Results of the search

{{#pick}} Our search identified {{input 'number' '[SEARCH TOTAL]' 'Initial number of articles in your PRISMA diagram'}} {{#pick}}articles / publications / records / studies{{/pick}}. The database searches found {{input 'number' '[SEARCH TOTAL]' 'Initial number of articles in your PRISMA diagram'}} {{#pick}}articles / publications / records / studies{{/pick}}. {{/pick}} {{#pick}} We found {{input 'number' '[SUPPLEMENTARY]' 'Number of additional articles found from other sources in your PRISMA diagram'}} citations from (additional/other) sources. Through supplementary methods, we identified {{input 'number' '[SUPPLEMENTARY]' 'Number of additional articles found from other sources in your PRISMA diagram'}} {{#pick}}articles / publications / records / studies{{/pick}}. We identified {{input 'number' '[SUPPLEMENTARY]' 'Number of additional articles found from other sources in your PRISMA diagram'}} additional {{#pick}}articles / publications / records / studies{{/pick}} through supplementary methods. {{/pick}} {{#pick}} The total number of {{#pick}}articles / publications / records / studies{{/pick}} found was {{input 'number' '[TOTAL]' 'Total number of articles identified'}}. In total we identified {{input 'number' '[TOTAL]' 'Total number of articles identified (sum of last two numbers)'}} {{#pick}}articles / publications / records / studies{{/pick}}. {{/pick}} {{#pick}} After removing duplicates the number of {{#pick}}articles / publications / records / studies{{/pick}} was {{input 'number' '[UNIQUE]' 'Number of articles remaining after de-duplication'}}. After removing duplicates the unique number of {{#pick}}articles / publications / records / studies{{/pick}} was {{input 'number' '[UNIQUE]' 'Number of articles remaining after de-duplication'}}. Removal of duplicates left {{input 'number' '[UNIQUE]' 'Number of articles remaining after de-duplication'}} {{#pick}}articles / publications / records / studies{{/pick}}. Removal of duplicates left unique {{input 'number' '[UNIQUE]' 'Number of articles remaining after de-duplication'}} {{#pick}}articles / publications / records / studies{{/pick}}. {{/pick}} {{#pick}} We screened by title and abstract, excluding {{input 'number' '[IRRELEVANT]' 'Number of articles excluded after full-text screening'}} to leave {{input 'number' '[FULL TEXT]' 'Final total of articles post title/abstract screening'}} which we screened in full-text. We screened those {{#pick}}articles / publications / records / studies{{/pick}} by title and abstract and excluded {{input 'number' '[IRRELEVANT]' 'Number of articles excluded after title/abstract screening'}}, leaving {{input 'number' '[FULL TEXT]' 'Final total of articles post title/abstract screening'}}, which were screened in full-text. {{/pick}} {{#pick}} During full-text screening we excluded {{input 'number' '[EXCLUDED]' 'Article count excluded during full-text screening'}} {{#pick}}articles / publications / records / studies{{/pick}}. We excluded {{input 'number' '[EXCLUDED]' 'Article count excluded during full-text screening'}} full-text {{#pick}}articles / publications / records / studies{{/pick}}. {{/pick}} {{#pick}} We included {{input 'number' '[INCLUDED]' 'Number of included studies'}} {{#pick}}articles / publications / records / studies{{/pick}}, and were able to extract outcome data from {{input 'number' '[EXTRACTED]' 'Number of studies it was possible to extract data from'}}. From {{input 'number' '[INCLUDED]' 'Number of included studies'}} included studies, {{input 'number' '[EXTRACTED]' 'Number of studies it was possible to extract data from'}} of studies had data that could be extracted. From {{input 'number' '[INCLUDED]' 'Number of included studies'}} included studies, {{input 'number' '[EXTRACTED]' 'Number of studies it was possible to extract data from'}} of studies had extractable data. {{/pick}}

Outcomes

{{#ifMultiple analysesAndData.comparison}} {{#pick}} For this review we generated {{formatNumber analysesAndData.comparison.length}} comparisons. {{formatNumber analysesAndData.comparison.length}} comparisons were used in this review. We specified {{formatNumber analysesAndData.comparison.length}} comparisons. We specified {{formatNumber analysesAndData.comparison.length}} comparisons in this review. {{formatNumber analysesAndData.comparison.length}} comparisons are included in this review. For this review we examined {{formatNumber analysesAndData.comparison.length}} comparisons. This review involved {{formatNumber analysesAndData.comparison.length}} comparisons. {{/pick}} {{/ifMultiple}} {{#ifSingle analysesAndData.comparison}} {{#pick}} For this review we generated a single comparison. A single comparison was used in this review. This review involved only one planned comparison. For this review we examined a single comparison. For this review we specified a single comparison. This review involved one planned comparison. A single comparison is included in this review. {{/pick}} {{/ifSingle}}

{{name}}

{{#ifSingle outcome}} {{#pick}} This comparison has a single outcome. There was only a single outcome for this comparison. There was one outcome for this comparison. There was a single outcome for this comparison. This comparison involved a single outcome only. There was one outcome for this comparison. {{/pick}} {{/ifSingle}} {{#ifMultiple outcome}} {{#pick}} This comparison has {{formatNumber outcome.length}} outcomes. There were {{formatNumber outcome.length}} outcome[s] for this comparison. This comparison involved {{formatNumber outcome.length}} outcomes. {{/pick}} {{/ifMultiple}}

{{name}}

{{#ifValue estimable '===' false}} {{#pick}} The effect estimate could not be calculated. The effect size could not be estimated. An effect estimate could not be calculated. An effect size could not be estimated. {{/pick}} {{/ifValue}} {{#ifValue estimable 'undefOr' true}} {{#ifNone study}} {{#pick}} No studies reported sufficient data for this outcome. No studies provided data relevant to this outcome. {{/pick}} {{/ifNone}} {{#ifSingle study}} {{#pick}} We identified one study relevant to this outcome. We identified one study that provided usable data relevant to this outcome. Data relevant to this outcome were reported in one study. {{/pick}} {{/ifSingle}} {{#ifMultiple study}} {{#pick}} We identified {{formatNumber studies}} studies relevant to this outcome. For this outcome we identified {{formatNumber studies}} relevant studies. {{formatNumber studies}} studies provided adequate data for this outcome. {{formatNumber studies}} studies provided relevant data for this outcome. We identified {{formatNumber studies}} studies that reported relevant data for this outcome. {{/pick}} {{/ifMultiple}} {{#pick}} , with a total of {{formatNumber participants}} participants. , involving a total of {{formatNumber participants}} participants. (total n={{formatNumber participants}}). involving {{formatNumber participants}} participants. {{/pick}} {{#ifSingle subgroup}} {{#pick}} This outcome contained a single subgroup. A single subgroup was used in this outcome. {{/pick}} {{/ifSingle}} {{#ifMultiple subgroup}} {{#pick}} We divided the data into {{formatNumber subgroup.length}} subgroups in accordance with our protocol. As specified in our protocol, we divided the data into {{formatNumber subgroup.length}} subgroups. The data were divided into {{formatNumber subgroup.length}} groups as specified in the protocol. Data were divided into the {{formatNumber subgroup.length}} pre-specified subgroups. We categorised data into {{formatNumber subgroup.length}} subgroups (in keeping with our protocol). There were {{formatNumber subgroup.length}} subgroups, as specified by the protocol. We categorised data into {{formatNumber subgroup.length}} subgroups as specified in our protocol. As specified in our protocol, we categorised data into {{formatNumber subgroup.length}} subgroups. {{/pick}} {{/ifMultiple}} {{#ifValue i2 'between' '30 and 70'}} {{#pick}} This outcome had moderate levels of heterogeneity (I² = {{formatNumber i2 2}}%). This outcome had moderate levels of heterogeneity, I² = {{formatNumber i2 2}}%. The heterogeneity of effects for this outcome was moderate (I² = {{formatNumber i2 2}}%). The heterogeneity of effects for this outcome [in this subgroup] was moderately high (I² = {{formatNumber i2 2}}%). Heterogeneity within this subgroup was moderate (I² = {{formatNumber i2 2}}%). This outcome had a moderate level of heterogeneity (I² = {{formatNumber i2 2}}%). The [subgroup] outcomes had moderate levels of heterogeneity (I² = {{formatNumber i2 2}}%). For this outcome heterogeneity is moderately high {{formatNumber i2 2}}, I² = {{formatNumber i2 2}}%. [In this subgroup], heterogeneity was found to be high (I² = {{formatNumber i2 2}}%) for this outcome. Outcomes for [this subgroup] had high heterogeneity (I² = {{formatNumber i2 2}}%). {{/pick}} {{/ifValue}} {{#ifValue i2 'gt' 70}} {{#pick}} This outcome had high levels of heterogeneity (I² = {{formatNumber i2 2}}%). This outcome had high levels of heterogeneity, I² = {{formatNumber i2 2}}%. For this outcome heterogeneity is high {{formatNumber i2 2}}, I² = {{formatNumber i2 2}}%. For this outcome heterogeneity was high {{formatNumber i2 2}}, I² = {{formatNumber i2 2}}%. Heterogeneity was found to be high {{formatNumber i2 2}}, I² = {{formatNumber i2 2}}% for this outcome. Outcomes for [this subgroup] had high heterogeneity (I² = {{formatNumber i2 2}}%). {{/pick}} {{/ifValue}} {{#ifValue p 'lt' 0.05}} {{#pick}} We found evidence of a clear difference between {{groupLabel1}} and {{groupLabel2}}, For this outcome, we found evidence that {{groupLabel1}} was clearly different in its effects compared with {{groupLabel2}}, For this outcome, we did find evidence that {{groupLabel1}} was clearly different in its effects compared with {{groupLabel2}}, There was a clear difference in this outcome between {{groupLabel1}} and {{groupLabel2}}, {{/pick}} {{/ifValue}} {{#ifValue p 'gte' 0.05}} {{#pick}} There was not a clear difference between {{groupLabel1}} and {{groupLabel2}}, We did not find evidence of a clear difference between the two treatments in this comparison, We did not find evidence of a difference between the two treatments in this comparison, {{/pick}} {{/ifValue}} {{formatLowerCase effectMeasureText}} was {{formatNumber effectSize 2}} ({{formatNumber ciTotal 2}}% CI {{formatNumber ciStart 2}} to {{formatNumber ciEnd 2}}, {{formatP p}}).

{{name}}

{{#ifSingle study}} {{#pick}} We found one trial to be relevant to this subgroup, There is a single trial in this subgroup, We found one trial with data relevant to this subgroup, {{/pick}} {{/ifSingle}} {{#ifMultiple study}} {{#pick}} We found {{formatNumber study.length}} trials to be relevant to this subgroup There are {{formatNumber study.length}} relevant trials in this subgroup {{formatNumber study.length}} studies reported the number of events {{/pick}} {{/ifMultiple}} {{#pick}} , which included a total of {{formatNumber participants}} participants. , with a total of {{formatNumber participants}} people. (total n={{formatNumber participants}}). , with a total of {{formatNumber participants}} study participants. {{/pick}} {{#ifValue i2 'between' '30 and 70'}} {{#pick}} This subgroup had moderate levels of heterogeneity (I² = {{formatNumber i2 2}}%). For this subgroup heterogeneity is moderately high (I² = {{formatNumber i2 2}}%). {{/pick}} {{/ifValue}} {{#ifValue i2 'gt' 70}} {{#pick}} This subgroup had high levels of heterogeneity (I² = {{formatNumber i2 2}}%). For this subgroup heterogeneity is high (I² = {{formatNumber i2 2}}%). {{/pick}} {{/ifValue}} {{#ifValue p 'lt' 0.05}} {{#pick}} We found evidence of a clear difference between {{groupLabel1}} and {{groupLabel2}} within this subgroup For this outcome, within this subgroup, we did find evidence that {{groupLabel1}} was clearly different in its effects compared with {{groupLabel2}} There was evidence that the intervention was clearly beneficial when compared with the control There was evidence that the intervention was clearly beneficial compared to the control {{/pick}} {{/ifValue}} {{#ifValue p 'gte' 0.05}} {{#pick}} There was not a clear difference between {{groupLabel1}} and {{groupLabel2}} within this subgroup For this subgroup, we did not find evidence of a clear difference between the two treatments {{/pick}} {{/ifValue}} {{formatLowerCase ../effectMeasureText}} was {{formatNumber effectSize 2}} ({{formatNumber ../ciTotal 2}}% CI {{formatNumber ciStart 2}} to {{formatNumber ciEnd 2}}, {{formatP ../p}}).

Risk of bias

{{#pick}} {{input 'number' '[INCLUDED]' 'The number of suitable studies from the Risk of Bias diagram'}} of included studies were mostly {{input 'choice' '[low / unclear / high]' 'The RoB on the majority of studies'}} risk of bias. {{/pick}} {{input 'text' '[RISK OF BIAS TEXT]' 'Comments on the Risk of Bias domains which have a large proportion of unclear or high risk judgements'}}