According to a commentary, what is puzzling about the trial is how badly all groups fared; none of the 21 patients in the antipsychotic group, six of 20 in the CBT group, and four of 21 in the combined group met response criteria at 1 year—around 16% overall. It suggests that this might be due to poor adherence to treatment, and also antipsychotic doses do not seem to have been increased despite non-response, and clozapine was not considered in anyone, despite clinical guidelines. It adds that the trial did not have a valid active control group or a psychosocial placebo group to measure possible non-specific effects of therapy and symptom fluctuations.