4.2.4.4 Outcomes

The review protocol should list all the outcomes considered. There is an international initiative known as The COMET (Core Outcome Measures in Effectiveness Trials) initiative, involved in the development and application of agreed standardized sets of outcomes for trials on specific conditions. Details are provided on the COMET website (http://www.comet-initiative.org/). Reviewers are encouraged to check the available standardized sets of outcomes for trials relevant for their reviews.

Outcomes should be measurable and appropriate to the review objectives and questions. Usually, only a limited number of primary outcomes and a limited number of secondary outcomes are considered for a review. Sometimes, if justified, it is acceptable to include multiple primary and secondary outcomes. However, the appropriateness of the number and scope of outcomes depend on the specifics of the review objectives and review questions (Aromataris 2015). The relevance of each outcome to the review objective/questions should be justified in the background section. Both beneficial outcomes (positive effects) and harms (negative effects, such as adverse effects or side effects) should be considered as outcomes (Aromataris 2015). Essentially, primary outcomes are those outcomes that are the most important outcomes informing the review questions and the conclusions about the beneficial and harmful effects of the intervention of interest for a review (Aromataris 2015). Secondary outcomes are all other outcomes not specified as primary outcomes. A fundamental distinction is that between true endpoints and surrogate outcomes; true endpoints reflect the effects of treatment on aspects of patients’ status considered the most important in terms of mortality and morbidity; surrogate outcomes are measured as “surrogates’ for true endpoints, for reasons related to complexity, time, and costs of measurement of true endpoints (Tufanaru 2016). Examples of true endpoints are survival time in cancer and bone fractures in osteoporosis; examples of surrogate outcomes are time to progress from one stage to another stage in cancer and bone mineral density in osteoporosis (Tufanaru 2016).

It is recommended that whenever possible true endpoints should be used as primary outcomes, and that if surrogate outcomes are used as primary outcomes then an explicit justification should be provided for the use of a surrogate outcomes instead of true endpoints (Tufanaru 2016). It is expected that all outcomes specified a priori in the review protocol, will be explicitly addressed in the systematic review report, regardless of the existence or not of data from included studies on these outcomes (Aromataris 2015).

A further critical aspect refers to the measurement of the specified outcomes. It is recommended that reviewers present explicit information on available measurement instruments, including details about the validity and reliability properties of these instruments (Aromataris 2015).

As JBI endorses the use of the GRADE approach known as the ‘Summary of findings’ table, reviewers should be aware that the most important outcomes, that is, the primary outcomes specified in the review protocol should be addressed in the review report and should be explicitly presented in the GRADE Summary of findings’ table. Details are provided in the GRADE Handbook (Schunnemann et al. 2013).