Introduction
Bladder cancer primarily affects older adults. End-of-life care planning is therefore central to comprehensive oncologic management. Place of death reflects access to hospice services, caregiver capacity, regional health system structure, and patient preference (Urological et al. Reference Urological, Barker and Soylu2024). Prior studies report low palliative care utilization among patients with advanced malignancy (Hugar et al. Reference Hugar, Lopa and Yabes2019). National longitudinal data describing trends in place of death among patients with bladder cancer remain limited. This study aimed to characterize national patterns in place of death among patients who died from bladder cancer between 2000 and 2020, and to examine variation across key demographic and geographic factors.
Methods
We conducted a retrospective population level study using the CDC WONDER database (Centers for Disease Control and Prevention 2026). We identified all deaths in the United States between 2000 and 2020 in which bladder cancer was listed as the underlying cause. Place of death was used as a proxy for preferred place of death, as direct data on patient preferences were not available in the dataset. A total of 293,906 deaths met inclusion criteria (see Figure 1). We categorized deaths by place of death, age group, race, census region, and year of death. Place of death categories included home, medical facility, nursing home, and hospice facility. Age groups were 35–44, 45–54, 55–64, 65–74, 75–84, and 85 years or older. Race categories included White, Black, and Asian or Pacific Islander. We performed descriptive analyses to summarize distributions across demographic strata. We assessed associations between categorical variables and place of death using chi square testing. We defined statistical significance as a 2-sided p-value < 0.05.
Geographical Distribution of Deaths Attributable to Bladder Cancer in the United States (2000-2020).
We constructed a multivariable binary logistic regression model to evaluate predictors of death at a hospice facility. The dependent variable was hospice facility death versus non-hospice death. Non-hospice deaths included home, medical facility, and nursing home deaths.
Independent variables included age group, race, census region, and year of death. We used age 85 years or older as the reference category for age. We used White race as the reference category for race. We modeled year of death as a continuous variable to estimate annual change in odds of hospice death. We report odds ratios with 95% confidence intervals when available.
Results
Between 2000 and 2020, 293,906 deaths from bladder cancer occurred in the United States. The South region accounted for the highest regional proportion at 35.6%. (Figure 1) Overall distribution of place of death was as follows: home 41.6%, medical facilities 24.4%, nursing homes 19.5%, and hospice facilities 9.0%. Place of death varied significantly by region, age, race, and year, all p < 0.001.
Home deaths increased from 4,281 in 2000 to 8,554 in 2020. Hospice deaths increased substantially over time (59 to 1934 from 2003 to 2020), while medical facility deaths declined from 3836 in 200 to 2558.
Age was strongly associated with place of death. Home deaths were most common among patients aged 75–84 years with 42,644 deaths. Nursing home/long-term care deaths were highest among patients aged 85 years or older with 26,216 deaths. Hospice deaths were most frequent among patients aged 75–84 years (n = 8,754). Age differences were statistically significant, (χ² p < 0.001).
White individuals accounted for 93.6% of deaths. Among White individuals, home was the most common place of death with 114,307 deaths. Black individuals had lower odds of hospice death compared with White individuals (OR: 0.699, p < 0.001). Asian or Pacific Islander individuals showed no significant difference in hospice use (p = 0.984). Race was significantly associated with place of death (p < 0.001).
In multivariable binary logistic regression assessing hospice versus non-hospice death, younger age was associated with higher odds of hospice death compared with age 85 years or older. Patients aged 45–54 years had OR: 1.36, 95% CI: 0.22–18.79, p < 0.001. Odds decreased progressively with advancing age. Black race remained independently associated with lower hospice use, OR: 0.699, p < 0.001. Year of death was associated with increasing hospice use, OR: 1.134 per year, p < 0.001 (see Table 1).
Multivariate logistic regression of predictors of hospice vs non-hospice death
Discussion
Our study demonstrates significant changes in place of death among patients with bladder cancer from 2000 to 2020. Home remained the most common location of death. Hospice use increased steadily over time, although hospice facilities accounted for only 9.0% of deaths.
The 13.4% annual increase in adjusted odds of hospice death indicates progressive growth in hospice utilization. Despite this increase, absolute hospice and palliative care use remained low relative to other settings (Ismaeel et al. Reference Ismaeel, Patil and Alemozaffar2021). Our findings suggest improvement over time but continued opportunity for expansion.
Age strongly influenced hospice utilization. Younger patients had markedly higher adjusted odds of hospice death compared with patients aged 85 years or older. The highest odds were observed in patients aged 35–44 years. Odds decreased progressively with advancing age. Although traditional patterns have suggested greater hospice utilization among older adults, recent shifts indicate increasing use among younger patients (Cagle et al. Reference Cagle, Lee and Ornstein2020). This trend may reflect differences in health literacy, access to information, or family support dynamics by patient age, though further investigation is warranted (Grant et al. Reference Grant, Back and Dettmar2021; Peng et al. Reference Peng, Feng and Cao2025). Similarly, nursing home deaths were concentrated in the oldest age group, indicating that age remains a major determinant of end-of-life setting.
Racial disparities persisted after adjustment. Black individuals had significantly lower odds of hospice death compared with White individuals. This difference remained statistically robust even after adjustment across multiple disease states. A prior study found that Black individuals were less likely to use 3 or less days of hospice compared to White individuals (Ornstein et al. Reference Ornstein, Roth and Huang2020). There may also be structural and system level factors that contribute to these disparities in care utilization by race. While not evaluated in this study, others have found that the type of government insurance coverage may play a role. Specifically, among Medicaid beneficiaries, Hispanic and non-Hispanic Black were less likely to use hospice compared to non-Hispanic White individuals (Robison et al. Reference Robison, Shugrue and Dillon2023). The data from our study demonstrate association and require targeted evaluation to identify modifiable causes.
Overall, medical facility deaths have declined over time while home and hospice deaths increased. In 2020, changes in place of death coincided with the COVID-19 pandemic. The degree to which pandemic related factors, such as policies on infection control or patient avoidance of medical facilities, influenced end-of-life location cannot be determined from place of death data alone (Teasdale et al. Reference Teasdale, Narayan and Harman2024). Future research should instead focus on examining ways to reduce barriers to accessing equitable end-of-life care.
This study has limitations. Death certificate data may misclassify place of death. Nonetheless, death certificate reporting follows standardized national procedures and provides complete population level coverage, which supports valid comparisons across regions and over time. The dataset does not capture patient stated preference, comorbid burden, socioeconomic status, or timing of hospice enrollment. However, the large national cohort of 293,906 deaths over 2 decades allows precise estimation of temporal trends and demographic associations at the population level. Place of death may not accurately reflect patient preferences, as it can be influenced by clinical, social, and health system factors beyond individual choice. Nevertheless, place of death is an established health services metric and offers an objective measure of how end-of-life care is delivered within the health-care system.
Conclusion
From 2000 to 2020, 293,906 deaths from bladder cancer occurred in the United States. Home was the most common place of death. Hospice use increased significantly, with a 13.4% annual rise in adjusted odds. Younger patients had higher odds of hospice death compared with patients aged 85 years or older. Black individuals had significantly lower odds of hospice utilization compared with White individuals. These findings demonstrate evolving patterns in end-of-life care and persistent disparities. Clinicians must ensure timely goals of care discussions and equitable access to hospice services to align care delivery with patient needs and established standards of practice.
Acknowledgments
Author Manas Pustake, MD presented a poster based on this study at the American Society of Clinical Oncology Genitourinary Symposium 2025 in San Francisco, CA, USA.
Funding
No external funding was received for this study.
Competing interests
The authors have no competing interests to declare.
Open access
