ABSTRACT

OBJECTIVE

Investigate factors associated with death at home among older adults who died of cancer in a large city.

METHODS

This is a descriptive study, including all cancer deaths (ICD C00-C97) occurring between 2006 and 2012, among residents of the city of São Paulo, 60 years of age or older. The data source was the Mortality Information System, and the proportion of deaths was calculated according to place of occurrence, gender, age, race/skin color, education, marital status, cancer type, hospital bed availability, and year of death. The chi-squared test was used to examine the associations between the place of death and sociodemographic and clinical variables. Logistic regression was used to identify factors associated with home death. Crude and adjusted odds ratios and the corresponding 95% confidence intervals were estimated.

RESULTS

Most of the deaths occurred in hospitals (88.2%). There was a significant association between the place of death and the following variables: gender, race/skin color, education, age, marital status, cancer type, hospital bed availability, and year of death. In the multivariate analysis, all variables, except the availability of hospital beds, remained as independent predictors of death at home.

CONCLUSIONS

There was a predominance of hospital deaths, with an increase in frequency in the period. Female gender, higher education, married or widowed status, and black race were associated with a decreased risk of death at home, while increasing age, Asian race, and solid neoplasms were associated with higher risk of dying at home.

Aged; Neoplasms, mortality; Hospital Mortality Terminal Care; Palliative Care; Home Nursing; Length of Stay

INTRODUCTION

Population aging is a global phenomenon, occurring not only in developed countries but also in developing countries. One of its consequences is the increase in the occurrence of chronic noncommunicable diseases, particularly cardiovascular diseases, cancer, and diabetes, which will account for more than 75% of all deaths worldwide¹1. Kalache A, Veras RP, Ramos L. O envelhecimento da população mundial: um desaffio novo. Rev Saude Publica. 1987;21(3):200-10. https://doi.org/10.1590/S0034-89101987000300005
https://doi.org/10.1590/S0034-8910198700... .

In Brazil, there was a rapid demographic transition, and in the city of São Paulo, between 1980 and 2012, the proportion of older adults in the population practically doubled, from 6.3% to 11.9%²2. Ministério da Saúde (BR), DATASUS. Informações de saúde (TABNET): demográficas e socioeconômicas. Brasília (DF); 2015 [cited 2017 Feb 2]. Available from: http://www2.datasus.gov.br/DATASUS/index.php?area=0206
http://www2.datasus.gov.br/DATASUS/index... .

In Brazil, for the year 2016, an estimated 600,000 new cases of cancer were estimated for both genders³3. Ministério da Saúde (BR), Instituto Nacional de Câncer José Alencar Gomes da Silva. Estimativa 2016: incidência de câncer no Brasil. Rio de Janeiro: INCA; 2015 [cited 2017 Feb 2]. Available from: http://www.inca.gov.br/estimativa/2016/estimativa-2016-v11.pdf
http://www.inca.gov.br/estimativa/2016/e... _, and 196,954 deaths from the disease were recorded in 2013⁴4. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde. Sistema de Informação sobre Mortalidade – SIM. Brasília (DF): DATASUS; 2015 [cited 2017 Feb 2]. Available from: http://sim.saude.gov.br/default.asp
http://sim.saude.gov.br/default.asp... . While about 59% of new cases registered in Brazil affect individuals over 60 years of age⁵5. Ministério da Saúde (BR), Instituto Nacional de Câncer José Alencar Gomes da Silva. Registros de câncer de base populacional. Rio de Janeiro: INCA; 2017 [cited 2017 Feb 2]. Available from: http://www2.inca.gov.br/wps/wcm/connect/estatisticas/site/home/rcbp/
http://www2.inca.gov.br/wps/wcm/connect/... , more than two-thirds of deaths are recorded in older adults⁴4. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde. Sistema de Informação sobre Mortalidade – SIM. Brasília (DF): DATASUS; 2015 [cited 2017 Feb 2]. Available from: http://sim.saude.gov.br/default.asp
http://sim.saude.gov.br/default.asp... . The city of São Paulo has high of cancer incidence rates (293.2 and 208.3/100,000 inhabitants, adjusted rates for the world population, all neoplasms except skin)⁵5. Ministério da Saúde (BR), Instituto Nacional de Câncer José Alencar Gomes da Silva. Registros de câncer de base populacional. Rio de Janeiro: INCA; 2017 [cited 2017 Feb 2]. Available from: http://www2.inca.gov.br/wps/wcm/connect/estatisticas/site/home/rcbp/
http://www2.inca.gov.br/wps/wcm/connect/... and neoplasms were the second cause of death among older adults residents in the city (20.7%), surpassed only by diseases of the circulatory system⁴4. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde. Sistema de Informação sobre Mortalidade – SIM. Brasília (DF): DATASUS; 2015 [cited 2017 Feb 2]. Available from: http://sim.saude.gov.br/default.asp
http://sim.saude.gov.br/default.asp... .

Due to the increase in cancer incidence and mortality in this age group, many studies have addressed the issue of the place of death for all age groups^6–15. However, the literature on this subject is still quite controversial concerning older adults. While some studies indicate a higher frequency of home deaths for this population group⁸8. Beccaro M, Costantini M, Giorgi Rossi P, Miccinesi G, Grimaldi M, Bruzzi P; ISDOC Research Group. Actual and preferred place of death of cancer patients. Results from the Italian survey of the dying of cancer (ISDOC). J Epidemiol Community Health. 2006;60(5):412-6. https://doi.org/10.1136/jech.2005.043646
https://doi.org/10.1136/jech.2005.043646... ^,99. Cárdenas-Turanzas M, Carrillo M T, Tovalin-Ahumada H, Elting L. Factors associated with place of death of cancer patients in the Mexico City Metropolitan area. Support Care Cancer. 2007;15(3):243-9. https://doi.org/10.1007/s00520-006-0152-4
https://doi.org/10.1007/s00520-006-0152-... , others indicate a higher frequency of hospital deaths⁶6. Bruera E, Sweeney C, Russell N, Willey JS., Palmer JL. Place of death of Houston area residents with cancer over a two-year period. J Pain Symptom Manage. 2003;26(1):637-43. https://doi.org/10.1016/S0885-3924(03)00204-5
https://doi.org/10.1016/S0885-3924(03)00... ^,77. Burge F, Lawson B, Johnston G. Trends in the place of death of cancer patients, 1992-1997. CMAJ. 2003;168(3):265-70.^,1414. Gao W, Ho YK, Verne J, Glickman M, Higginson IJ. Changing patterns in place of cancer death in England: a population-based study. PLoS Med. 2013;10(3):e1001410. https://doi.org/10.1371/journal.pmed.1001410
https://doi.org/10.1371/journal.pmed.100... .

The city of São Paulo is one of the largest cities in the world, and the older population almost doubled in the period from 1980 to 2012²2. Ministério da Saúde (BR), DATASUS. Informações de saúde (TABNET): demográficas e socioeconômicas. Brasília (DF); 2015 [cited 2017 Feb 2]. Available from: http://www2.datasus.gov.br/DATASUS/index.php?area=0206
http://www2.datasus.gov.br/DATASUS/index... . With the technological advance, medicalization costs have been increasing, and many resources have been spent on patients at the end of their lives¹⁶16. Khan SA, Gomes B, Higginson IJ. End-of-life care: what do cancer patients want? Nature Rev Clin Oncol. 2014;11(2):100-8. https://doi.org/10.1038/nrclinonc.2013.217
https://doi.org/10.1038/nrclinonc.2013.2... . The identification of factors associated with place of death in the older population with cancer in the largest city in the country, as well as the monitoring of changes in these patterns, can help formulate public policies and allocate resources related to palliative care¹¹11. Cohen J, Pivodic L, Miccinesi G, Onwuteaka-Philipsen BD, Naylor WA, Wilson DM, et al. International study of the place of death of people with cancer: a population-level comparison of 14 countries across 4 continents using death certificate data. Br J Cancer. 2015:113(9):1397-404. https://doi.org/10.1038/bjc.2015.312
https://doi.org/10.1038/bjc.2015.312... .

METHODS

This is a descriptive study using all cancer deaths (as the underlying cause, ICD-10 C00–C97) occurred in individuals aged ≥ 60 years old, living in the city of São Paulo, from 2006 to 2012. The data source is the Mortality Information System (MIS). Mortality data in the city of São Paulo are considered to be of good quality. During the study period (2006–2012) the percentage of unknown values for the variables gender (0.01%), age (0.3%) and race/skin color (4.1%) was small, as well as the percentage of deaths due to ill-defined causes (only 1.5% of all deaths)⁴4. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde. Sistema de Informação sobre Mortalidade – SIM. Brasília (DF): DATASUS; 2015 [cited 2017 Feb 2]. Available from: http://sim.saude.gov.br/default.asp
http://sim.saude.gov.br/default.asp... .

The following data were retrieved from the MIS databases: place of death (hospital, home, other health facilities, public road, others, unknown); year of death; administrative district of residence; ICD-10 (hematological malignancies = ICD-10 codes (Leukemias and Lymphomas = C81 to C97) and solid tumors = ICD-10 codes (C00 to C79), age group, categorized as: 60-69 years, 70-79 years, 80-89 years, 90 years or more; gender (male, female); race/skin color (white, black, yellow/Asian, mixed, indigenous, unknown); education (years of study completed) (none, 1 to 3, 4 to 7, 8 to 11, 12 and over, unknown); marital status (single, married, widower, separated or divorced, unknown); and number of beds/1,000 inhabitants per district of residence (for the year 2012) (the cut-offs for this variable were established based on the quartiles)¹⁷17. Rede Social Brasileira por Cidades Justas e Sustentáveis. Leitos Hospitalares - São Paulo (SP). São Paulo: Observatório Cidadão Rede Nossa São Paulo; 2016 [cited 2017 Feb 2]. Available from: http://www.redesocialdecidades.org.br/br/SP/sao-paulo/leitos-hospitalares
http://www.redesocialdecidades.org.br/br... .

Bivariate analyses were restricted to deaths in hospitals and households (n = 61,573, corresponding to 97.2% of all cancer deaths among older people in the period) and in all analyses the unknown values were excluded (for each variable).

The association between the independent variables and the place of death was evaluated in the bivariate analysis by the chi-squared test. Crude odds ratios and corresponding 95% confidence intervals were estimated. Multiple logistic regression was used to identify the independent factors associated with death at home and calculation of adjusted odds ratios. In the multiple model, we have included all variables that had a p < 0.25 in the bivariate analysis, but only those with p < 0.05 remained in the final model. The goodness of fit of the final model was assessed using the Hosmer and Lemeshow test.

We have calculated the Annual Percent Change (APC) in the proportion of home deaths in the period 2006–2012 using the Joinpoint method, having the calendar year as the regressor variable.

Statistical analyses were performed with the software Stata for Mac version 13.0 and Joinpoint Regression Program version 3.3. For all statistical tests, an alpha error = 5% was established.

RESULTS

In the period of 2006–2012, 63,343 cancer deaths were recorded among older adults living in the city of São Paulo. The age of the individuals ranged from 60 to 104 years (median = 74 years) and there was a predominance of males (51.3%), white (75.7%), and married individuals (47.4%), with one to three complete years of education (25.1%) (Table 1). Most of the deaths occurred in hospitals (88.2%).

Lung cancer accounted for 14% of cancer deaths among older adults. Among men, the highest frequencies were observed for deaths from lung (16.2%), prostate (15.1%), colon or rectum (10.9%), stomach (9.8%) and pancreatic cancer (5.0%). In females, there was a predominance of deaths due to breast cancer (15.2%), followed by malignant neoplasms of the colon or rectum (13.1%), and lung (10.9%) (Table 2).

There was a significant association between gender and place of death (p < 0.001), with a lower frequency of home deaths among (8.8%) (OR = 0.89, 95%CI 0.85–0.95). There was also a significant association between the marital status and the place of death, with singles (10.0%) and widowers (10.0%) as the categories with higher frequencies of home deaths, compared to married or separated/divorced (p < 0.001). Education was also associated with the place of death, and older people with 12 years or more of education presented the higher percentage of home deaths (9.2%). We have also observed that, with aging, there was an increase in the number of older persons with cancer dying at home (16.1% for the group aged 90 years or older) (p < 0.001). There was also a significant association between race/skin color and the place of death, with Asians (13.8%) presenting a higher risk of dying at home (OR = 1.55, 95%CI 1.36–1.78, compared to whites), while those Blacks/mixed had a lower frequency of death at home (8.0%), compared to whites (9.8%) (OR = 0.84, 95%CI 0.78–0.91). In addition, there was a significant association between the type of neoplasm and the place of death, with a higher frequency of home death among those with solid tumors (9.5%) compared to those with hematological malignancies (4.8%) (OR = 2.09, 95%CI 1.83–2.41). Finally, we observed a statistically significant association between the availability of hospital beds in the district of residence and the place of death. The extreme categories, that is, with the lowest (zero bed/1,000 inhabitants) and the highest availability of beds (≥ 3.82 beds/1,000 inhabitants) were the ones that presented the highest percentages of home deaths (9.7% and 9.6%, respectively) (Table 3).

In the multivariate analysis, except for the availability of hospital beds, all other variables (gender, age, marital status, race/skin color, education, year of death and type of neoplasm) remained as predictors of death at home (Table 4).

It was also observed that, in relation to the year of death, the deaths of older adults with cancer at home decreased significantly over time, from 10.3% in 2006 to 8.1% in 2012 (APC = -4.38%, p < 0.001) (Figure 1).

DISCUSSION

This population-based study showed that, among older adults with cancer living in the city of São Paulo, most of the deaths occurred in hospitals and that only 9% of deaths occurred at home. This is the first study that analyzes the place of death among individuals with cancer, focusing on the older population residing in a large Brazilian metropolis.

Our results corroborate findings from other studies. A survey conducted in England, which included all cancer deaths recorded from 1993 to 2010, also reported a predominance of hospital deaths among adults over 24 years of age (48%)¹⁴14. Gao W, Ho YK, Verne J, Glickman M, Higginson IJ. Changing patterns in place of cancer death in England: a population-based study. PLoS Med. 2013;10(3):e1001410. https://doi.org/10.1371/journal.pmed.1001410
https://doi.org/10.1371/journal.pmed.100... . However, our results are not concordant with findings from studies conducted in Mexico City and Italy, where there was a predominance of home deaths (54% and 57.9%, respectively)⁸8. Beccaro M, Costantini M, Giorgi Rossi P, Miccinesi G, Grimaldi M, Bruzzi P; ISDOC Research Group. Actual and preferred place of death of cancer patients. Results from the Italian survey of the dying of cancer (ISDOC). J Epidemiol Community Health. 2006;60(5):412-6. https://doi.org/10.1136/jech.2005.043646
https://doi.org/10.1136/jech.2005.043646... ^,99. Cárdenas-Turanzas M, Carrillo M T, Tovalin-Ahumada H, Elting L. Factors associated with place of death of cancer patients in the Mexico City Metropolitan area. Support Care Cancer. 2007;15(3):243-9. https://doi.org/10.1007/s00520-006-0152-4
https://doi.org/10.1007/s00520-006-0152-... . Although except for the Mexican and Italian studies, a predominance of hospital deaths was observed, the magnitude of the difference between the proportion of hospital and home deaths in our study was significantly higher. However, it is worth noting that none of the previously mentioned studies included only the older population.

The results of our study showed that, in the city of São Paulo, older women who died from cancer had an 11% lower chance of dying at home compared to men. These findings are similar to those of a European study which reported that women living in the Netherlands, Norway, England, and Wales were less likely to die at home compared to men¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... . However, our findings are different from those observed in Canadian⁷7. Burge F, Lawson B, Johnston G. Trends in the place of death of cancer patients, 1992-1997. CMAJ. 2003;168(3):265-70. and UK¹⁴14. Gao W, Ho YK, Verne J, Glickman M, Higginson IJ. Changing patterns in place of cancer death in England: a population-based study. PLoS Med. 2013;10(3):e1001410. https://doi.org/10.1371/journal.pmed.1001410
https://doi.org/10.1371/journal.pmed.100... studies, which recorded a higher percentage of home deaths among women. A possible explanation for the lower frequency of home death among women with cancer would be the lack of a husband or partner to care for them at home since the life expectancy is higher for women¹⁵15. Higginson IJ, Astin P, Dolan S. Where do cancer patients die? Ten-year trends in the place of death of cancer patients in England. Palliative Med. 1998;12(5):353-63. https://doi.org/10.1191/026921698672530176
https://doi.org/10.1191/0269216986725301... . The presence of a gender issue is also considered since studies in the general population show a higher probability of death at home when the primary caregiver is a woman; besides, women are more often placed in the position of caregivers than in the position of recipients of care. However, it is essential to consider that it is not only a matter related to gender but is also necessary to consider the cultural and family context of these women¹⁸18. Grande GE, Addington-Hall JM, Todd CJ. Place of death and access to home care services: are certain patient groups at a disadvantage? Socl Sci Med. 1998;47(5):565-79. https://doi.org/10.1016/S0277-9536(98)00115-4
https://doi.org/10.1016/S0277-9536(98)00... .

This study demonstrated that married or separated/divorced older adults with cancer had a lower chance of dying at home compared to single people. These findings are opposed to the results of a study developed in London, where it was observed that married individuals had a higher chance of dying at home than singles¹⁹19. Gomes B, Calanzani N, Koffman J, Higginson IJ. Is dying in hospital better than home in incurable cancer and what factors influence this? A population-based study. BMC Med. 2015;13:235. https://doi.org/10.1186/s12916-015-0466-5
https://doi.org/10.1186/s12916-015-0466-... . Cohen et al.¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... also reported that married cancer patients living in Belgium, the Netherlands, and Norway had a higher chance of death at home and the same was observed for widowers residing in the Netherlands and Norway. However, it is important to note that, in that study, the authors used separated or divorced individuals as the reference category¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... . Apparently, marital status is a mediator of gender differences in the care for terminally ill patients. Differences in the place of death can be explained by the discrepancies in the availability of informal caregivers, and the literature has shown that women more often play this role²⁰20. Wachterman MW, Sommers BD. The impact of gender and marital status on end-of-life care: evidence from the National Mortality Follow-Back Survey. J Palliative Med. 2006;9(2):343-52. https://doi.org/10.1089/jpm.2006.9.343
https://doi.org/10.1089/jpm.2006.9.343... .

Our results showed a significant association between death and schooling, with a higher frequency of home deaths among older adults with higher education. Therefore, our data are similar to those described in Belgium, Italy, and Norway, where individuals with cancer and higher education had up to a 98% higher chance of dying at home, compared to individuals with a level of education less or equal to elementary school¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... . In the international study that evaluated the place of death for cancer patients residing in 14 countries and four continents, a higher level of schooling was associated with a higher chance of home death in the European countries (Italy, Spain, Belgium, and the Czech Republic), while the opposite was observed in Mexico, the United States, and South Korea¹¹11. Cohen J, Pivodic L, Miccinesi G, Onwuteaka-Philipsen BD, Naylor WA, Wilson DM, et al. International study of the place of death of people with cancer: a population-level comparison of 14 countries across 4 continents using death certificate data. Br J Cancer. 2015:113(9):1397-404. https://doi.org/10.1038/bjc.2015.312
https://doi.org/10.1038/bjc.2015.312... . We believe that the observation of the increase in the frequency of home deaths concomitant with the increase in education can be seen as an indicator of the possibility (or not) of payment for private home care.

In our study, there was an increase in home deaths concomitant with increasing age. These findings are similar to those described in a Mexican study, where older adults with cancer died more frequently in the households⁹9. Cárdenas-Turanzas M, Carrillo M T, Tovalin-Ahumada H, Elting L. Factors associated with place of death of cancer patients in the Mexico City Metropolitan area. Support Care Cancer. 2007;15(3):243-9. https://doi.org/10.1007/s00520-006-0152-4
https://doi.org/10.1007/s00520-006-0152-... . In contrast to these findings, a European study reported an increase in the frequency of home deaths, concomitantly with the increase in age only in Italy¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... , attributing this fact to the insufficient number of long-term care institutions for older population in that country, as this is still a responsibility of families²¹21. Ribbe MW, Ljunggren G, Steel K, Topinková E, Hawes C, Ikegami N, et al. Nursing homes in 10 nations: a comparison between countries and settings. Age Ageing. 1997;26 Suppl 2:3-12. https://doi.org/10.1093/ageing/26.1.3
https://doi.org/10.1093/ageing/26.1.3... .

In Brazil, the legislation establishes that the care of dependent members should be a responsibility of the families. According to the National Sanitary Surveillance Agency (ANVISA), long-term care institutions for older people are public or private institutions, designed to serve as a collective residence for people aged 60 years and above, with or without family support, in a condition of freedom, dignity, and citizenship. Despite this official definition, there is still no consensus on what a long-term institution for older population is, and since its origin is linked to asylums (institutions initially created by Christian charity for the underprivileged population), there is some prejudice against this model of care²²22. Camarano AA, Kanso S. As instituições de longa permanência para idosos no Brasil. Rev Bras Estud Popul. 2010;27(1):233-5. https://doi.org/10.1590/S0102-30982010000100014
https://doi.org/10.1590/S0102-3098201000... . The relationship between age and home deaths found in our study could be explained by this context.

Our study revealed that older persons with Asian ancestry had a higher frequency of home deaths compared to the other races (white, black/mixed and indigenous). Ethnic differences in the place of death among older people with cancer have been described in several studies. In the United States, according to the results of a study conducted in the Houston area, the risk of hospital death was higher among black individuals (OR = 1.51, 95%CI 1.37–1.66) compared to whites, Hispanics, and Asians⁶6. Bruera E, Sweeney C, Russell N, Willey JS., Palmer JL. Place of death of Houston area residents with cancer over a two-year period. J Pain Symptom Manage. 2003;26(1):637-43. https://doi.org/10.1016/S0885-3924(03)00204-5
https://doi.org/10.1016/S0885-3924(03)00... . In Japan, a national survey data published in 2000 revealed that approximately two-thirds of Japanese when diagnosed with a terminal illness, prefer to die in their homes. In addition, most Japanese identify themselves as Buddhists and Shintoists, whose dogmas claim that death is part of a natural process and that the person who dies will remain present in the family as a spirit²³23. Long SO. Cultural scripts for a good death in Japan and the United States: similarities and differences. Soc Sci Med. 2004;58(5):913-28. https://doi.org/10.1016/j.socscimed.2003.10.037
https://doi.org/10.1016/j.socscimed.2003... ^,2424. Davis AJ, Konishi E, Mitoh T. The telling and knowing of dying: philosophical bases for hospice care in Japan. Int Nurs Rev. 2002;49(4):226-33. https://doi.org/10.1046/j.1466-7657.2002.00126.x
https://doi.org/10.1046/j.1466-7657.2002... . In the last demographic census (2010), 2.2% of the inhabitants of the city of São Paulo declared themselves Asians, and among older adults, this percentage increases to 4.4%. The city of São Paulo has about one million Japanese and their descendants, and our results may be reflecting the cultural patterns of this population²⁵25. São Paulo (Estado). 1908-2008: 100 anos de imigração Japão-Brasil. São Paulo; 2008 [cited 2017 Feb 2]. Available from: http://www.saopaulo.sp.gov.br/imigracaojaponesa/historia.php
http://www.saopaulo.sp.gov.br/imigracaoj... .

Our findings corroborate the literature regarding the type of neoplasm. There is almost a consensus in the literature regarding the higher frequency of home deaths among those with solid tumors compared to those with hematological malignancies. It is known that, among individuals with hematological malignancies, hospital deaths are much more frequent because of the risk of intercurrences (bleeding, thromboembolic events, febrile neutropenia) and consequent need for antibiotics, more aggressive interventions and hospitalization after performing clinical procedures⁶6. Bruera E, Sweeney C, Russell N, Willey JS., Palmer JL. Place of death of Houston area residents with cancer over a two-year period. J Pain Symptom Manage. 2003;26(1):637-43. https://doi.org/10.1016/S0885-3924(03)00204-5
https://doi.org/10.1016/S0885-3924(03)00... .

Our results concerning the lack of association between the place of death and the availability of hospital deaths are not in agreement with the results of a Mexican study; in that study, a significant association between the density of public hospital beds and the place of death was recorded and also a higher chance of hospital death for those individuals living in places with higher bed availability⁹9. Cárdenas-Turanzas M, Carrillo M T, Tovalin-Ahumada H, Elting L. Factors associated with place of death of cancer patients in the Mexico City Metropolitan area. Support Care Cancer. 2007;15(3):243-9. https://doi.org/10.1007/s00520-006-0152-4
https://doi.org/10.1007/s00520-006-0152-... . On the other hand, in the study that analyzed the predictive factors of home death in Europe, there was no influence of the availability of hospital beds in this outcome in three countries (the Netherlands, Norway, and England)¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... . The authors pointed out that the relationship between the availability of hospital beds and place of death may not be so linear and possibly confounded by other factors, such as the availability of home care, cultural and religious issues, and even the communication between relatives¹⁰10. Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol. 2010;28(13):2267-73. https://doi.org/10.1200/JCO.2009.23.2850
https://doi.org/10.1200/JCO.2009.23.2850... .

We observed that the frequency of home deaths among older adults with cancer decreased significantly over time, from 10.3% in 2006 to 8.1% in 2012. This result differs from most reports in the literature, which describes the reduction of hospital deaths over time⁷7. Burge F, Lawson B, Johnston G. Trends in the place of death of cancer patients, 1992-1997. CMAJ. 2003;168(3):265-70.^,1414. Gao W, Ho YK, Verne J, Glickman M, Higginson IJ. Changing patterns in place of cancer death in England: a population-based study. PLoS Med. 2013;10(3):e1001410. https://doi.org/10.1371/journal.pmed.1001410
https://doi.org/10.1371/journal.pmed.100... . Although it has been shown that the choice of cancer patients is to die at home²⁶26. Higginson IJ, Sen-Gupta GJ. Place of care in advanced cancer: a qualitative systematic literature review of patient preferences. J Palliative Med. 2000;3(3):287-300. https://doi.org/10.1089/jpm.2000.3.287
https://doi.org/10.1089/jpm.2000.3.287... , preferences are not always respected and may reflect much more the practices and characteristics of the local health system. According to the World Health Organization, the provision of palliative care in Brazil is still isolated and disproportionate to the size of the population, and there is also a limitation in the availability of morphine²⁷27. Connor SR, Bermedo MCS, editors. Global atlas of palliative care at the end of life. Geneva: World Health Organization; WPCA; 2014 [cited 2017 Feb 2]. Available from: http://www.who.int/nmh/Global_Atlas_of_Palliative_Care.pdf
http://www.who.int/nmh/Global_Atlas_of_P... . It should be noted that the situation observed reflects the profile of the largest and most developed city in the country. In Brazil, in the same period, the percentage of home deaths for older adults with cancer ranged from 10.5% (Federal District) to 52.0% (Piauí)⁴4. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde. Sistema de Informação sobre Mortalidade – SIM. Brasília (DF): DATASUS; 2015 [cited 2017 Feb 2]. Available from: http://sim.saude.gov.br/default.asp
http://sim.saude.gov.br/default.asp... . This large variation raises the hypothesis that the factors associated with home death in this group may be entirely different in distinct regions of the country.

The possible bias arising from unknown data in the death certificate and underreporting of cancer deaths can be listed as limitations of our study. However, mortality data in the city of São Paulo has reliable quality, and we do not believe that the results of the study would be substantially distorted.

Our results bring to light the need for a closer look at the quality of end-of-life care for people with cancer. We hope that the information generated in this study can be used to support public policies aimed at improving the offer of palliative care for older persons with cancer, benefiting both patients and their families with quality of life.

REFERENCES

Publication Dates

History