Flach, ClareFrench, PaulDunn, GrahamFowler, DavidGumley, Andrew I.Birchwood, MaxStewart, Suzanne L. K.Morrison, Anthony P.2014-12-102014-12-102015-05-21The British Journal of Psychiatry, 207(2), 123-1290007-125010.1192/bjp.bp.114.153320http://hdl.handle.net/10034/337035This is an author-produced electronic version of an article accepted for publication in the British Journal of Psychiatry. The definitive publisher-authenticated version is available online at http://bjp.rcpsych.orgBackground: Research suggests that the way in which cognitive therapy is delivered is an important factor in determining outcomes. We test the hypotheses that the development of a shared problem list, use of case formulation, homework tasks and active intervention strategies will act as process variables. Methods: Presence of these components during therapy is taken from therapist notes. The direct and indirect effect of the intervention is estimated by an instrumental variable analysis. Results: A significant decrease in symptom score for case formulation (coefficient=-23, 95%CI -44 to -1.7, p=0.036) and homework (coefficient=-0.26, 95%CI -0.51 to -0.001, p=0.049) is found. Improvement with the inclusion of active change strategies is of borderline significance (coefficient= -0.23, 95%CI -0.47 to 0.005, p=0.056). Conclusions: There is a greater treatment effect if formulation and homework are involved in therapy. However, high correlation between components means that these may be indicators of overall treatment fidelity.entherapyArticle1472-1465British Journal of Psychiatry