National trends and predictors of inpatient palliative care utilization in metastatic prostate cancer.

71 Background: Men with metastatic prostate cancer (mPC) experience high symptom burden and often require intensive end-of-life care. Inpatient palliative care (IPC) utilization patterns, including the impact of epidural spinal cord compression (ESCC), are incompletely studied. Methods: We used the National Inpatient Sample database (2016–2022) to identify adult mPC hospitalizations with and without IPC using ICD-10 codes. We compared cohort demographics using t-tests and chi-square tests. We performed multivariate logistic models assessing predictors of IPC and mortality with IPC as the exposure, adjusting for age, race, primary payer, income quartile, hospital characteristics, and Elixhauser Comorbidity Index (ECI). We evaluated ESCC as a primary exposure for IPC use and mortality. p <0.05 was considered significant. Results: We identified 93,018 mPC hospitalizations; IPC occurred in 15,633 (16.8%). mPC admissions rose from 12,050 to 14,471 (APC 2.4%; p =0.019). IPC increased numerically from 1,812 to 2,742, without a significant APC. Versus non-IPC, IPC admissions more often involved age ≥60 (92.4% vs 91.4%; p <0.001), urban hospitals (93.6% vs 91.8%; p <0.001), teaching hospitals (76.4% vs 73.8%; p <0.001), higher comorbidity (ECI≥3: 89.2% vs 86.1%; p <0.001), longer LOS (7.33 vs 5.93 days; p <0.001), and higher unadjusted mortality (24.2% vs 3.3%; p <0.001). In adjusted models, higher IPC use was associated with age ≥60 (aOR 1.187; p <0.001), Black vs White race (aOR 1.09; p <0.001), and Medicaid, private insurance, and self-pay vs Medicare (all p <0.001); lower use occurred in the Midwest and West vs Northeast (aOR 0.87 each; p <0.001), rural vs urban hospitals (aOR 0.83; p <0.001), and non-teaching vs teaching hospitals (aOR 0.89; p <0.001). Larger hospitals had higher use (aOR 1.09; p <0.001). Higher comorbidity was associated with lower IPC (ECI Q3 aOR 0.91; p =0.0015; ECI Q4 aOR 0.92; p =0.009). ESCC independently predicted IPC use (aOR 1.17; p <0.001). In the mortality model, IPC was associated with higher in-hospital death (aOR 9.37; p <0.001). Additional mortality associations included age ≥60 (aOR 1.26; p <0.001), Asian/Pacific Islander race (aOR 1.30; p =0.002), private insurance (aOR 1.34; p <0.001), Midwest region (aOR 1.19; p <0.001), rural location (aOR 1.39; p <0.001), non-teaching hospitals (aOR 1.09; p =0.026) and higher ECI (Q3 aOR 1.24; Q4 aOR 1.55; both <0.001). ESCC was associated with lower inpatient mortality (aOR 0.71; p <0.001). Conclusions: IPC was used in ~1 in 6 mPC admissions, increased over time, and varied by age, race, payer, and hospital factors. ESCC was an independent predictor of IPC use, yet it was associated with lower in-hospital mortality. Geographic and institutional disparities may reflect differences in palliative care availability. The observed IPC use aligns with estimates of need among patients with advanced cancer.