Chemotherapy-Related Hospitalizations Common in Palliative Care

Patients receiving chemotherapy solely for palliative intent have a high risk of chemotherapy-related hospitalization, defeating the goal of the care and increasing health care costs. A study presented by Ankit J. Kansagra, MD, (Abstract 3) found several independent predictors of such hospitalization, including Eastern Cooperative Oncology Group (ECOG) performance status score, cancer site, and other factors. 

“There is an increasing use of palliative chemotherapy,” said Dr. Kansagra, a hematology/oncology fellow at Tufts University School of Medicine. If risks of toxicity of that chemotherapy are underappreciated or misunderstood, the goals of palliative care are undermined.

Kansagra and colleagues conducted a nested case-control study of adult patients with cancer who received chemotherapy between January 2003 and December 2011. Of the 6,850 patients at Northshore Medical Center who received chemotherapy, 2,559 (37.3%) underwent the therapy with palliative intent. Of that cohort, 230 patients (9%) needed a chemotherapy-related hospitalization; 199 of those cases were included in the analysis. The study also included two matched controls for each case, yielding 398 patients in the control group who underwent the same lines of chemotherapy treatment but did not require hospitalization; three cases and 27 patients in the control group were excluded from the analysis due to insufficient data. 

The most common cancers among those hospitalized were gastrointestinal malignancies (36% of patients hospitalized) and lung (32%), followed by breast cancer (14%). The type of chemotherapy was also widely distributed, with platinum-based regimens the most commonly used (50%) followed by taxanes (27%), fluorouracil (25%), and camptothecins (19%).

Virtual Meeting Clip


Of the hospitalizations, 76.5% occurred during the first three cycles of chemotherapy, and the mean length of stay at the hospital was 5 days. Multivariable regression showed that several factors were significant predictors of hospitalization, including ECOG score (p = 0.07), Charlson comorbidity index (p = 0.02), abnormal creatinine levels (p = 0.001), treatment with camptothecins (p = 0.003), low albumin (p = 0.007), and the use of multiple agents (p = 0.03). 

“Patients receiving palliative chemotherapy are at a substantial risk for chemotherapy-related hospitalization,” Dr. Kansagra concluded. “Identified predictors of severe toxicity may help caregivers and their patients make informed decisions about treatment options.”

Robert Siegel, MD, of Hartford Hospital, was the Discussant for the session and noted that, “The administration of palliative chemotherapy cannot and should not happen in a void.” There is evidence that palliative chemotherapy often does actually extend patients’ lives rather than only provide relief of symptoms, but there is often patient misunderstanding that the goal of the treatment is to actually cure the disease.

“Clearly, some of the misconceptions at the time of diagnosis lead to determination of what kind of chemotherapy to get, and when,” Dr. Siegel said. He added that identifying the issue of chemotherapy-related hospitalizations is important, but how to change the problem remains murky. “We need to move from quality measurement… to vigorous efforts at quality improvement. Defining a problem and solving it are two different things.”