Predictors of Response to an Evidence-Based Behavioral Cancer Pain Management Intervention: An Exploratory Analysis...

Interventions that teach patients cognitive and behavioral strategies for managing cancer pain have demonstrated effectiveness. Systematic reviews of such interventions call for research to inform their implementation in practice, including investigations into which patients are most likely to benefit.

We aimed to identify predictors of response to an evidence-based behavioral intervention for cancer pain, pain coping skills training (PCST).

We conducted an exploratory secondary analysis of a randomized noninferiority trial comparing in-person to videoconference-based PCST. Using modified Poisson regression, we estimated the adjusted associations of patient characteristics with clinically meaningful reductions (≥30%) in pain severity and pain interference.

Of the 178 patients who were randomized, 135 completed at least one follow-up assessment and were included in this analysis. Proportions of patients experiencing reductions in pain severity and pain interference were 34% and 46%, respectively. In multivariable analysis, education level was associated with a reduction in pain severity (adjusted relative risk, some college or technical school vs. college or higher: 0.49, 95% CI: 0.26-0.93). Patients with colorectal cancer were 61% more likely to experience a reduction in pain interference than patients with breast cancer (95% CI:1.21-2.34). Marital status was also statistically significantly associated with pain interference reduction, with married patients less likely to experience a reduction in pain interference (adjusted relative risk, married vs. not: 0.69, 95% CI: 0.49-0.98).

Our findings elucidate several subgroups of patients who may be especially likely to benefit from PCST, informing both targeted implementation efforts and opportunities to improve delivery for diverse patients.