidence.

Data on severe all‐cause diarrhoea was reported in one trial. This suggested a protective effect, but the results were not statistically significant: RR 0.80, 95% CI 0.58 to 1.11; 4085 participants, 1 trial; moderate‐certainty evidence.

In children up to two years

RV5 prevents 41% of cases of severe rotavirus diarrhoea: RR 0.59, 95% CI 0.43 to 0.82; 5885 participants, 2 trials; high‐certainty evidence.

RV5 prevents 15% of cases of severe all‐cause diarrhoea: RR 0.85, 95% CI 0.75 to 0.98; 5977 participants, 2 trials; high‐certainty evidence.

For all‐cause death, an effect of the vaccine has not been shown: RR 0.92, 95% CI 0.68 to 1.24; 6806 participants, 3 trials; low‐certainty evidence.

For serious adverse events, an effect of the vaccine has not been shown: RR 0.92, 95% CI 0.66 to 1.28; 6830 participants, 4 trials; moderate‐certainty evidence.

For intussusception, RV5 was not associated with a higher risk: no cases were reported, 6588 participants, 2 trials; low‐certainty evidence.

See summary of findings Table 4.

5. Rotavac in countries with high child mortality (WHO stratum D)

Four trials were conducted in India.

In infants under one year

Rotavac prevents 57% of cases of severe rotavirus diarrhoea: RR 0.43, 95% CI 0.30 to 0.60; 6799 participants, 1 trial; moderate‐certainty evidence.

In children up to two years

Rotavac prevents 54% of cases of severe rotavirus diarrhoea: RR 0.46, 95% CI 0.35 to 0.60; 6541 participants, 1 trial; moderate‐certainty evidence.

Rotavac prevents 16% of cases of severe all‐cause diarrhoea: RR 0.84, 95% CI 0.71 to 0.98; 6799 participants, one trial; moderate‐certainty evidence.

For all‐cause death, an effect of the vaccine has not been shown: RR 0.92, 95% CI 0.52 to 1.62; 8155 participants, 2 trials; very low‐certainty evidence.

For serious adverse events, an effect of the vaccine has not been shown: RR 0.93, 95% CI 0.85 to 1.02; 8210 participants, 3 trials; moderate‐certainty evidence.

For intussusception, Rotavac was not associated with a higher risk: RR 1.33, 95% CI 0.35 to 5.02; 8582 participants, 4 trials; very low‐certainty evidence.

See summary of findings Table 5.

Overall completeness and applicability of evidence

We carried out this systematic review using RCTs. All the included trials were placebo‐controlled, except for two RV1 trials that compared vaccine to no intervention (RV1 Colgate 2016‐BGD; RV1 Zaman 2017‐BGD). We could not evaluate potential herd protection afforded by vaccination. The trials provided only limited data for special groups of children, such as malnourished or immunocompromised children.

Efficacy by setting

RV1 and RV5 were highly efficacious in reducing severe rotavirus diarrhoea episodes in low‐mortality countries; widespread roll‐out of rotavirus vaccines has led to major reductions in rotavirus hospitalizations in such settings (Hungerford 2015; Jonesteller 2017). In contrast, trials of RV1 and RV5 in high‐mortality countries in Africa and Asia demonstrated a relatively lower vaccine efficacy. However, because of the higher burden of rotavirus disease in such countries, the absolute number of events prevented by vaccination is greater than in low‐mortality countries (RV1 Madhi 2010‐AF).

Efficacy by age

Results from RV1 and RV5 found higher vaccine efficacy against severe rotavirus diarrhoea in the first year compared to the cumulative efficacy for the first and second years. The efficacy was lower but the differences between the first and second years were greater in high‐mortality (RV1: 63% up to one year versus 54% up to two years; RV5: 57% versus 41%) compared to low‐mortality countries (RV1: 84% up to one year versus 82% up to two years; RV5: 92% versus 82%). Trials with Rotavac were not carried out in any low‐mortality country.

Reduced vaccine efficacy in high‐mortality countries in trials reporting two years of follow‐up could be explained either by waning of vaccine‐induced immunity, or some protection in the placebo group resulting from more frequent exposure to natural rotavirus infection (RV1 Madhi 2010‐AF). Post‐introduction studies have shown reduced effectiveness in the second year of life in some, but not all, high‐burden settings (Bar‐Zeev 2015; Groome 2014). Additional vaccine doses have been explored to extend the duration of protection in high disease‐burden settings (Cunliffe 2016).

Efficacy by schedule

Children in trials performed in low‐mortality countries received the vaccines according to the country's immunization schedule. Trials performed in high‐mortality countries examined the efficacy of RV1 when administered at 10 to 14 weeks of age, a later age than is recommended in the Expanded Programme on Immunization (EPI) schedule. However, the 6‐ and 10‐week RV1 schedule used in EPI programmes has now been extensively evaluated following vaccine roll‐out in high‐mortality countries in Africa, with effectiveness comparable to efficacy trial estimates (Bar‐Zeev 2015).

All‐cause diarrhoea

The impact of rotavirus vaccination on severe all‐cause diarrhoea from a public health perspective is important, as laboratories in low‐income countries may not routinely test for rotavirus infection. The effect on all‐cause diarrhoea is a function of the contribution of rotavirus to all diarrhoea and the efficacy of the vaccine against rotavirus. Surprisingly, few trials reported vaccine efficacy against all‐cause diarrhoea. Vaccine efficacy against all‐cause diarrhoea of any severity was lower, meaning that vaccination may not have a noticeable impact on milder episodes of diarrhoea occurring in the community (Hungerford 2018).

Mortality data

The included trials were not individually powered to detect a mortality effect. This review did not detect a difference in the number of deaths for children receiving any of the vaccines or placebo. Two post‐vaccine implementation national surveillance studies from Mexico and Brazil reported that the introduction of RV1 into the national immunization programme was associated with a decline in the number of diarrhoea‐related deaths (Do Carmo 2011; Richardson 2010) in comparison with historical controls. A study from rural Malawi showed that diarrhoea deaths reduced by a third following RV1 introduction (Bar‐Zeev 2018).

Safety data

There was no detectable difference in the number of cases of intussusception for children receiving vaccine or placebo. While both RV1 and RV5 have been associated with a low risk of intussusception in post‐marketing studies in Europe, Americas and Australia, the benefits of vaccination are considered to outweigh the risk of vaccine‐associated intussusception (Yen 2016). However, the risk of intussusception after administration of RV1 was not higher than the background risk of intussusception in seven lower‐income sub‐Saharan African countries (Tate 2018).

Subgroup analyses

Rotavirus G‐types

All three rotavirus vaccines showed efficacy against most of the specific rotavirus G‐types that were assessed (G1, G2, G3, G4, G8, G9, and G12), although results were often inconsistent between different countries and imprecise due to few events.

Immunocompromised children

One RV1 trial and two RV5 trials reported on immunocompromised children, all exposed to or infected with HIV. We found no differences for efficacy or safety, but samples were not sufficiently powered. It is now strongly recommended that all HIV‐infected or HIV‐exposed infants be vaccinated with oral rotavirus vaccine, unless severely immunocompromised (Calles 2010). While we lack specific information on many immunodeficiencies, infants with known severe combined immunodeficiency should not receive live rotavirus vaccine (Pinto 2016; Vesikari 2015).

Children with malnutrition

One RV1 trial (RV1 Salinas 2005‐LA) found that RV1 was significantly better than placebo in preventing rotavirus diarrhoea in a subgroup of malnourished children.

Certainty of the evidence

The trials included in this updated review were placebo‐controlled (53 trials) or compared vaccine to no intervention (RV1 Colgate 2016‐BGD; RV1 Zaman 2017‐BGD), were conducted in Latin America, North America, Europe, Asia, and Africa, and the largest included over 60,000 children (RV1 Ruiz‐Palac 06‐LA/EU; RV5 Vesikari 2006b‐INT); we identified the need for such trials in the original version of the review (Soares‐Weiser 2004). However, most children were followed for safety outcomes only.

The certainty of the evidence for efficacy outcomes (rotavirus diarrhoea of any severity and severe, and all‐cause diarrhoea of any severity and severe) was either high or moderate. This was because most trials were assessed at low risk of bias, especially more recent trials, and pooled samples were usually large enough to generate more precise estimates. When we downgraded efficacy outcomes to moderate certainty, this was due to selective reporting bias (only half of the studies reporting on severe rotavirus diarrhoea reported on severe all‐cause diarrhoea), imprecision (low number of events), attrition bias (incomplete outcome data were not clearly reported), or indirectness (only one study carried out in one high‐mortality country or neighbouring high‐mortality countries makes it difficult to generalize to any high‐mortality country).

The certainty of the evidence for all‐cause mortality was low because the trials were not powered to detect an effect on mortality, and results were consequently imprecise with wide 95% CIs.

The certainty of the evidence for all serious adverse events was mostly high but downgraded to moderate for RV5 in high‐mortality countries due to imprecise results, and for Rotavac due to indirectness (all trials were carried out in India). For the rare serious adverse event intussusception, evidence was of low certainty for RV1 and RV5 due to imprecision because trials were not powered to detect an association between RV1 and intussusception. For Rotavac evidence on intussusception was of very low certainty, due to imprecision and indirectness as previously described.

Potential biases in the review process

We stratified all analyses by WHO mortality strata, which may not reflect the current situation in the member countries. The use of the strata may not be sensitive enough to show differences at the country level, and perhaps stratifying by prevalence/burden of rotavirus may be a better method to group the analyses. In addition, not all countries are represented by the studies performed, and some strata (e.g. C) are lacking sufficient data.

Agreements and disagreements with other studies or reviews

We identified three systematic reviews of RCTs evaluating RV1 or RV5 or both that have been conducted since the 2012 update of this Cochrane Review:

  • Lamberti 2016 included RCTs and observational studies and evaluated region‐specific effectiveness of RV1, RV5 and Rotavac. The systematic review found that rotavirus vaccination was both efficacious and effective in preventing rotavirus diarrhoea, severe rotavirus diarrhoea and rotavirus hospitalizations among children under five across all regions, with higher efficacy in more developed regions.

  • Velázquez 2017 included RCTs and post‐licensure observational studies from Latin America and the Caribbean, and found that RV1 reduced the risk of any‐severity rotavirus‐related gastroenteritis by 65% and of severe gastroenteritis by 82% versus placebo. Both RV1 and RV5 vaccines significantly reduced the risk of hospitalization and emergency visits by 85% for RV1 and by 90% for RV5. Vaccination with RV5 or RV1 did not increase the risk of death, intussusception, or other severe adverse events.

  • Buyse 2014 presented an integrated meta‐analysis of safety and reactogenicity data of 28 RV1 RCTs and found that RV1 has a reactogenicity and safety profile similar to placebo.

The findings of these systematic reviews agree with the findings of our review, although the scope of these reviews was narrower; they reviewed efficacy or safety only, or were limited to a specific geographical region, or reviewed only one of the vaccines. Consequently, we included more trials in our review. Finally, the major findings of this review update, including new evidence from 14 trials of RV1, RV5, and Rotavac, are not significantly different from the previous Soares‐Weiser 2012b review.

Relationship to current policies

The data in this review support the WHO's Strategic Advisory Group of Experts (SAGE) on Immunization’s recommendation for "the inclusion of rotavirus vaccination of infants into all national immunization programmes" with a stronger recommendation for countries where "diarrhoeal deaths account for ≥10% of mortality among children aged <5 years" (SAGE 2009).

A simplified diagram of the location of rotavirus structural proteins (source: Graham Cohn, Wikipedia (public domain image)): Rotaviruses are segmented, double‐stranded RNA viruses. The mature, triple‐layered virus particle comprises a core (which contains the viral genome), a middle layer (comprised of viral protein (VP)6, and an outer layer (comprised of VP7 and VP4) as shown in the figure. VP6 defines rotavirus group, and most rotaviruses that infect humans are of group A. The two outer capsid proteins independently induce neutralizing antibodies: VP7, a glycoprotein, defines G‐serotype; and the protease‐sensitive VP4 protein defines P‐serotype. G‐serotype determined by serological methods correlates precisely with G‐genotype obtained through molecular assays, whereas there is an imperfect correlation of P‐serotype and P‐genotype; P‐genotype is thus included in square brackets.
Figures and Tables -
Figure 1

A simplified diagram of the location of rotavirus structural proteins (source: Graham Cohn, Wikipedia (public domain image)): Rotaviruses are segmented, double‐stranded RNA viruses. The mature, triple‐layered virus particle comprises a core (which contains the viral genome), a middle layer (comprised of viral protein (VP)6, and an outer layer (comprised of VP7 and VP4) as shown in the figure. VP6 defines rotavirus group, and most rotaviruses that infect humans are of group A. The two outer capsid proteins independently induce neutralizing antibodies: VP7, a glycoprotein, defines G‐serotype; and the protease‐sensitive VP4 protein defines P‐serotype. G‐serotype determined by serological methods correlates precisely with G‐genotype obtained through molecular assays, whereas there is an imperfect correlation of P‐serotype and P‐genotype; P‐genotype is thus included in square brackets.

Navigate to figure in ReviewOpen in new tab

PRISMA diagram.
Figures and Tables -
Figure 2

PRISMA diagram.

Navigate to figure in ReviewOpen in new tab

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Figures and Tables -
Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.
Figures and Tables -
Figure 4

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Navigate to figure in ReviewOpen in new tab

Funnel plot of comparison: 1 RV1 versus placebo, outcome: 1.1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).
Figures and Tables -
Figure 5

Funnel plot of comparison: 1 RV1 versus placebo, outcome: 1.1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.1

Comparison 1 RV1 versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.2

Comparison 1 RV1 versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.3

Comparison 1 RV1 versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.4

Comparison 1 RV1 versus placebo, Outcome 4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 5 All‐cause diarrhoea: severe episodes (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.5

Comparison 1 RV1 versus placebo, Outcome 5 All‐cause diarrhoea: severe episodes (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 6 All‐cause diarrhoea: severe episodes (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.6

Comparison 1 RV1 versus placebo, Outcome 6 All‐cause diarrhoea: severe episodes (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 7 All‐cause death.
Figures and Tables -
Analysis 1.7

Comparison 1 RV1 versus placebo, Outcome 7 All‐cause death.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 8 All serious adverse events.
Figures and Tables -
Analysis 1.8

Comparison 1 RV1 versus placebo, Outcome 8 All serious adverse events.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 9 Serious adverse events: intussusception.
Figures and Tables -
Analysis 1.9

Comparison 1 RV1 versus placebo, Outcome 9 Serious adverse events: intussusception.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 10 Serious adverse events: Kawasaki disease.
Figures and Tables -
Analysis 1.10

Comparison 1 RV1 versus placebo, Outcome 10 Serious adverse events: Kawasaki disease.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 11 Serious adverse events requiring hospitalization.
Figures and Tables -
Analysis 1.11

Comparison 1 RV1 versus placebo, Outcome 11 Serious adverse events requiring hospitalization.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 12 Rotavirus diarrhoea: of any severity (up to 2 months follow‐up).
Figures and Tables -
Analysis 1.12

Comparison 1 RV1 versus placebo, Outcome 12 Rotavirus diarrhoea: of any severity (up to 2 months follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 13 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.13

Comparison 1 RV1 versus placebo, Outcome 13 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 14 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.14

Comparison 1 RV1 versus placebo, Outcome 14 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 15 All‐cause diarrhoea: all cases (up to 2 months follow‐up).
Figures and Tables -
Analysis 1.15

Comparison 1 RV1 versus placebo, Outcome 15 All‐cause diarrhoea: all cases (up to 2 months follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 16 All‐cause diarrhoea: all cases (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.16

Comparison 1 RV1 versus placebo, Outcome 16 All‐cause diarrhoea: all cases (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 17 All‐cause diarrhoea: all cases (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.17

Comparison 1 RV1 versus placebo, Outcome 17 All‐cause diarrhoea: all cases (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 18 All‐cause diarrhoea: all episodes (up to 1 year follow‐up).
Figures and Tables -
Analysis 1.18

Comparison 1 RV1 versus placebo, Outcome 18 All‐cause diarrhoea: all episodes (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 19 All‐cause diarrhoea: all episodes (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.19

Comparison 1 RV1 versus placebo, Outcome 19 All‐cause diarrhoea: all episodes (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 20 All‐cause hospitalizations (up to 2 years follow‐up).
Figures and Tables -
Analysis 1.20

Comparison 1 RV1 versus placebo, Outcome 20 All‐cause hospitalizations (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 21 Rotavirus diarrhoea: requiring hospitalization.
Figures and Tables -
Analysis 1.21

Comparison 1 RV1 versus placebo, Outcome 21 Rotavirus diarrhoea: requiring hospitalization.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 22 Rotavirus diarrhoea: requiring medical attention.
Figures and Tables -
Analysis 1.22

Comparison 1 RV1 versus placebo, Outcome 22 Rotavirus diarrhoea: requiring medical attention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 23 All‐cause diarrhoea: cases requiring hospitalization.
Figures and Tables -
Analysis 1.23

Comparison 1 RV1 versus placebo, Outcome 23 All‐cause diarrhoea: cases requiring hospitalization.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 24 All‐cause diarrhoea: episodes requiring hospitalization.
Figures and Tables -
Analysis 1.24

Comparison 1 RV1 versus placebo, Outcome 24 All‐cause diarrhoea: episodes requiring hospitalization.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 25 Reactogenicity: fever.
Figures and Tables -
Analysis 1.25

Comparison 1 RV1 versus placebo, Outcome 25 Reactogenicity: fever.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 26 Reactogenicity: diarrhoea.
Figures and Tables -
Analysis 1.26

Comparison 1 RV1 versus placebo, Outcome 26 Reactogenicity: diarrhoea.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 27 Reactogenicity: vomiting.
Figures and Tables -
Analysis 1.27

Comparison 1 RV1 versus placebo, Outcome 27 Reactogenicity: vomiting.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 28 Adverse events requiring discontinuation (end of follow‐up).
Figures and Tables -
Analysis 1.28

Comparison 1 RV1 versus placebo, Outcome 28 Adverse events requiring discontinuation (end of follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 29 Immunogenicity: rotavirus vaccine shedding (end of follow‐up).
Figures and Tables -
Analysis 1.29

Comparison 1 RV1 versus placebo, Outcome 29 Immunogenicity: rotavirus vaccine shedding (end of follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 30 Immunogenicity: seroconversion.
Figures and Tables -
Analysis 1.30

Comparison 1 RV1 versus placebo, Outcome 30 Immunogenicity: seroconversion.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 31 Dropouts before the end of the trial.
Figures and Tables -
Analysis 1.31

Comparison 1 RV1 versus placebo, Outcome 31 Dropouts before the end of the trial.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 32 Subgroup analysis: rotavirus diarrhoea of any severity (by G type).
Figures and Tables -
Analysis 1.32

Comparison 1 RV1 versus placebo, Outcome 32 Subgroup analysis: rotavirus diarrhoea of any severity (by G type).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 33 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type).
Figures and Tables -
Analysis 1.33

Comparison 1 RV1 versus placebo, Outcome 33 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type).

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 34 Subgroup analysis: rotavirus diarrhoea in malnourished children.
Figures and Tables -
Analysis 1.34

Comparison 1 RV1 versus placebo, Outcome 34 Subgroup analysis: rotavirus diarrhoea in malnourished children.

Navigate to figure in ReviewOpen in new tab

Comparison 1 RV1 versus placebo, Outcome 35 Subgroup analysis: rotavirus diarrhoea in HIV‐infected children.
Figures and Tables -
Analysis 1.35

Comparison 1 RV1 versus placebo, Outcome 35 Subgroup analysis: rotavirus diarrhoea in HIV‐infected children.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).
Figures and Tables -
Analysis 2.1

Comparison 2 RV5 versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).
Figures and Tables -
Analysis 2.2

Comparison 2 RV5 versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).
Figures and Tables -
Analysis 2.3

Comparison 2 RV5 versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up).
Figures and Tables -
Analysis 2.4

Comparison 2 RV5 versus placebo, Outcome 4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 5 All‐cause death.
Figures and Tables -
Analysis 2.5

Comparison 2 RV5 versus placebo, Outcome 5 All‐cause death.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 6 All serious adverse events.
Figures and Tables -
Analysis 2.6

Comparison 2 RV5 versus placebo, Outcome 6 All serious adverse events.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 7 Serious adverse events: intussusception.
Figures and Tables -
Analysis 2.7

Comparison 2 RV5 versus placebo, Outcome 7 Serious adverse events: intussusception.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 8 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).
Figures and Tables -
Analysis 2.8

Comparison 2 RV5 versus placebo, Outcome 8 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 9 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).
Figures and Tables -
Analysis 2.9

Comparison 2 RV5 versus placebo, Outcome 9 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 10 All‐cause diarrhoea: of any severity (up to 1 year follow‐up).
Figures and Tables -
Analysis 2.10

Comparison 2 RV5 versus placebo, Outcome 10 All‐cause diarrhoea: of any severity (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 11 All‐cause diarrhoea: of any severity (up to 2 years follow‐up).
Figures and Tables -
Analysis 2.11

Comparison 2 RV5 versus placebo, Outcome 11 All‐cause diarrhoea: of any severity (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 12 All‐cause hospitalizations (up to 2 years follow‐up).
Figures and Tables -
Analysis 2.12

Comparison 2 RV5 versus placebo, Outcome 12 All‐cause hospitalizations (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 13 Rotavirus diarrhoea: requiring hospitalization.
Figures and Tables -
Analysis 2.13

Comparison 2 RV5 versus placebo, Outcome 13 Rotavirus diarrhoea: requiring hospitalization.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 14 Rotavirus diarrhoea: requiring medical attention.
Figures and Tables -
Analysis 2.14

Comparison 2 RV5 versus placebo, Outcome 14 Rotavirus diarrhoea: requiring medical attention.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 15 Reactogenicity: fever.
Figures and Tables -
Analysis 2.15

Comparison 2 RV5 versus placebo, Outcome 15 Reactogenicity: fever.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 16 Reactogenicity: diarrhoea.
Figures and Tables -
Analysis 2.16

Comparison 2 RV5 versus placebo, Outcome 16 Reactogenicity: diarrhoea.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 17 Reactogenicity: vomiting.
Figures and Tables -
Analysis 2.17

Comparison 2 RV5 versus placebo, Outcome 17 Reactogenicity: vomiting.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 18 Adverse events requiring discontinuation (end of follow‐up).
Figures and Tables -
Analysis 2.18

Comparison 2 RV5 versus placebo, Outcome 18 Adverse events requiring discontinuation (end of follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 19 Immunogenicity: rotavirus vaccine shedding (after dose 3).
Figures and Tables -
Analysis 2.19

Comparison 2 RV5 versus placebo, Outcome 19 Immunogenicity: rotavirus vaccine shedding (after dose 3).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 20 Immunogenicity: seroconversion.
Figures and Tables -
Analysis 2.20

Comparison 2 RV5 versus placebo, Outcome 20 Immunogenicity: seroconversion.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 21 Dropouts before the end of the trial.
Figures and Tables -
Analysis 2.21

Comparison 2 RV5 versus placebo, Outcome 21 Dropouts before the end of the trial.

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 22 Subgroup analysis: rotavirus diarrhoea of any severity (by G type).
Figures and Tables -
Analysis 2.22

Comparison 2 RV5 versus placebo, Outcome 22 Subgroup analysis: rotavirus diarrhoea of any severity (by G type).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 23 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type).
Figures and Tables -
Analysis 2.23

Comparison 2 RV5 versus placebo, Outcome 23 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type).

Navigate to figure in ReviewOpen in new tab

Comparison 2 RV5 versus placebo, Outcome 24 Subgroup analysis: HIV‐infected children.
Figures and Tables -
Analysis 2.24

Comparison 2 RV5 versus placebo, Outcome 24 Subgroup analysis: HIV‐infected children.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).
Figures and Tables -
Analysis 3.1

Comparison 3 Rotavac versus placebo, Outcome 1 Rotavirus diarrhoea: severe (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).
Figures and Tables -
Analysis 3.2

Comparison 3 Rotavac versus placebo, Outcome 2 Rotavirus diarrhoea: severe (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).
Figures and Tables -
Analysis 3.3

Comparison 3 Rotavac versus placebo, Outcome 3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 4 All‐cause death.
Figures and Tables -
Analysis 3.4

Comparison 3 Rotavac versus placebo, Outcome 4 All‐cause death.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 5 All serious adverse events.
Figures and Tables -
Analysis 3.5

Comparison 3 Rotavac versus placebo, Outcome 5 All serious adverse events.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 6 Serious adverse events: intussusception.
Figures and Tables -
Analysis 3.6

Comparison 3 Rotavac versus placebo, Outcome 6 Serious adverse events: intussusception.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 7 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).
Figures and Tables -
Analysis 3.7

Comparison 3 Rotavac versus placebo, Outcome 7 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 8 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).
Figures and Tables -
Analysis 3.8

Comparison 3 Rotavac versus placebo, Outcome 8 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 9 Rotavirus diarrhoea: requiring medical attention.
Figures and Tables -
Analysis 3.9

Comparison 3 Rotavac versus placebo, Outcome 9 Rotavirus diarrhoea: requiring medical attention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 10 Reactogenicity: fever.
Figures and Tables -
Analysis 3.10

Comparison 3 Rotavac versus placebo, Outcome 10 Reactogenicity: fever.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 11 Reactogenicity: diarrhoea.
Figures and Tables -
Analysis 3.11

Comparison 3 Rotavac versus placebo, Outcome 11 Reactogenicity: diarrhoea.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 12 Reactogenicity: vomiting.
Figures and Tables -
Analysis 3.12

Comparison 3 Rotavac versus placebo, Outcome 12 Reactogenicity: vomiting.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 13 Immunogenicity: rotavirus vaccine shedding (end of follow‐up).
Figures and Tables -
Analysis 3.13

Comparison 3 Rotavac versus placebo, Outcome 13 Immunogenicity: rotavirus vaccine shedding (end of follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 14 Immunogenicity: seroconversion.
Figures and Tables -
Analysis 3.14

Comparison 3 Rotavac versus placebo, Outcome 14 Immunogenicity: seroconversion.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 15 Dropouts before the end of the trial.
Figures and Tables -
Analysis 3.15

Comparison 3 Rotavac versus placebo, Outcome 15 Dropouts before the end of the trial.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 16 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 1 year follow‐up).
Figures and Tables -
Analysis 3.16

Comparison 3 Rotavac versus placebo, Outcome 16 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 1 year follow‐up).

Navigate to figure in ReviewOpen in new tab

Comparison 3 Rotavac versus placebo, Outcome 17 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 2 years follow‐up).
Figures and Tables -
Analysis 3.17

Comparison 3 Rotavac versus placebo, Outcome 17 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 2 years follow‐up).

Navigate to figure in ReviewOpen in new tab
Summary of findings for the main comparison. RV1 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries

Patient or population: children
Setting: low‐mortality countries (WHO strata A and B)
Intervention: RV1
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV1

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

13 per 1000

2 per 1000
(1 to 3)

RR 0.16
(0.09 to 0.26)

43,779
(7 studies)

⊕⊕⊕⊕
higha

RV1 reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

One study (RV1 Vesikari 2007a‐EU) reported higher efficacy compared to the pooled data. When we excluded this study from the analysis, there was no heterogeneity observed in the pooled data

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

24 per 1000

4 per 1000
(3 to 5)

RR 0.18
(0.14 to 0.23)

36,002
(9 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

41 per 1000

24 per 1000
(19 to 30)

RR 0.59
(0.47 to 0.74)

28,051
(3 studies)

⊕⊕⊕⊝
moderateb

due to reporting bias

RV1 probably reduces severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

Severe episodes of all‐cause diarrhoea
Follow‐up: up to 2 years

39 per 1000

24 per 1000
(22 to 28)

Rate Ratio 0.63
(0.56 to 0.71)

39,091
(2 studies)

⊕⊕⊕⊝
moderatec

due to reporting bias

RV1 probably reduces severe all‐cause diarrhoea compared to placebo at up to two years follow‐up.

Three additional studies reported on cases of children with severe all‐cause diarrhoea (RR 0.60, 95% CI 0.36 to 1.02; 9417 participants); these data could not be pooled with the studies reporting on number of episodes

All‐cause death
Follow‐up: 2 months to 2 years

1 per 1000

2 per 1000
(1 to 2)

RR 1.22
(0.87 to 1.71)

97,597
(22 studies)

⊕⊕⊝⊝
lowd

due to imprecision

RV1 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

45 per 1000

40 per 1000
(37 to 42)

RR 0.88
(0.83 to 0.93)

96,233
(24 studies)

⊕⊕⊕⊕
high

RV1 slightly reduces serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

1 per 1000

1 per 1000
(0 to 1)

RR 0.69
(0.45 to 1.04)

96,513
(17 studies)

⊕⊕⊝⊝
lowe

due to imprecision

RV1 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aWe observed heterogeneity (I2 statistic = 61%) in the pooled data, but given the strength of the evidence, and that estimates were all in the same direction, we did not downgrade the outcome.
bDowngraded by one for risk of selective reporting bias. Only three of the seven studies reporting on severe rotavirus diarrhoea provided data for this outcome.
cDowngraded by one for risk of selective reporting bias. Only five of the nine studies reporting on severe rotavirus diarrhoea provided data for this outcome.
dDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
eDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Figures and Tables -
Summary of findings for the main comparison. RV1 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries
Navigate to table in Review
Summary of findings 2. RV1 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children
Settings: high‐mortality countries (WHO strata D and E)
Intervention: RV1
Comparison: placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo or no intervention

RV1

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

60 per 1000

22 per 1000
(14 to 36)

RR 0.37
(0.23 to 0.60)

6114
(3 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo or no intervention at up to one year follow‐up.

We did not downgrade for inconsistency as the heterogeneity observed in the pooled data (I2 statistic = 57%) was due to within‐study heterogeneity (RV1 Madhi 2010‐AF results split by country)

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

43 per 1000

28 per 1000
(22 to 35)

RR 0.65
(0.51 to 0.83)

13,768**
(2 studies)

⊕⊕⊕⊕
high

RV1 reduces severe rotavirus diarrhoea compared to placebo or no intervention at up to two years follow‐up.

Sensitivity analysis excluding the cluster‐RCT (RV1 Zaman 2017‐BGD) that contributed data to this outcome showed no significant change in effect estimate or 95% CI (RR 0.58, 95% CI 0.42 to 0.79, n = 2764, 1 RCT)

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

176 per 1000

129 per 1000
(99 to 167)

RR 0.73
(0.56 to 0.95)

5639
(2 studies)

⊕⊕⊕⊕
high

RV1 reduces severe all‐cause diarrhoea compared to placebo or no intervention at up to one year follow‐up.

We did not downgrade for inconsistency as the heterogeneity observed in the pooled data (I2 statistic = 75%) was due to within‐study heterogeneity (RV1 Madhi 2010‐AF results split by country)

Severe cases of all‐cause diarrhoea
Follow‐up: up to 2 years

233 per 1000

191 per 1000
(166 to 222)

RR 0.82
(0.71 to 0.95)

2764
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

RV1 probably slightly reduces severe all‐cause diarrhoea compared to placebo or no intervention at up to two years follow‐up.

All‐cause death
Follow‐up: 2 months to 2 years

24 per 1000

21 per 1000
(16 to 30)

RR 0.88
(0.64 to 1.22)

8181
(8 studies)

⊕⊕⊝⊝
lowb

due to imprecision

RV1 may make little or no difference to all‐cause death compared to placebo or no intervention.

All serious adverse events
Follow‐up: 2 months to 2 years

95 per 1000

84 per 1000
(72 to 99)

RR 0.89
(0.76 to 1.04)

7481
(7 studies)

⊕⊕⊕⊕
high

RV1 makes little or no difference to serious adverse events compared to placebo or no intervention.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

0 per 100,000

0 per 100,000
(0 to 0)

RR 1.49
(0.06 to 36.63)

17,492**
(4 studies)

⊕⊕⊝⊝
lowc

due to imprecision

RV1 may make little or no difference to intussusception compared to placebo or no intervention.

Sensitivity analysis excluding the cluster‐RCT (RV1 Zaman 2017‐BGD) that contributed data to this outcome showed no change in effect estimate or 95% CI

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**Number of participants in this table shows the true number of participants for this outcome; the number of events and the number of participants in the analysis has been adjusted for the included cluster trial RV1 Zaman 2017‐BGD using a design effect of 2.53.
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for indirectness. Trials were conducted in Malawi and South Africa, so generalization to any high‐mortality country is difficult.
bDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
cDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Figures and Tables -
Summary of findings 2. RV1 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries
Navigate to table in Review
Summary of findings 3. RV5 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries

Patient or population: children
Settings: low‐mortality countries (WHO strata A and B)
Intervention: RV5
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV5

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

17 per 1000

1 per 1000
(1 to 5)

RR 0.08
(0.03 to 0.22)

4132
(5 studies)

⊕⊕⊕⊝
moderatea

due to imprecision

RV5 probably reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

25 per 1000

4 per 1000
(2 to 10)

RR 0.18
(0.08 to 0.39)

7318
(4 studies)

⊕⊕⊕⊝
moderateb

due to inconsistency

RV5 probably reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe all‐cause diarrhoea
Follow‐up: up to 1 year

‐

‐

‐

‐

‐

We found no studies that reported on this outcome in this setting

Severe all‐cause diarrhoea
Follow‐up: up to 2 years

‐

‐

‐

‐

‐

We found no studies that reported on this outcome in this setting

All‐cause death
Follow‐up: 2 months to 2 years

1 per 1000

1 per 1000
(0 to 1)

RR 1.13
(0.65 to 1.96)

77,642
(9 studies)

⊕⊕⊝⊝
lowc

due to imprecision

RV5 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

27 per 1000

25 per 1000
(23 to 28)

RR 0.93
(0.86 to 1.02)

75,672
(8 studies)

⊕⊕⊕⊕
high

RV5 makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

1 per 1000

0 per 1000
(0 to 1)

RR 0.77
(0.41 to 1.45)

78,907
(12 studies)

⊕⊕⊝⊝
lowd

due to imprecision

RV5 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for imprecision. The total number of events was very low.
bDowngraded by one for inconsistency. We found substantial heterogeneity (I2 statistic = 44%). Consistency was restored when removing the one study carried out only in a very low‐mortality (stratum A) country, with results then showing a slightly smaller effect (RR 0.22, 95% CI 0.13 to 0.36, 6291 participants, 3 studies).
cDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Figures and Tables -
Summary of findings 3. RV5 compared to placebo for preventing rotavirus diarrhoea in low‐mortality countries
Navigate to table in Review
Summary of findings 4. RV5 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children
Settings: high‐mortality countries (WHO strata D and E)
Intervention: RV5
Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

Number of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

RV5

Severe cases of rotavirus diarrhoea
Follow‐up: up to 1 year

30 per 1000

13 per 1000
(9 to 19)

RR 0.43
(0.29 to 0.62)

5916
(2 studies)

⊕⊕⊕⊕
high

RV5 reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea
Follow‐up: up to 2 years

63 per 1000

37 per 1000
(27 to 51)

RR 0.59
(0.43 to 0.82)

5885
(2 studies)

⊕⊕⊕⊕
high

RV5 reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 1 year

77 per 1000

62 per 1000
(45 to 85)

RR 0.8
(0.58 to 1.11)

4085
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

RV5 probably makes little or no difference to severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of all‐cause diarrhoea
Follow‐up: up to 2 years

130 per 1000

110 per 1000
(97 to 127)

RR 0.85
(0.75 to 0.98)

5977
(2 studies)

⊕⊕⊕⊕
high

RV5 slightly reduces severe all‐cause diarrhoea compared to placebo at up to two years follow‐up.

All‐cause death
Follow‐up: 2 months to 2 years

26 per 1000

23 per 1000
(17 to 32)

RR 0.92
(0.68 to 1.24)

6806
(3 studies)

⊕⊕⊝⊝
lowb

due to imprecision

RV5 may make little or no difference to all‐cause death compared to placebo.

All serious adverse events
Follow‐up: 2 months to 2 years

21 per 1000

19 per 1000
(14 to 27)

RR 0.92
(0.66 to 1.28)

6830
(4 studies)

⊕⊕⊕⊝
moderatec

due to imprecision

RV5 probably makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception
Follow‐up: 2 months to 2 years

See comment

See comment

Not estimable

6588
(2 studies)

⊕⊕⊝⊝
lowd

due to imprecision

No events were reported. RV5 may make little or no difference to intussusception compared to placebo.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; RR: risk ratio

GRADE Working Group grades of evidence
High‐certainty: further research is very unlikely to change our confidence in the estimate of effect.
Moderate‐certainty: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low‐certainty: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low‐certainty: we are very uncertain about the estimate.

aDowngraded by one for indirectness. Single trial conducted in three African countries (Mali, Ghana, and Kenya), so generalization to any high‐mortality country is difficult.
bDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
cDowngraded by one for imprecision. The 95% CI includes both no effect and appreciable harm.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), so these trials were not powered to detect an association between RV1 and intussusception.

Figures and Tables -
Summary of findings 4. RV5 compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries
Navigate to table in Review
Summary of findings 5. Rotavac compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries

Patient or population: children

Settings: one high‐mortality country (India) (WHO stratum D)

Intervention: Rotavac

Comparison: placebo

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of Participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Placebo

Rotavac

Severe cases of rotavirus diarrhoea

follow‐up: up to 1 year

31 per 1000

13 per 1000
(9 to 19)

RR 0.43
(0.30 to 0.60)

6799
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably reduces severe rotavirus diarrhoea compared to placebo at up to one year follow‐up.

Severe cases of rotavirus diarrhoea follow‐up: up to 2 years

47 per 1000

21 per 1000
(16 to 28)

RR 0.46
(0.35 to 0.60)

6541
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably reduces severe rotavirus diarrhoea compared to placebo at up to two years follow‐up.

Severe cases of all‐cause diarrhoea

follow‐up: up to 2 years

93 per 1000

78 per 1000
(66 to 91)

RR 0.84
(0.71 to 0.98)

6799
(1 study)

⊕⊕⊕⊝
moderatea

due to indirectness

Rotavac probably slightly reduces severe all‐cause diarrhoea compared to placebo at up to one year follow‐up.

All‐cause death

follow‐up: up to 2 years

7 per 1000

6 per 1000
(4 to 11)

RR 0.92
(0.52 to 1.62)

8155
(2 studies)

⊕⊝⊝⊝
very lowb,c

due to indirectness and imprecision

We are uncertain whether Rotavac reduced all‐cause death as the certainty of the evidence is very low.

All serious adverse events

follow‐up: up to 2 years

204 per 1000

189 per 1000
(173 to 208)

RR 0.93
(0.85 to 1.02)

8210
(3 studies)

⊕⊕⊕⊝
moderateb

due to indirectness

Rotavac probably makes little or no difference to serious adverse events compared to placebo.

Serious adverse events: intussusception

follow‐up: up to 2 years

1 per 1000

1 per 1000
(0 to 5)

RR 1.33
(0.35 to 5.02)

8582
(4 studies)

⊕⊝⊝⊝
very lowb,d

due to indirectness and imprecision

No events were reported in three of the four studies. We are uncertain whether Rotavac has an effect on intussusception as the certainty of the evidence is very low.

*The basis for the assumed risk is the control group risk across studies included in the meta‐analysis. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk Ratio

GRADE Working Group grades of evidence
High certainty: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate certainty: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low certainty: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low certainty: We are very uncertain about the estimate.

aDowngraded by one for indirectness. Single trial conducted in India, so generalization to any high‐mortality country is difficult.
bDowngraded by one for indirectness. All trials were conducted in India, so generalization to any high‐mortality country is difficult.
cDowngraded by two for imprecision. These trials were not powered to detect an effect on mortality.
dDowngraded by two for imprecision. There was a 1:10,000 to 1:32,000 increased risk of intussusception with a previous rotavirus vaccine (Bines 2005), therefore, these trials were not powered to detect an association between Rotavac and intussusception.

Figures and Tables -
Summary of findings 5. Rotavac compared to placebo for preventing rotavirus diarrhoea in high‐mortality countries
Navigate to table in Review
Comparison 1. RV1 versus placebo

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Rotavirus diarrhoea: severe (up to 1 year follow‐up) Show forest plot

11

49893

Risk Ratio (M‐H, Random, 95% CI)

0.22 [0.14, 0.34]

1.1 Low‐mortality countries (WHO strata A & B)

7

43779

Risk Ratio (M‐H, Random, 95% CI)

0.16 [0.09, 0.26]

1.2 High‐mortality countries (WHO strata D & E)

4

6114

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.23, 0.60]

2 Rotavirus diarrhoea: severe (up to 2 years follow‐up) Show forest plot

12

Risk Ratio (Fixed, 95% CI)

0.34 [0.29, 0.41]

2.1 Low‐mortality countries (WHO strata A & B)

9

Risk Ratio (Fixed, 95% CI)

0.18 [0.14, 0.23]

2.2 High‐mortality countries (WHO strata D & E)

3

Risk Ratio (Fixed, 95% CI)

0.65 [0.51, 0.83]

3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up) Show forest plot

6

33690

Risk Ratio (M‐H, Random, 95% CI)

0.66 [0.54, 0.80]

3.1 Low‐mortality countries (WHO strata A & B)

3

28051

Risk Ratio (M‐H, Random, 95% CI)

0.59 [0.47, 0.74]

3.2 High‐mortality countries (WHO strata D & E)

3

5639

Risk Ratio (M‐H, Random, 95% CI)

0.73 [0.56, 0.95]

4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up) Show forest plot

5

12181

Risk Ratio (M‐H, Random, 95% CI)

0.70 [0.54, 0.92]

4.1 Low‐mortality countries (WHO strata A & B)

3

9417

Risk Ratio (M‐H, Random, 95% CI)

0.60 [0.36, 1.02]

4.2 High‐mortality countries (WHO strata D & E)

2

2764

Risk Ratio (M‐H, Random, 95% CI)

0.83 [0.72, 0.96]

5 All‐cause diarrhoea: severe episodes (up to 1 year follow‐up) Show forest plot

1

Rate Ratio (Fixed, 95% CI)

Totals not selected

5.1 Low‐mortality countries (WHO strata A & B)

1

Rate Ratio (Fixed, 95% CI)

0.0 [0.0, 0.0]

6 All‐cause diarrhoea: severe episodes (up to 2 years follow‐up) Show forest plot

2

Rate Ratio (Fixed, 95% CI)

Subtotals only

6.1 Low‐mortality countries (WHO strata A & B)

2

Rate Ratio (Fixed, 95% CI)

0.63 [0.56, 0.71]

7 All‐cause death Show forest plot

30

105778

Risk Ratio (M‐H, Fixed, 95% CI)

1.03 [0.82, 1.30]

7.1 Low‐mortality countries (WHO strata A & B)

22

97597

Risk Ratio (M‐H, Fixed, 95% CI)

1.22 [0.87, 1.71]

7.2 High‐mortality countries (WHO strata D & E)

8

8181

Risk Ratio (M‐H, Fixed, 95% CI)

0.88 [0.64, 1.22]

8 All serious adverse events Show forest plot

31

103714

Risk Ratio (M‐H, Fixed, 95% CI)

0.88 [0.83, 0.93]

8.1 Low‐mortality countries (WHO strata A & B)

24

96233

Risk Ratio (M‐H, Fixed, 95% CI)

0.88 [0.83, 0.93]

8.2 High‐mortality countries (WHO strata D & E)

7

7481

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.76, 1.04]

9 Serious adverse events: intussusception Show forest plot

21

Risk Ratio (Fixed, 95% CI)

0.70 [0.46, 1.05]

9.1 Low‐mortality countries (WHO strata A & B)

17

Risk Ratio (Fixed, 95% CI)

0.69 [0.45, 1.04]

9.2 High‐mortality countries (WHO stratum E)

4

Risk Ratio (Fixed, 95% CI)

1.49 [0.06, 36.63]

10 Serious adverse events: Kawasaki disease Show forest plot

3

13117

Risk Ratio (M‐H, Fixed, 95% CI)

1.79 [0.30, 10.61]

11 Serious adverse events requiring hospitalization Show forest plot

2

63675

Risk Ratio (M‐H, Fixed, 95% CI)

0.88 [0.81, 0.96]

12 Rotavirus diarrhoea: of any severity (up to 2 months follow‐up) Show forest plot

12

4294

Risk Ratio (M‐H, Fixed, 95% CI)

1.17 [0.69, 2.00]

12.1 Low‐mortality countries (WHO strata A & B)

9

3537

Risk Ratio (M‐H, Fixed, 95% CI)

1.28 [0.66, 2.50]

12.2 High‐mortality countries (WHO strata D & E)

3

757

Risk Ratio (M‐H, Fixed, 95% CI)

1.0 [0.41, 2.41]

13 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up) Show forest plot

8

15197

Risk Ratio (M‐H, Random, 95% CI)

0.34 [0.23, 0.50]

13.1 Low‐mortality countries (WHO strata A & B)

4

9083

Risk Ratio (M‐H, Random, 95% CI)

0.22 [0.13, 0.40]

13.2 High‐mortality countries (WHO stratum E)

4

6114

Risk Ratio (M‐H, Random, 95% CI)

0.49 [0.35, 0.68]

14 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up) Show forest plot

7

11692

Risk Ratio (M‐H, Random, 95% CI)

0.36 [0.28, 0.47]

14.1 Low‐mortality countries (WHO strata A & B)

6

10441

Risk Ratio (M‐H, Random, 95% CI)

0.35 [0.25, 0.48]

14.2 High‐mortality countries (WHO stratum E)

1

1251

Risk Ratio (M‐H, Random, 95% CI)

0.41 [0.28, 0.62]

15 All‐cause diarrhoea: all cases (up to 2 months follow‐up) Show forest plot

7

3132

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.72, 1.10]

15.1 Low‐mortality countries (WHO strata A & B)

6

3032

Risk Ratio (M‐H, Fixed, 95% CI)

0.86 [0.67, 1.09]

15.2 High‐mortality countries (WHO stratum E)

1

100

Risk Ratio (M‐H, Fixed, 95% CI)

1.04 [0.69, 1.58]

16 All‐cause diarrhoea: all cases (up to 1 year follow‐up) Show forest plot

3

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

16.1 Low‐mortality countries (WHO strata A & B)

2

2204

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.82, 1.03]

16.2 High‐mortality countries (WHO strata D & E)

1

700

Risk Ratio (M‐H, Fixed, 95% CI)

0.99 [0.93, 1.05]

17 All‐cause diarrhoea: all cases (up to 2 years follow‐up) Show forest plot

3

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

17.1 Low‐mortality countries (WHO strata A & B)

3

5937

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.87, 1.00]

18 All‐cause diarrhoea: all episodes (up to 1 year follow‐up) Show forest plot

2

Rate Ratio (Fixed, 95% CI)

Subtotals only

18.1 Low‐mortality countries (WHO strata A & B)

2

Rate Ratio (Fixed, 95% CI)

0.98 [0.88, 1.10]

19 All‐cause diarrhoea: all episodes (up to 2 years follow‐up) Show forest plot

1

Rate Ratio (Fixed, 95% CI)

Totals not selected

19.1 Low‐mortality countries (WHO strata A & B)

1

Rate Ratio (Fixed, 95% CI)

0.0 [0.0, 0.0]

20 All‐cause hospitalizations (up to 2 years follow‐up) Show forest plot

2

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

20.1 Low‐mortality countries (WHO strata A & B)

2

65646

Risk Ratio (M‐H, Random, 95% CI)

0.63 [0.27, 1.47]

21 Rotavirus diarrhoea: requiring hospitalization Show forest plot

11

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

21.1 Up to 1 year follow‐up (at least 1 rotavirus season)

8

48718

Risk Ratio (M‐H, Random, 95% CI)

0.18 [0.09, 0.33]

21.2 Second year follow‐up (at least 2 rotavirus seasons)

7

35331

Risk Ratio (M‐H, Random, 95% CI)

0.15 [0.11, 0.22]

22 Rotavirus diarrhoea: requiring medical attention Show forest plot

3

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

22.1 Up to 1 year follow‐up (at least 1 rotavirus season)

1

3874

Risk Ratio (M‐H, Fixed, 95% CI)

0.08 [0.04, 0.16]

22.2 Second year follow‐up (at least 2 rotavirus seasons)

3

7017

Risk Ratio (M‐H, Fixed, 95% CI)

0.22 [0.16, 0.31]

23 All‐cause diarrhoea: cases requiring hospitalization Show forest plot

2

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

23.1 Up to one year of follow‐up (at least 1 rotavirus season)

2

14393

Risk Ratio (M‐H, Random, 95% CI)

0.43 [0.17, 1.11]

23.2 Second year of follow‐up (at least 2 rotavirus seasons)

2

14367

Risk Ratio (M‐H, Random, 95% CI)

0.52 [0.27, 0.99]

24 All‐cause diarrhoea: episodes requiring hospitalization Show forest plot

1

Rate Ratio (Fixed, 95% CI)

Subtotals only

24.1 Up to 1 year of follow‐up (at least 1 rotavirus season)

1

Rate Ratio (Fixed, 95% CI)

0.58 [0.47, 0.71]

24.2 Second year of follow‐up (at least 2 rotavirus seasons)

1

Rate Ratio (Fixed, 95% CI)

0.53 [0.46, 0.61]

25 Reactogenicity: fever Show forest plot

28

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

25.1 After dose 1

25

16192

Risk Ratio (M‐H, Random, 95% CI)

1.06 [0.97, 1.17]

25.2 After dose 2

24

15630

Risk Ratio (M‐H, Random, 95% CI)

0.99 [0.92, 1.06]

25.3 After dose 3

4

1390

Risk Ratio (M‐H, Random, 95% CI)

0.98 [0.86, 1.13]

25.4 End of follow‐up

18

11926

Risk Ratio (M‐H, Random, 95% CI)

0.97 [0.93, 1.01]

26 Reactogenicity: diarrhoea Show forest plot

27

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

26.1 After dose 1

25

18732

Risk Ratio (M‐H, Random, 95% CI)

1.01 [0.88, 1.17]

26.2 After dose 2

24

15630

Risk Ratio (M‐H, Random, 95% CI)

1.02 [0.86, 1.21]

26.3 After dose 3

4

1390

Risk Ratio (M‐H, Random, 95% CI)

0.69 [0.35, 1.36]

26.4 End of follow‐up

17

14305

Risk Ratio (M‐H, Random, 95% CI)

0.95 [0.84, 1.08]

27 Reactogenicity: vomiting Show forest plot

27

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

27.1 After dose 1

25

18732

Risk Ratio (M‐H, Random, 95% CI)

1.03 [0.94, 1.12]

27.2 After dose 2

24

15630

Risk Ratio (M‐H, Random, 95% CI)

0.92 [0.81, 1.05]

27.3 After dose 3

4

1390

Risk Ratio (M‐H, Random, 95% CI)

1.34 [0.71, 2.50]

27.4 End of follow‐up

17

14305

Risk Ratio (M‐H, Random, 95% CI)

0.93 [0.84, 1.04]

28 Adverse events requiring discontinuation (end of follow‐up) Show forest plot

26

94980

Risk Ratio (M‐H, Fixed, 95% CI)

1.03 [0.83, 1.26]

29 Immunogenicity: rotavirus vaccine shedding (end of follow‐up) Show forest plot

16

2638

Risk Ratio (M‐H, Random, 95% CI)

10.94 [4.90, 24.43]

30 Immunogenicity: seroconversion Show forest plot

31

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

30.1 After dose 1

9

2537

Risk Ratio (M‐H, Random, 95% CI)

20.39 [8.48, 49.01]

30.2 After dose 2

27

8742

Risk Ratio (M‐H, Random, 95% CI)

11.44 [8.01, 16.32]

30.3 After dose 3

5

1137

Risk Ratio (M‐H, Random, 95% CI)

6.89 [3.59, 13.24]

31 Dropouts before the end of the trial Show forest plot

28

93106

Risk Ratio (M‐H, Fixed, 95% CI)

0.95 [0.90, 1.00]

32 Subgroup analysis: rotavirus diarrhoea of any severity (by G type) Show forest plot

6

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

32.1 G1

6

27583

Risk Ratio (M‐H, Random, 95% CI)

0.21 [0.10, 0.44]

32.2 G2

5

26835

Risk Ratio (M‐H, Random, 95% CI)

0.41 [0.31, 0.56]

32.3 G3

4

8968

Risk Ratio (M‐H, Random, 95% CI)

0.14 [0.05, 0.39]

32.4 G4

2

5720

Risk Ratio (M‐H, Random, 95% CI)

0.20 [0.07, 0.59]

32.5 G9

3

8868

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.18, 0.75]

33 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type) Show forest plot

8

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

33.1 G1

7

39428

Risk Ratio (M‐H, Random, 95% CI)

0.24 [0.16, 0.38]

33.2 G2

7

44682

Risk Ratio (M‐H, Random, 95% CI)

0.30 [0.18, 0.50]

33.3 G3

5

20505

Risk Ratio (M‐H, Random, 95% CI)

0.17 [0.05, 0.56]

33.4 G4

1

2421

Risk Ratio (M‐H, Random, 95% CI)

0.12 [0.00, 2.95]

33.5 G8

2

4417

Risk Ratio (M‐H, Random, 95% CI)

0.22 [0.02, 2.37]

33.6 G9

6

26815

Risk Ratio (M‐H, Random, 95% CI)

0.23 [0.13, 0.40]

33.7 G12

2

4417

Risk Ratio (M‐H, Random, 95% CI)

0.47 [0.23, 0.97]

34 Subgroup analysis: rotavirus diarrhoea in malnourished children Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

34.1 Up to 1 year of follow‐up (at least 1 rotavirus season)

1

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

35 Subgroup analysis: rotavirus diarrhoea in HIV‐infected children Show forest plot

1

100

Risk Ratio (M‐H, Fixed, 95% CI)

1.0 [0.26, 3.78]
Figures and Tables -
Comparison 1. RV1 versus placebo
Navigate to table in Review
Comparison 2. RV5 versus placebo

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Rotavirus diarrhoea: severe (up to 1 year follow‐up) Show forest plot

9

10048

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.22, 0.44]

1.1 Low‐mortality countries (WHO strata A & B)

5

4132

Risk Ratio (M‐H, Fixed, 95% CI)

0.08 [0.03, 0.22]

1.2 High‐mortality countries (WHO strata D & E)

4

5916

Risk Ratio (M‐H, Fixed, 95% CI)

0.43 [0.29, 0.62]

2 Rotavirus diarrhoea: severe (up to 2 years follow‐up) Show forest plot

8

13203

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.23, 0.60]

2.1 Low‐mortality countries (WHO strata A & B)

4

7318

Risk Ratio (M‐H, Random, 95% CI)

0.18 [0.08, 0.39]

2.2 High‐mortality countries (WHO strata D & E)

4

5885

Risk Ratio (M‐H, Random, 95% CI)

0.59 [0.43, 0.82]

3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up) Show forest plot

3

4085

Risk Ratio (M‐H, Random, 95% CI)

0.80 [0.58, 1.11]

3.1 Low‐mortality countries (WHO stratum A)

0

0

Risk Ratio (M‐H, Random, 95% CI)

0.0 [0.0, 0.0]

3.2 High‐mortality countries (WHO strata D & E)

3

4085

Risk Ratio (M‐H, Random, 95% CI)

0.80 [0.58, 1.11]

4 All‐cause diarrhoea: severe cases (up to 2 years follow‐up) Show forest plot

4

5977

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.75, 0.98]

4.1 Low‐mortality countries (WHO strata A & B)

0

0

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

4.2 High‐mortality countries (WHO strata D & E)

4

5977

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.75, 0.98]

5 All‐cause death Show forest plot

14

84448

Risk Ratio (M‐H, Fixed, 95% CI)

0.96 [0.74, 1.25]

5.1 Low‐mortality countries (WHO strata A & B)

9

77642

Risk Ratio (M‐H, Fixed, 95% CI)

1.13 [0.65, 1.96]

5.2 High‐mortality countries (WHO strata D & E)

5

6806

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.68, 1.24]

6 All serious adverse events Show forest plot

14

82502

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.86, 1.01]

6.1 Low‐mortality countries (WHO strata A & B)

8

75672

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.86, 1.02]

6.2 High‐mortality countries (WHO strata D & E)

6

6830

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.66, 1.28]

7 Serious adverse events: intussusception Show forest plot

16

85495

Risk Ratio (M‐H, Fixed, 95% CI)

0.77 [0.41, 1.45]

7.1 Low‐mortality countries (WHO strata A & B)

12

78907

Risk Ratio (M‐H, Fixed, 95% CI)

0.77 [0.41, 1.45]

7.2 High‐mortality countries (WHO strata D & E)

4

6588

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

8 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up) Show forest plot

8

13450

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.28, 0.50]

8.1 Low‐mortality countries (WHO strata A & B)

5

8644

Risk Ratio (M‐H, Random, 95% CI)

0.30 [0.25, 0.37]

8.2 High‐mortality countries (WHO strata D & E)

3

4806

Risk Ratio (M‐H, Random, 95% CI)

0.52 [0.28, 0.94]

9 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up) Show forest plot

7

12888

Risk Ratio (M‐H, Random, 95% CI)

0.46 [0.33, 0.65]

9.1 Low‐mortality countries (WHO strata A & B)

3

6144

Risk Ratio (M‐H, Random, 95% CI)

0.34 [0.26, 0.43]

9.2 High‐mortality countries (WHO strata D & E)

4

6744

Risk Ratio (M‐H, Random, 95% CI)

0.61 [0.45, 0.83]

10 All‐cause diarrhoea: of any severity (up to 1 year follow‐up) Show forest plot

1

1059

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.61, 1.11]

10.1 Low‐mortality countries (WHO strata A & B)

0

0

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

10.2 High‐mortality countries (WHO stratum E)

1

1059

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.61, 1.11]

11 All‐cause diarrhoea: of any severity (up to 2 years follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

11.1 High‐mortality countries (WHO stratum E)

1

1059

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.68, 1.16]

12 All‐cause hospitalizations (up to 2 years follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

12.1 High‐mortality countries (WHO strata D & E)

1

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

13 Rotavirus diarrhoea: requiring hospitalization Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

13.1 Up to 1 year of follow‐up

1

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

14 Rotavirus diarrhoea: requiring medical attention Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

14.1 Up to 1 year of follow‐up

1

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

15 Reactogenicity: fever Show forest plot

12

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

15.1 After dose 1

4

7124

Risk Ratio (M‐H, Random, 95% CI)

1.15 [0.91, 1.45]

15.2 After dose 2

2

4322

Risk Ratio (M‐H, Random, 95% CI)

0.83 [0.69, 1.01]

15.3 After dose 3

2

4294

Risk Ratio (M‐H, Random, 95% CI)

1.07 [0.90, 1.27]

15.4 End of follow‐up

11

18391

Risk Ratio (M‐H, Random, 95% CI)

1.01 [0.94, 1.09]

16 Reactogenicity: diarrhoea Show forest plot

10

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

16.1 After dose 1

2

4745

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.95, 1.32]

16.2 After dose 2

1

3905

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.72, 1.10]

16.3 End of follow‐up

10

17087

Risk Ratio (M‐H, Fixed, 95% CI)

1.04 [0.98, 1.10]

17 Reactogenicity: vomiting Show forest plot

9

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

17.1 After dose 1

2

4745

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.63, 1.12]

17.2 After dose 2

1

3905

Risk Ratio (M‐H, Fixed, 95% CI)

0.69 [0.32, 1.49]

17.3 After dose 3

1

3878

Risk Ratio (M‐H, Fixed, 95% CI)

0.46 [0.16, 1.32]

17.4 End of follow‐up

9

16294

Risk Ratio (M‐H, Fixed, 95% CI)

0.98 [0.90, 1.06]

18 Adverse events requiring discontinuation (end of follow‐up) Show forest plot

10

15471

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.57, 1.39]

19 Immunogenicity: rotavirus vaccine shedding (after dose 3) Show forest plot

5

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected

20 Immunogenicity: seroconversion Show forest plot

10

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected

20.1 After dose 3

10

Risk Ratio (M‐H, Random, 95% CI)

0.0 [0.0, 0.0]

21 Dropouts before the end of the trial Show forest plot

13

85855

Risk Ratio (M‐H, Random, 95% CI)

0.98 [0.90, 1.08]

22 Subgroup analysis: rotavirus diarrhoea of any severity (by G type) Show forest plot

4

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

22.1 G1

4

11022

Risk Ratio (M‐H, Random, 95% CI)

0.26 [0.21, 0.32]

22.2 G2

3

9907

Risk Ratio (M‐H, Random, 95% CI)

0.35 [0.16, 0.78]

22.3 G3

4

11022

Risk Ratio (M‐H, Random, 95% CI)

0.40 [0.08, 2.02]

22.4 G4

3

9907

Risk Ratio (M‐H, Random, 95% CI)

0.41 [0.13, 1.33]

22.5 G9

2

9537

Risk Ratio (M‐H, Random, 95% CI)

0.33 [0.20, 0.54]

23 Subgroup analysis: severe cases of rotavirus diarrhoea (by G type) Show forest plot

3

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

23.1 G1

3

76606

Risk Ratio (M‐H, Random, 95% CI)

0.23 [0.03, 1.74]

23.2 G2

3

76606

Risk Ratio (M‐H, Random, 95% CI)

0.41 [0.13, 1.37]

23.3 G3

3

76606

Risk Ratio (M‐H, Random, 95% CI)

0.38 [0.05, 2.74]

23.4 G4

3

76606

Risk Ratio (M‐H, Random, 95% CI)

0.12 [0.03, 0.46]

23.5 G9

3

76606

Risk Ratio (M‐H, Random, 95% CI)

0.13 [0.05, 0.34]

24 Subgroup analysis: HIV‐infected children Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

24.1 Rotavirus diarrhoea: severe (up to two years follow‐up)

1

38

Risk Ratio (M‐H, Fixed, 95% CI)

2.45 [0.11, 56.68]

24.2 All‐cause diarrhoea: severe (up to two years follow‐up)

1

38

Risk Ratio (M‐H, Fixed, 95% CI)

4.05 [0.52, 31.43]

24.3 All‐cause death

2

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.29 [0.51, 3.21]

24.4 Serious adverse events (up to 24 weeks)

2

113

Risk Ratio (M‐H, Fixed, 95% CI)

1.53 [0.59, 3.97]
Figures and Tables -
Comparison 2. RV5 versus placebo
Navigate to table in Review
Comparison 3. Rotavac versus placebo

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Rotavirus diarrhoea: severe (up to 1 year follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2 Rotavirus diarrhoea: severe (up to 2 years follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3 All‐cause diarrhoea: severe cases (up to 1 year follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4 All‐cause death Show forest plot

2

8155

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.52, 1.62]

5 All serious adverse events Show forest plot

3

8210

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.85, 1.02]

6 Serious adverse events: intussusception Show forest plot

4

8582

Risk Ratio (M‐H, Fixed, 95% CI)

1.33 [0.35, 5.02]

7 Rotavirus diarrhoea: of any severity (up to 1 year follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

8 Rotavirus diarrhoea: of any severity (up to 2 years follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

9 Rotavirus diarrhoea: requiring medical attention Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

9.1 Up to 1 year follow‐up (at least 1 rotavirus season)

1

Risk Ratio (M‐H, Fixed, 95% CI)

0.0 [0.0, 0.0]

10 Reactogenicity: fever Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

10.1 After dose 1

2

427

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.35, 1.94]

10.2 After dose 2

1

356

Risk Ratio (M‐H, Fixed, 95% CI)

0.77 [0.33, 1.77]

10.3 After dose 3

1

358

Risk Ratio (M‐H, Fixed, 95% CI)

1.11 [0.52, 2.36]

11 Reactogenicity: diarrhoea Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

11.1 After dose 1

2

427

Risk Ratio (M‐H, Fixed, 95% CI)

0.90 [0.62, 1.30]

11.2 After dose 2

1

356

Risk Ratio (M‐H, Fixed, 95% CI)

1.55 [1.00, 2.41]

11.3 After dose 3

1

358

Risk Ratio (M‐H, Fixed, 95% CI)

4.09 [2.11, 7.92]

12 Reactogenicity: vomiting Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

12.1 After dose 1

2

427

Risk Ratio (M‐H, Fixed, 95% CI)

1.34 [0.71, 2.55]

12.2 After dose 2

1

356

Risk Ratio (M‐H, Fixed, 95% CI)

1.53 [0.64, 3.66]

12.3 After dose 3

1

358

Risk Ratio (M‐H, Fixed, 95% CI)

1.02 [0.39, 2.66]

13 Immunogenicity: rotavirus vaccine shedding (end of follow‐up) Show forest plot

2

427

Risk Ratio (M‐H, Random, 95% CI)

9.86 [2.58, 37.63]

14 Immunogenicity: seroconversion Show forest plot

3

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

14.1 After dose 1

1

121

Risk Ratio (M‐H, Fixed, 95% CI)

3.58 [2.03, 6.29]

14.2 After dose 2

1

117

Risk Ratio (M‐H, Fixed, 95% CI)

2.97 [1.78, 4.98]

14.3 After dose 3

3

1699

Risk Ratio (M‐H, Fixed, 95% CI)

2.82 [2.26, 3.51]

15 Dropouts before the end of the trial Show forest plot

3

8215

Risk Ratio (M‐H, Fixed, 95% CI)

0.81 [0.62, 1.06]

16 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 1 year follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

16.1 G1P[8]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.36, 1.20]

16.2 G2P[4]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.39 [0.22, 0.69]

16.3 G12P[6]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.13, 0.74]

16.4 G12P[8]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.30 [0.07, 1.26]

17 Subgroup analysis: severe cases of rotavirus diarrhoea by G and P types (up to 2 years follow‐up) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

17.1 G1P[8]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.59 [0.38, 0.93]

17.2 G2P[4]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.37 [0.23, 0.62]

17.3 G9P[4]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

4.52 [0.57, 35.66]

17.4 G12P[6]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.13, 0.74]

17.5 G12P[8]

1

6541

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.10, 0.96]
Figures and Tables -
Comparison 3. Rotavac versus placebo
Navigate to table in Review