The selection of adverse outcomes to include in a review can be difficult. Specific adverse effects associated with an intervention may be known in advance of the review; others will not. Which effects will be most relevant to the review may be uncertain beforehand. The following general strategies may be used depending on the study question and the therapeutic or preventive context.
A detailed analysis of one or two known or a few of the most serious adverse effects that are of special concern to patients and health professionals.
Advantages: Easiest approach, especially with regard to data collection. Can focus on important adverse effects and reach a meaningful conclusion on issues that have a major impact on the treatment decision (McIntosh 2004).
Disadvantages: Scope may be too narrow. Method is only really suitable for adverse events that are known in advance.
To detect a variety of adverse effects, whether known or previously unrecognized.
Advantages: Wider coverage, and can evaluate new adverse effects that we may not have previously been aware of.
Disadvantages: Potentially large volume of work with particular difficulties in the data collection process. Some researchers have found broad, non-specific evaluations to be very resource-intensive, with little useful information to show for the effort expended (McIntosh 2004). These researchers also point out that previously unrecognized adverse effects may be best detected through primary surveillance, rather than in a systematic review.
In order to address adverse effects in a more organized manner, review authors may choose to narrow down the broad focus into some of the following areas:
The five to ten most frequent adverse effects;
All adverse effects that either the patient or the clinician considers to be serious;
By category, for example:
Diagnosed by lab results (e.g. hypokalaemia);
Patient-reported symptoms (e.g. pain).