Prior authorization for supportive care medications at a large academic cancer center.

350 Background: For patients with cancer, supportive care medications (meds) improve quality of life and capacity to receive life-sustaining anti-cancer therapy. Adequate treatment of nausea and pain can decrease hospitalizations and improve therapy tolerance. Supportive care meds, however, often require prior authorization (PA), for which providers must obtain payor pre-approval. PAs may result in delayed or denied care. We aimed to characterize the burden of PA for supportive care meds for patients with cancer. Methods: At an urban comprehensive cancer center, we identified all PA requests for outpatient long-acting opioid (buprenorphine, fentanyl, hydromorphone, methadone, morphine, oxycodone, oxymorphone, tapentadol), key non-opioid pain (pregabalin, duloxetine), and key anti-nausea (ondansetron, granisetron) prescriptions from 1-9/2024 and described frequency of PA outcomes including approvals, denials, and turnaround time (TT) from prescription to decision of approval or denial. For patients with multiple supportive care PA requests, only the first was considered. Fisher’s Exact test and Wilcoxon rank sum test compared TT between categorical variables and continuous variables, respectively. Outcomes at 90 days post-index prescription [urgent care (UCC) or symptom care clinic (SCC) visits, hospitalization, death] were noted through end of 2024. Results: Over 9 months, 897 patients had at ≥1 PA for supportive care meds (714 opioid, 38 non-opioid, 145 nausea). Patients were a median of 62 years (IQR 54-68), 52% female, 80% White race (9% Black, 8% Asian), 90% non-Hispanic, and mostly partnered (66%). Most had private insurance (58%) or Medicare (39%); few had Medicaid (2%). PA requests had a median TT of 0 days (IQR 0-2); 6% of PAs were denied. For denials, median TT was 7 days (IQR 1-14). Patients with longer delays (≥1 TT vs 0 days) were younger (median 59 vs 63 years); 86% of patients with Medicaid had ≥1 day TT vs 33% of those with Medicare (p < 0.001 for both comparisons). There were no differences by race, ethnicity, sex, or marital status. 16% with PA for supportive care meds died within 90 days of PA. One-quarter visited the SCC within 90 days of a PA (27% opioid, 21% non-opioid, 28% nausea) with slightly fewer visiting the UCC (24% opioid, 11% non-opioid, 12% nausea). 22% of those with PA for opioid meds were hospitalized within 90 days, more than for non-opioid (13%) or nausea meds (14%). About one-third prescribed pain meds saw a pain specialist within 90 days (28% opioid, 37% non-opioid vs 8% nausea). Conclusions: In a large cancer center with streamlined PA processes, PAs for supportive care meds typically resulted in brief delays and high authorization rates. However, the need burden of patients was high. Within 90 days of a PA request, up to 1 in 5 were hospitalized and 1 in 6 had died. Patients with cancer needing supportive care medications represent a vulnerable population for which PA is an additional burden.