Hospice palliative care utilization among patients with cancer and non-cancer lung diseases: a retrospective cohort study

Abstract Background Disparities in hospice palliative care utilization rates between terminally ill cancer and non-cancer patients are evident. However, limited research has examined factors associated with hospice palliative care utilization among patients with cancer and non-cancer lung diseases. We aimed to compare hospice palliative care utilization between patients with cancer and those with non-cancer lung diseases and to identify factors associated with hospice palliative care utilization. Methods We conducted a retrospective cohort study of deceased patients with primary lung disease diagnoses at a southern Taiwanese tertiary teaching hospital between January 2019 and December 2020. Demographic and clinical data were extracted from the hospital’s medical information system. Multivariate logistic regression analyses were performed to identify factors associated with hospice palliative care utilization. Results The study included 464 non-cancer and 269 cancer patients, averaging 77.4 ± 13.4 years. While there was no significant difference in the rate of the Do-Not-Resuscitate code (95.08% vs. 91.81%, p = 0.065), the hospice palliative care rate differed significantly (36.89% vs. 55.97%, p < 0.001) between the two groups. Multiple logistic regression analysis revealed increased odds of hospice palliative care utilization associated with cancer diagnosis, higher education level, urban residency, and autonomous Do-Not-Resuscitate decisions. Decreased odds were associated with being male, emergency hospital admission, and Charlson comorbidity index scores ≥ 2. Based on the logistic regression findings, we create a nomogram that serves as a predictive tool for assessing patients’ utilization of hospice palliative care services. Conclusion By analyzing medical records, this study highlights a disparity in utilization between cancer and non-cancer patients. The development of a nomogram prediction model can enable the identification of patients whose hospice palliative care needs might have been inadvertently overlooked at admission. We hope to encourage medical practitioners to be more aware of patients’ needs early to ease physical and emotional distress.