First do no harm: how can systematic reviewers do justice to harms? Deciding which harms to search for and how

Session Type
Workshop - training
Adverse effects
Target audience
Information Specialists, Review Authors
Level of difficulty

Background: Any intervention that can have an effect has the potential to have an adverse effect. All systematic reviews of interventions should at least consider the harms of that intervention. Harms are important to patients, impacting on morbidity and mortality. There are many differences (as well as similarities) in how researchers should approach a review question on harms, as opposed to a question on benefit. Formulating the question on harms is a challenging but key step in the review process. Review teams need to decide whether to focus on specific harms (and if so which ones) or whether to have a broader objective. Balancing the quest for an answerable, useful question with limited resources is difficult. Searching for included studies on harms can also be problematic and often requires a different approach to searching for effectiveness studies. Obstacles include the lack of harm information in journal abstracts, inconsistent terminology and indexing and the need to search a range of study designs beyond randomised controlled trials (RCTs).
Objectives: To give guidance to review authors on formulating the question, signposting the available data sources on harms and retrieving data for incorporation into a systematic review. This will include the development of optimal search techniques for different interventions and discussion of the impact of CONSORT-harms.
Description: We will facilitate discussion on the advantages and disadvantages of a range of approaches to question formulation in systematic reviews incorporating harms. Next, participants will receive a number of scenarios and will work together in small groups to plan a search strategy for a comprehensive evaluation of harms. The scenarios will be drawn from real-life situations to cover a wide range of potential harms. At the end, the groups will feed back on their search protocols and any points raised will be discussed further. Examples from existing reviews will be provided by the facilitator, as well as tips and tricks to solve specific issues, and examples of sources of support and help available.

Kwong J1
1Cochrane Central Editorial Service,