“…Specifically, future studies should: (1) employ methods consistent with contemporary standards of psychotherapy research, including randomization of subjects and use of control/comparison groups; (2) use standardized, verifiable, and reproducible therapeutic procedures; (3) test interventions (including those dealing with cognition and behavior or with expressed emotion) likely to be particularly pertinent to individuals with bipolar disorder; (4) enroll enough subjects to assure sufficient statistical power to detect changes related to experimental interventions; and (5) empsychotherapeutic interventions, although reliable assessment of medicine-taking, particularly among outpatients, is difficult. 29,[65][66][67]73,78,88,97 A particularly reassuring finding in all 11 studies that considered the matter was reduced risk of rehospitalization with various psychosocial interventions (see Table 4). 29,[58][59][60][61][67][68][69]76,77,81 Interpretation and generalization of these findings are complicated by major changes in indications for and financing of psychiatric hospitalization in recent years.…”