In this section, and in Sections 14.2.2 and 14.2.3, we describe three broad strategies that a Cochrane review may use to address adverse effects. The first strategy is to assess intended (beneficial) and unintended (adverse) effects together using the same methodology, applying common eligibility criteria (in terms of types of studies, types of participants and types of interventions).
This approach implies that a single search strategy may be used. A critical issue is how review authors deal with the three datasets that may potentially arise:
(a) Studies that report both the beneficial effects and adverse effects of interest;
(b) Studies that report beneficial effects but not adverse effects;
(c) Studies that report adverse effects, but not the beneficial outcomes of interest.
Studies of type (a) have the important advantage that benefits and adverse effects can be compared directly, since the data are derived from the same population and setting. Furthermore, evidence on benefits and adverse effects arises from studies with similar designs and quality. However, data on adverse effects may be very limited and in particular may be restricted to short-term harms because of the relatively short duration of included studies.
Evaluation of benefits and adverse effects using some combination of the three types of study (rather than (a) alone) will increase the amount of information available. For instance, datasets (a) and (b) could be used to evaluate beneficial effects, while (a) and (c) could be used to assess adverse effects. However, as the studies addressing adverse effects differ from those addressing beneficial effects, authors should note that it is difficult to compare benefits and adverse effects directly.