Outcomes of patients receiving chemotherapy at the end of life.

Abstract

70 Background: Chemotherapy within 14 days of death indicates poor quality end-of-life (EOL) care. Despite this, approximately 25% of patients receive EOL chemotherapy. This study identifies determinants of EOL chemotherapy. Methods: We identified patients who died in the hospital within 2 weeks of receipt of chemotherapy in 2012 and conducted a detailed chart review to identify demographic and clinical factors, clinical intent, and cause of death. Provider reasoning and assessments were collected. We identified and grouped cause of death into: treatment related, progression of disease, and sudden/unexpected. We used descriptive statistics to document factors associated with receipt of EOL chemotherapy. Results: Of 41 patients, 73% were male; mean age was 63 years; 71% had a hematological malignancy. Whereas solid tumor patients overwhelmingly died of progression of disease (83%), the majority of hematologic patients died of treatment related causes (62%) such as neutropenic sepsis and graft versus host disease. Furthermore, 41% of hematologic malignancy decedents were undergoing induction therapy. Similarly, though all solid tumor chemotherapy was palliative, 33% was for a new diagnosis. All solid tumor patients had a palliative care consultation, whereas 28% of hematologic malignancy patients did. Only 17% of induction chemotherapy patients received a palliative care consult. 76% of hematologic patients were DNR at time of death, while all solid tumor patients were DNR. Site of death varied with 52% of hematologic patients dying in ICU and 10% on the inpatient hospice unit. No solid tumor patients died in the ICU; 67% died on inpatient hospice. Only 10% of all patients had performance status documented. Clinical reasoning for chemotherapy included: disease modification, bridge to clinical trial or transplant, and palliation of symptoms. Conclusions: Patients with hematologic malignancies who received EOL chemotherapy had high ICU utilization and were less likely to have a palliative care consultation. Clinical intent, including early line therapy, may foster unrealistic clinician expectations. An improved understanding of avoidable causes of receipt of chemotherapy at EOL is important in addressing this gap in quality of care.