Inequality in Palliative Radiation Therapy: A Population-based Analysis

Abstract

Purpose/Objective(s)One-half of prescribed radiation therapy (RT) is delivered with palliative intent. The purpose of this study was to define the patterns of delivery of palliative RT across the United States.Materials/MethodsThe Surveillance Epidemiology and End Results (SEER)-Medicare linked database was used to identify patients with metastatic breast, prostate, lung or colorectal cancer diagnosed between 2000 and 2007, and followed through 2009. The receipt of palliative RT was abstracted from SEER and Medicare claims data. Multivariate logistic regression determined the predictors of palliative RT. Multivariate Cox-regression determined predictors of death after palliative RT.ResultsA total of 62,521 patients were identified from the SEER-Medicare database, of which 25,888 (41%) received palliative radiation therapy. Fifty percent of lung cancer patients received palliative RT, followed by breast (42%), prostate (40%), and colorectal cancer (14%). Across all disease sites, multivariate analysis revealed that older patients (p < 0.0001), and those with higher Charlson comorbidity scores (p < 0.0001) were less likely to receive palliative RT. With prostate cancer, black patients were 21% less likely than white patients to receive palliative RT (relative risk [RR] = 0.79, 95% CI = 0.68-0.90, p = 0.0002). With colorectal cancer, black patients were 29% less likely than white patients to receive palliative RT (RR = 0.71, 95% CI = 0.59-0.84, p < 0.0001). No differences were seen between races with breast, or lung cancer patients. Median time from the end of palliative RT through death for the entire population was 98 days. Twenty-seven percent of patients died within 1 month of RT, and 16% died within 14 days of RT. The median time to from RT to death was shortest for lung cancer (77 days), followed by colorectal (6.4 months), prostate (10 months), and breast cancer (1 year). Multivariate analysis revealed that significant predictors (p < 0.05) of death after RT include older age (hazard ratio [HR] = 1.4), lung cancer (HR = 2.5), colorectal cancer (HR = 1.3), prostate cancer (HR = 1.08), and a higher comorbidity score (HR = 1.2).ConclusionsThis study characterizes the delivery of palliative radiation therapy across the United States. Inequality exists among the elderly, patients with comorbid conditions, and black prostate and colorectal cancer patients. Additionally, a significant number of patients die within 1 month of radiation therapy. Understanding these patterns of care, and further research into the underlying causes, will improve access and quality of palliative radiation. Purpose/Objective(s)One-half of prescribed radiation therapy (RT) is delivered with palliative intent. The purpose of this study was to define the patterns of delivery of palliative RT across the United States. One-half of prescribed radiation therapy (RT) is delivered with palliative intent. The purpose of this study was to define the patterns of delivery of palliative RT across the United States. Materials/MethodsThe Surveillance Epidemiology and End Results (SEER)-Medicare linked database was used to identify patients with metastatic breast, prostate, lung or colorectal cancer diagnosed between 2000 and 2007, and followed through 2009. The receipt of palliative RT was abstracted from SEER and Medicare claims data. Multivariate logistic regression determined the predictors of palliative RT. Multivariate Cox-regression determined predictors of death after palliative RT. The Surveillance Epidemiology and End Results (SEER)-Medicare linked database was used to identify patients with metastatic breast, prostate, lung or colorectal cancer diagnosed between 2000 and 2007, and followed through 2009. The receipt of palliative RT was abstracted from SEER and Medicare claims data. Multivariate logistic regression determined the predictors of palliative RT. Multivariate Cox-regression determined predictors of death after palliative RT. ResultsA total of 62,521 patients were identified from the SEER-Medicare database, of which 25,888 (41%) received palliative radiation therapy. Fifty percent of lung cancer patients received palliative RT, followed by breast (42%), prostate (40%), and colorectal cancer (14%). Across all disease sites, multivariate analysis revealed that older patients (p < 0.0001), and those with higher Charlson comorbidity scores (p < 0.0001) were less likely to receive palliative RT. With prostate cancer, black patients were 21% less likely than white patients to receive palliative RT (relative risk [RR] = 0.79, 95% CI = 0.68-0.90, p = 0.0002). With colorectal cancer, black patients were 29% less likely than white patients to receive palliative RT (RR = 0.71, 95% CI = 0.59-0.84, p < 0.0001). No differences were seen between races with breast, or lung cancer patients. Median time from the end of palliative RT through death for the entire population was 98 days. Twenty-seven percent of patients died within 1 month of RT, and 16% died within 14 days of RT. The median time to from RT to death was shortest for lung cancer (77 days), followed by colorectal (6.4 months), prostate (10 months), and breast cancer (1 year). Multivariate analysis revealed that significant predictors (p < 0.05) of death after RT include older age (hazard ratio [HR] = 1.4), lung cancer (HR = 2.5), colorectal cancer (HR = 1.3), prostate cancer (HR = 1.08), and a higher comorbidity score (HR = 1.2). A total of 62,521 patients were identified from the SEER-Medicare database, of which 25,888 (41%) received palliative radiation therapy. Fifty percent of lung cancer patients received palliative RT, followed by breast (42%), prostate (40%), and colorectal cancer (14%). Across all disease sites, multivariate analysis revealed that older patients (p < 0.0001), and those with higher Charlson comorbidity scores (p < 0.0001) were less likely to receive palliative RT. With prostate cancer, black patients were 21% less likely than white patients to receive palliative RT (relative risk [RR] = 0.79, 95% CI = 0.68-0.90, p = 0.0002). With colorectal cancer, black patients were 29% less likely than white patients to receive palliative RT (RR = 0.71, 95% CI = 0.59-0.84, p < 0.0001). No differences were seen between races with breast, or lung cancer patients. Median time from the end of palliative RT through death for the entire population was 98 days. Twenty-seven percent of patients died within 1 month of RT, and 16% died within 14 days of RT. The median time to from RT to death was shortest for lung cancer (77 days), followed by colorectal (6.4 months), prostate (10 months), and breast cancer (1 year). Multivariate analysis revealed that significant predictors (p < 0.05) of death after RT include older age (hazard ratio [HR] = 1.4), lung cancer (HR = 2.5), colorectal cancer (HR = 1.3), prostate cancer (HR = 1.08), and a higher comorbidity score (HR = 1.2). ConclusionsThis study characterizes the delivery of palliative radiation therapy across the United States. Inequality exists among the elderly, patients with comorbid conditions, and black prostate and colorectal cancer patients. Additionally, a significant number of patients die within 1 month of radiation therapy. Understanding these patterns of care, and further research into the underlying causes, will improve access and quality of palliative radiation. This study characterizes the delivery of palliative radiation therapy across the United States. Inequality exists among the elderly, patients with comorbid conditions, and black prostate and colorectal cancer patients. Additionally, a significant number of patients die within 1 month of radiation therapy. Understanding these patterns of care, and further research into the underlying causes, will improve access and quality of palliative radiation.