Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group


First Advisor

Salimah H. Meghani


Pain is one of the most burdensome symptoms for patients with cancer. Per cancer pain guidelines, opioids remain one of the primary modalities for managing moderate to severe cancer pain. Analgesic nonadherence is common among cancer patients despite unmanaged pain symptoms. We investigated how patients prioritized analgesic treatment beliefs for cancer pain and whether those beliefs predicted objective analgesic adherence behaviors.

This is a secondary analysis of an existing dataset (n=207) that used a three-month prospective observational design. Subjects were from outpatient oncology clinics of a large Philadelphia health system and were > 18 years, self-identified as African-American or White, diagnosed with solid tumor or multiple myeloma, and prescribed at least one around-the-clock analgesic for reported cancer pain.

We conducted three studies to achieve the aims. First, we performed a concept analysis (Chapter 2) of analgesic nonadherence for cancer pain and qualified its utility in the context of the United States opioid epidemic. In Chapter 3, we used maximum difference scaling to identify how patients traded-off on analgesic treatment beliefs. Utilities (importance scores) were ranked using a k means cluster analysis; clusters were compared in terms of key variables. Finally, we employed general linear modeling to evaluate if analgesic belief clusters predicted analgesic adherence behaviors, assessed longitudinally using electronic medication monitoring while accounting for relevant confounders (Chapter 4).

Initial results showed beliefs weigh significantly in subjective analgesic trade-offs. We identified two distinct belief clusters. Side effect severity was the only variable that significantly differed between clusters. Subjects mostly traded-off based on the belief, ‘pain medicines keep you from knowing what is going on in your body.’ Addiction was not a top priority. Belief clusters did not predict analgesic adherence. However, in an adjusted analysis, it was the experiential variables (e.g., side effects, most potent analgesia, pain relief with analgesics, duration of disease), as well as patients’ race, that were statistically significant in explaining analgesic adherence.

Our findings suggest that experiential variables rather than analgesic beliefs were associated with analgesic adherence in this sample of cancer outpatients. Additional studies should explore patients’ cancer pain self-management practices while considering patient, provider, and system/ structural factors to optimizing cancer pain management.