Using a remote health monitoring system to understand the pain experience and pain management decision-making among advanced cancer patients in the home.

317 Background: Examining the pain experiences of advanced cancer patients in the home is essential to improve palliative care and inform tailored interventions. Methods: This analysis uses data from a remote health monitoring system designed by our interdisciplinary team, the Behavioral and Environmental Sensing and Intervention for Cancer. Using a custom smartwatch application, patients record episodes of breakthrough pain via ecological momentary assessments (EMAs) over 14-days. Patients were recruited from a palliative care clinic and local hospice, had locally advanced or metastatic cancer, and pain requiring PRN opioids. Through initial (on-demand) and 40-minute scheduled follow-up EMAs, patients self-report pain severity (0-10) and use of pharmacologic or non-pharmacologic pain management approaches. We used R to calculate descriptive statistics and run generalized linear and mixed models to describe patterns in pain severity and pain management decision making. Results: A total of 19 patients reported 732 breakthrough pain episodes, 450 of which were complete (i.e., an initial EMA plus its associated follow-up EMA). Patients reported an average of 29.2 episodes (95% CI: 22.2-38.5) over the 2-week period with an estimated rate of 2.1 pain episodes per day (95% CI: 1.6-2.8). Average initial and follow-up pain severity were 4.85 (95% CI: 4.06-5.64) and 3.76 (95% CI: 2.89-4.63), respectively. Of complete episodes, 56.4% showed a decrease in pain severity. On average patients experienced a decrease of 1.02 points in pain severity (95% CI: -1.51-(-0.53)) from initial to follow-up. In 33.3% (n = 150) of complete episodes, patients had no change in pain severity score. In 67.8% of complete episodes, patients took an opioid, with an average probability of taking an opioid of 74.7% (95% CI: 0.59-0.86). Patients reported taking an opioid in 52.7% of episodes with no change in pain severity. In 39.1% of episodes patients reported using non-opioid or non-pharmacologic approaches (in 73.3% of those episodes, patients also took an opioid). Conclusions: Our findings may indicate gaps in pain management education (particularly around non-opioid or non-pharmacologic pain management approaches), sub-optimal opioid prescribing, or possible self-rationing of opioids. Additional research is needed to identify opportunities for intervening when patients with advanced cancer experience inadequate pain management.