This is an archived version. For the current version, please go to training.cochrane.org/handbook/current.

As the discussion in Section 17.8.1 pointed out, when pooling across PROs the mean difference is no longer a possible measure of effect and we therefore replace it with the standardized mean difference (SMD) (see Chapter 9, Section 9.2.3). Unfortunately, there are no fully satisfactory ways of providing a sense of the magnitude of effect in a PRO when one has had to resort to SMD to generate a summary. One can offer readers standard rules of thumb in interpretation of effect sizes (for instance 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect (Cohen 1988) or some variation (<0.41 = small, 0.40 to 0.70 = moderate, >0.70 = large). Another, perhaps even less satisfactory, approach suggests that a standardized mean difference of 0.5 approximates, in many cases, to a minimal important difference (Norman 2003).

General methods of reporting and interpreting PROs, and other clinical outcomes, with respect to drawing inferences and conclusions are discussed in Chapter 12 (Section 12.6).