Journal of Pain Research (Dec 2017)
Psychosocial predictors of posttreatment pain after nonmetastatic breast cancer treatment: a systematic review and meta-analysis of prospective studies
Abstract
M Johannsen,1,2 Y Frederiksen,1,2 AB Jensen,3 R Zachariae1,2 1Unit for Psychoncology and Health Psychology, Department of Oncology, Aarhus University Hospital, 2Department of Psychology, Aarhus University, 3Department of Oncology, Aarhus University Hospital, Aarhus C, Denmark Background: The search for risk factors of pain after breast cancer, which affects a considerable proportion of the women, has primarily focused on clinical factors. The aim of this meta-analysis was to explore the less well-studied psychosocial predictors of pain after breast cancer treatment. Methods: Two independent searches were conducted in PubMed, PsycINFO, Web of Science, and CINAHL. Eligible studies were prospective, observational studies of women aged ≥18 years, diagnosed and treated for nonmetastatic breast cancer ≥3 months previously. Additional inclusion criteria were that studies had assessed at least one pain outcome and at least one psychosocial predictor. The psychosocial predictors investigated included: 1) psychological–behavioral states, 2) psychological traits, and 3) social support. Effect size correlations (ESr) were chosen as the effect size and pooled using a random effects model. We also explored a number of study characteristics as possible moderators of the effect with meta-regression. Results: Of the total of 13 eligible studies identified, most studies measured psychosocial predictors at presurgery. Neither psychological–behavioral states (ESr: 0.05; p=0.13; K=11) nor psychological traits (ESr: 0.02; p=0.48; K=6) emerged as statistically significant predictors of pain. In contrast, higher levels of social support were statistically significantly associated with less pain (ESr: −0.24; p<0.001; K=4). In studies of psychological–behavioral states, longer follow-up was associated with smaller effect sizes (p=0.023). Furthermore, older mean sample age was associated with larger effect sizes for both psychological–behavioral states (p=0.0004) and psychological traits (p=0.035).Conclusion: The results of this meta-analysis suggest that psychosocial factors measured at presurgery may only be of modest predictive value in identifying women at risk of developing pain after breast cancer treatment. While speculative, psychosocial factors may play a larger role in the postsurgery trajectory, which could be valuable to investigate in future studies. Keywords: breast cancer, pain, predictors, personality traits, social support, behavioral symptoms
