What factors influence hospitalizations among dying cancer patients? An analysis of aggressive end-of-life cancer care

Aggressive end-of-life (EOL) care is usually considered to be of poor quality and little value, but cancer patients make avoidable and distressing hospital visits shortly before death. Cancer patient quality of life has been more extensively studied than the quality of dying. This study aimed to reveal factors that influence aggressive EOL cancer care, with the hope of directing efforts toward eliminating disparities. Linked New York State Cancer Registry (NYSCR), Medicaid, and Statewide Planning and Research Cooperative System (SPARCS) databases were used to evaluate predictors of aggressive EOL care in a cohort of NYS breast and colorectal cancer decedents diagnosed 2004-2006. Logistic regression was used to determine risk factors for multiple hospitalizations, intensive care unit (ICU) admissions, and emergency room (ER) visits in the final month of life. Being ≥ 85-years-old was negatively associated with hospitalizations (OR 0.37, 95% CI 0.29-0.46), ER visits (OR 0.40, 95% CI 0.29-0.56), and ICU admissions (OR 0.55, 95% CI 0.44-0.70). Private insurance was positively associated with hospitalizations (OR 2.22, 95% CI 1.94-2.54). African American women had an increased risk of being admitted to the ICU (OR 1.39, 95% CI 1.19-1.63). There are clear subgroups at an increased risk of receiving poor quality EOL cancer care. This was NYS's first study examining the quality of dying for cancer patients using linked databases. Future research may uncover mechanisms underlying the disparities highlighted by these results. Using previously linked databases, I performed all subsequent data preparation, predictive modeling, and analyses of results using SAS software.