Moderators of Psycho-Oncology Therapy Effectiveness: Addressing Design Variable Confounds in Meta-Analysis

This meta-analysis begins to organize the complex field of psycho-oncology intervention for psychological distress using a two-part analysis strategy: preliminary analyses of primary study design variables, reported here, followed by the analysis of variables of substantive interest (person and therapy characteristics) using a structure which accounts for the otherwise confounding potential of moderators found in preliminary analyses. Using 146 published and unpublished prospective controlled trials, outcomes were anxiety, depression, and general distress. Two design variables produced higher effects: studies recruiting patients with elevated baseline distress and/or with no history of distress; and studies with an untreated control group. Exclusion of nonrandomized trials was not empirically justified and exclusion for lack of blindness was also seen to be inappropriate.