What is driving declines in oncology-prescribed opioids?

6583 Background: Patients with cancer have experienced sharp declines in opioid prescribing during the past decade, raising concerns about insufficient pain management in a population with high rates of undertreated pain. It is unclear whether these declines are driven by certain oncology subspecialties, payors, or clinical guidelines regarding opioid prescribing released in/after March 2016. Methods: We conducted time series analyses using 2006-2019 IQVIA Longitudinal Prescription Data, which represents up to 92% of all U.S. prescription fills. Using log-linear regression stratified by group, we estimated annual percent change in opioid receipt by: subspecialty (medical oncology MO, radiation oncology RO, and surgical oncology SO); payor (third-party; Medicare, including Medicare Part D; Medicaid; cash/out-of-pocket); and pre- versus post-March 2016. Opioid receipt outcomes included: monthly fills of any opioid; monthly fills of extended-release/long-acting (ER/LA) opioids; and monthly number of patients receiving long-term opioid therapy (LTOT). Outcomes were per 1,000 survivors to account for temporal changes in cancer prevalence. Results: Between 2006 and 2019, 14,301,900 opioid fills were prescribed by oncologists to 3,476,354 distinct patients. Across all outcomes, MO had the highest levels of dispensing, followed by RO, then SO; for example, in 2006 MO dispensed 6.2 fills per 1,000 survivors (compared to 1.9 fills among RO, and 0.5 among SO). We observed substantial declines in all opioid outcomes. Per 1,000 survivors, there was an annual decline of: 5.7% (95CI: 5.1-6.3; total unadjusted decline 2006-2019=70.2%) in the rate of all opioid fills; 4.9% (95CI: 4.1-5.6; total unadjusted decline=66.8%) in ER/LA fills; 3.2% (95CI: 1.5-5.0; total unadjusted decline=56.2%) in LTOT fills, and 1.9% in average daily dose (95CI: 1.9-2.0; total unadjusted decline=29.8%). The annual decline in opioid fills prescribed by MO (7.1%, 95CI: 6.6-7.7) was sharper than for RO (5.9%, 95CI: 5.4-6.4) and SO (5.2%, 95CI: 4.7-5.7). Annual declines were steepest among fills paid out-of-pocket (14.5%, 95CI: 13.9-15.0), followed by those paid by Medicaid (13.1%, 95CI: 12.5-13.6), third-party payors (9.3%, 95CI: 8.7-9.9), and Medicare (2.4%, 95CI: 1.8-3.1). Declines in every outcome accelerated following clinical guidelines released in/after 2016. Conclusions: Opioids prescribed by oncologists declined dramatically across groups and outcomes. Medical oncologists were responsible for a disproportionate share of opioid fills in 2006, and for the sharpest declines. Out-of-pocket fills declined more sharply than fills covered by insurance and clinical guidelines may have contributed to accelerating declines. While de-escalation of opioid therapy may reduce risk of opioid-related harms to cancer survivors, care is needed to ensure cancer-related pain is appropriately treated.