We planned to do analyses for outcomes closest to 12 weeks, 26 weeks, and 52 weeks, but few data were available for 26 weeks and 52 weeks and analyses at these timepoints were, therefore, not possible. This scarcity of data reflects ethical issues associated with doing long-term, placebo-controlled, randomised controlled trials of effective treatments. Thus, our findings can inform only the choice of short-term medication treatment for ADHD. Moreover, because of a paucity of data, we were unable to properly undertake all the planned sensitivity analyses. Finally, we did not include studies of antipsychotic or tricyclic antidepressant compounds because, although commonly prescribed for patients with ADHD, they are not used routinely to treat ADHD core symptoms, and their inclusion would, therefore, violate the assumption of transitivity in the networks.
