Sixteen studies were included, eight of which provided data on relapse. Although heterogeneity in the data warrants caution,psychoeducation appeared to be effective in preventing any relapse[n = 7; OR: 1.98–2.75; number needed to treat (NNT): 5–7, depending on the method of analysis] and manic/hypomanic relapse (n = 8; OR:
1.68–2.52; NNT: 6–8), but not depressive relapse. Group, but not individually, delivered interventions were effective against both poles of relapse; the duration of follow-up and hours of therapy explained some of the heterogeneity. Psychoeducation improved medication adherence
and short-term knowledge about medication. No consistent effects on
mood symptoms, quality of life, or functioning were found