Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina.
Division of Medical Oncology and Palliative Care, Duke Cancer Institute, Durham, North Carolina.
Cancer. 2018 Nov 15;124(22):4408-4419. doi: 10.1002/cncr.31737. Epub 2018 Oct 21.
Place of death is an essential component of high quality cancer care and comprehensive national trends and disparities in place of death are unknown.
Deidentified death certificate data were obtained via the National Center for Health Statistics. All cancer deaths from 1999 through 2015 were included. Multivariate logistic regression was used to test for disparities in place of death associated with sociodemographic variables.
From 1999 through 2015, a total of 9,646,498 cancer deaths occurred. Hospital deaths decreased (from 36.6% to 24.6%), whereas the rate of home deaths (38.4% to 42.6%) and hospice facility deaths (0% to 14.0%) both increased (all P<.001). On multivariate logistic regression, all assessed variables were found to be associated with place of death. Specifically, younger age (age birth-14 years: odds ratio [OR], 2.39; age 25-44 years: OR, 1.62), black (OR, 1.83) or Asian (OR, 1.74) race, and Hispanic ethnicity (OR, 1.41) were associated with hospital death. Being married (OR, 2.17) or widowed (OR, 1.56) was associated with home death whereas increasing educational level (OR, 1.15-1.19) was associated with hospice death (all P<.001). Despite overall improvements, certain disparities were found to increase. For young patients, the likelihood of a hospital death increased from 2.3 times to 3.4 times that of older patients (50.9% for those aged 15-24 years vs 15.0% for those aged ≥85 years in 2015). For black patients, the likelihood of a hospital death increased from 1.29 times to 1.42 times that of white patients (32.8% for black patients vs 23.1% for white patients in 2015).
Hospital cancer deaths decreased by approximately one-third with commensurate increases in home and hospice facility deaths. Many sociodemographic groups experience significant disparities with regard to place of death and may benefit from targeted efforts to improve goal-concordant care.
死亡地点是高质量癌症护理的一个重要组成部分，目前尚不清楚全国范围内死亡地点的综合趋势和差异。
通过国家卫生统计中心获得了经身份识别的死亡证明数据。纳入了 1999 年至 2015 年期间所有的癌症死亡病例。采用多变量逻辑回归来检验与社会人口统计学变量相关的死亡地点差异。
1999 年至 2015 年期间，共有 9646498 例癌症死亡病例。医院死亡人数减少（从 36.6%降至 24.6%），而家中死亡人数（从 38.4%增加到 42.6%）和临终关怀机构死亡人数（从 0%增加到 14.0%）均有所增加（均 P<.001）。在多变量逻辑回归中，所有评估变量均与死亡地点相关。具体而言，年龄较小（出生至 14 岁：比值比[OR]，2.39；25-44 岁：OR，1.62）、黑种人（OR，1.83）或亚裔（OR，1.74）、西班牙裔（OR，1.41）与医院死亡相关。已婚（OR，2.17）或丧偶（OR，1.56）与家中死亡相关，而教育程度提高（OR，1.15-1.19）与临终关怀机构死亡相关（均 P<.001）。尽管总体情况有所改善，但某些差异仍有所增加。对于年轻患者，其在医院死亡的可能性从是老年患者的 2.3 倍增加到 3.4 倍（2015 年，15-24 岁年龄组的患者为 50.9%，而≥85 岁年龄组的患者为 15.0%）。对于黑种人患者，其在医院死亡的可能性从是白种人患者的 1.29 倍增加到 1.42 倍（2015 年，黑种人患者为 32.8%，白种人患者为 23.1%）。
随着家中和临终关怀机构死亡人数的相应增加，医院癌症死亡人数减少了约三分之一。许多社会人口统计学群体在死亡地点方面存在显著差异，可能需要有针对性的努力来改善与目标一致的护理。
