Tendency to adhere to provider-recommended treatments and subsequent pain severity among individuals with cancer

Anthony Jerant¹, Peter Franks¹, Daniel J Tancredi², Naomi Saito³, Richard L Kravitz^4
1Department of Family and Community Medicine, Center for Healthcare Policy and Research, University of California Davis School of Medicine, Sacramento, CA, USA; ²Department of Pediatrics, Center for Healthcare Policy and Research, University of California Davis School of Medicine, Sacramento, CA, USA; ³Center for Healthcare Policy and Research, University of California Davis School of Medicine, Sacramento, CA, USA; ⁴Department of Internal Medicine, Division of General Medicine, Center for Healthcare Policy and Research, University of California Davis School of Medicine, Sacramento, CA, USA

Background: Patients’ general tendency to adhere to health care provider-recommended treatments is associated with a number of health outcomes, but whether it influences pain severity over time among individuals with cancer is unclear. We explored the relationship between adherence tendency and subsequent pain severity among cancer patients participating in a randomized controlled trial of coaching to enhance communication with physicians and reduce pain severity.
Methods: Associations between baseline Medical Outcomes Study general adherence measure score and pain severity over 12 weeks were examined with repeated-measures regression models. Model 1 included sociodemographics, study group and site, follow-up point, and baseline pain; Model 2 included these variables plus partner status, physical and mental health status (12-item Short Form Health Survey [SF-12®]), and pain control self-efficacy.
Results: Of 307 patients randomized, 224 (73%) had at least one follow-up pain severity assessment plus complete data for other model variables and were included in the analyses. In Model 1, adherence tendency was associated with less subsequent pain severity: a one standard deviation increase in adherence tendency was associated with a 0.22-point adjusted mean decrease in pain severity on a 0–10 scale (95% confidence interval 0.40, 0.03). The association was diminished and not statistically significant in Model 2, primarily due to adjustment for the SF-12.
Conclusion: Tendency to adhere to provider-recommended treatments was associated with subsequent pain severity among individuals with cancer, suggesting a potential way of predicting and intervening to improve cancer pain control. However, the association was attenuated after adjusting for health status, suggesting mediation or confounding of the relationship by health status.

Keywords: health status, longitudinal studies, neoplasms, outcome assessment (health care)