Article type
Year
Abstract
Background: Recent initiatives have highlighted the importance of standardised outcome measures in clinical trials to reduce reporting bias, misinformation and wasted resources. Although it has received less attention, the principles behind these efforts remain the same for systematic reviews.
Objectives: To determine which outcomes are assessed and how they are defined in systematic reviews of paediatric surgical interventions.
Methods: MEDLINE, EMBASE, DARE, PROSPERO and the Cochrane Database of Systematic Reviews were searched for any systematic reviews of paediatric surgical interventions published before 2011. All specified outcomes and related definitions were recorded.
Results: Fifteen reviews were identified of which 13 were Cochrane reviews. The mean number of outcomes per review was six, with a range of two to 14 and a mean of two primary outcomes. Outcome definitions were provided in three reviews. Time-points for outcome assessment were only pre-specified in one review. Mortality, adverse events (only one review was specific about which adverse events were to be recorded) and quality of life (only one specified for whom quality of life was being assessed) were the most commonly reported outcomes (five reviews each).
Conclusions: Outcomes assessed in systematic reviews of paediatric surgical interventions are inconsistent, poorly defined and reported, and are therefore susceptible to a high risk of outcome reporting bias. While we have assessed reviews covering a range of surgical treatments, there is a core set of outcomes that should be assessed in all reviews of paediatric surgery, e.g. mortality, adverse events and quality of life for both the child and carer. As the Cochrane Collaboration publishes the most systematic reviews of paediatric surgical interventions, we suggest that they develop and promote a set of core outcomes for inclusion in systematic reviews of paediatric surgical interventions. In addition, a set of standardised outcome definitions may improve the reporting quality of systematic reviews.
Objectives: To determine which outcomes are assessed and how they are defined in systematic reviews of paediatric surgical interventions.
Methods: MEDLINE, EMBASE, DARE, PROSPERO and the Cochrane Database of Systematic Reviews were searched for any systematic reviews of paediatric surgical interventions published before 2011. All specified outcomes and related definitions were recorded.
Results: Fifteen reviews were identified of which 13 were Cochrane reviews. The mean number of outcomes per review was six, with a range of two to 14 and a mean of two primary outcomes. Outcome definitions were provided in three reviews. Time-points for outcome assessment were only pre-specified in one review. Mortality, adverse events (only one review was specific about which adverse events were to be recorded) and quality of life (only one specified for whom quality of life was being assessed) were the most commonly reported outcomes (five reviews each).
Conclusions: Outcomes assessed in systematic reviews of paediatric surgical interventions are inconsistent, poorly defined and reported, and are therefore susceptible to a high risk of outcome reporting bias. While we have assessed reviews covering a range of surgical treatments, there is a core set of outcomes that should be assessed in all reviews of paediatric surgery, e.g. mortality, adverse events and quality of life for both the child and carer. As the Cochrane Collaboration publishes the most systematic reviews of paediatric surgical interventions, we suggest that they develop and promote a set of core outcomes for inclusion in systematic reviews of paediatric surgical interventions. In addition, a set of standardised outcome definitions may improve the reporting quality of systematic reviews.