Cancer Treatment Side Effects: A Meta-analysis of the Relationship Between Response Expectancies and Experience

Abstract

Context: Although previous research has, overall, suggested a moderate relationship between response expectancies (REs) and cancer treatment-related side effects, empirical results have been mixed.

Objectives: We aimed to further explore these relationships, hypothesizing that REs would predict subsequent toxicities with the inclusion of more recent studies, across a broader range of side effects, while incorporating the impact of potential moderators including patients' experience with treatment and measurement methods. We further investigated the impact of REs across individual toxicities.

Methods: A systematic search and analysis were conducted across four databases (PsychInfo, PubMed, CINAHL, and Embase) and reference lists, from 1985 to February 2016. This provided 27 eligible studies with 4474 participants, through which the main analysis, moderator analyses, and individual side-effect analyses were explored.

Results: REs were moderately related to side effects overall (r = 0.26), and effect sizes were significantly influenced by sample diagnostic homogeneity, whereas differences between type and timing of measurement showed trends. Of the 16 toxicities examined, 15 demonstrated significant relationships between REs and side-effect experience, with hair loss (r = 0.48) the strongest. No clear difference emerged between objective and subjective side effects; however, significant differences across individual toxicities were revealed.

Conclusion: Findings support a relationship between REs and a wide range of subsequent side effects, yet differences between individual RE-toxicity associations emerged. These findings provide direction for the measurement of side effects and REs and support REs as potential targets for intervention during the informed consent process.

Keywords: Medical oncology; chemotherapy; informed consent; placebo response; toxicity.