Supplementary Material for:
International study of the place of death of people with cancer: a population-level
comparison of fourteen countries across four continents using death certificate
data
J. Cohen1*; L. Pivodic1*; G Miccinesi2; B.D. Onwuteaka-Philipsen3; W.A. Naylor4; D.M.
Wilson5; M. Loucka6; A. Csikos7; K. Pardon1; L. Van den Block1; M. Ruiz-Ramos8; M.
Cardenas-Turanzas9; Y. Rhee10; R. Aubry11; K. Hunt12; J. Teno13; D. Houttekier1**; L.
Deliens1,14**
Methods
About different data sources and the levels of linkage of data
Information about the source of the healthcare resource statistics, the degree of
urbanization and details of the level of linkage are provided in Table S1.
About the logistic regression analyses
The analyses were performed on cancer patients only. This has as the advantage that we
obtained a homogeneous risk population in terms of disease-specific characteristics.
Additionally, cancer patients have a relatively predictable course of dying allowing for
better planning for end-of-life care and for dying in the place of choice which means that
analyses of cancer patients only can provide meaningful insight into the effects of certain
factors on the place of death in different countries.(Cohen et al, 2010)
To construct a parsimonious model and select the relevant variables for each country,
several alternative logistic regression models were constructed and evaluated for
goodness of fit and regression diagnostics. A stepwise variable selection (forward and
backward) was used to obtain different candidate models and guide the variable
selection. After the variable selection only covariates with p<0.05 were allowed to stay
on the final logistic regression model using a backward conditional procedure. As we
aimed for similar regression models across countries, in order to enhance comparability
of the effects between countries, we only selected variables that were available in at least
a number of countries.
Reference
Cohen J, Houttekier D, Onwuteaka-Philipsen B, Miccinesi G, Addington-Hall J, Kaasa S,
Bilsen J, Deliens L (2010) Which patients with cancer die at home? A study of six
European countries using death certificate data. J Clin Oncol 28: 2267–2273,
doi:10.1200/JCO.2009.23.2850.
Table S1: Sources of the death certificate data, the data on healthcare supply, the ecological levels of the healthcare supply data linked to
the individual death certificates, degree of urbanisation and population size per health region
Belgium
Canada
Czech Republic
England
France
Flemish Agency for Care and
Health, Brussels Health and
Social Observatory, French
Community of Belgium, data for
2008
Statistics Canada, data for
2008. Quebec death
certificate data (managed by
the Quebec province) are
not integrated with the
death certificate data of the
rest of the country
(managed by Statistics
Canada).
Institute of Health
Information and Statistics
of the Czech Republic
(UZIS), data for 2008
Office for National
Statistics (ONS), data for
2008
Inserm-CépiDc (Centre
d'épidémiologie sur les
causes médicales de décès,
Institut national de la santé et
de la recherche médicale),
data for 2008
source of data
Federal Public Service Health,
Food Chain Safety And
Environment, data for 2008
Statistics Canada, data for
2008
Institute of Health
Information and Statistics
of the Czech Republic
(UZIS), data for 2008
Department of Health
(form KH03), data for
2008
Ecosanté (institut de
recherche et documentation
en economie de la santé ),
data for 2008
area level of linkage
For Flanders and Brussels: health
care region, large city level (14
regions); for Wallonia: province
(5 regions, no health care regions
exist in Wallonia)
Province (12 regions,
excluding Quebec province)
Country-level
Strategic Health
Authority (10 regions)
Administrative region (22
regions)
source of data
National institute for health and
disability insurance (INAMIRIZIV), data for 2008
Statistics Canada, data for
2008
Institute of Health
Information and Statistics
of the Czech Republic
(UZIS), data for 2008
Care Standards Act data
(care homes active as of
31st March 2008) / CQC
database)
Ecosanté (Institut de
recherche et documentation
en economie de la santé )
area level of linkage
For Flanders and Brussels: Health
Care Region, small city level (60
regions); for Wallonia: borough
(arrondissement) (20 regions, no
health care regions exist in
Wallonia)
Province (12 regions,
excluding Quebec province)
Country-level
Strategic health
Authority (10 regions)
Administrative region (22
regions)
Death certificate data
Hospital beds
LTC beds
Belgium
Canada
Czech Republic
England
France
General practitioners
source of data
National institute for health and
disability insurance (INAMIRIZIV), data for 2008
Statistics Canada, data for
2008
Institute of Health
Information and Statistics
of the Czech Republic
(UZIS), data for 2008
Department of Health,
data for 2008
area level of linkage
For Flanders and Brussels: Health
Care Region, small city level (60
regions); for Wallonia: borough
(arrondissement) (20 regions, no
health care regions exist in
Wallonia)
Statistics Belgium, data for 2008
Province (12 regions,
excluding Quebec province)
Country-level
Strategic health
Authority (10 regions)
Statistics Canada, data for
2008
Czech Statistical Office,
data for 2008
Office for National
Statistics (mid-2008
population estimates)
Statistics Belgium (calculated at
the municipality level)
Statistics Canada
n/a
n/a
Population size per health
region
Degree of urbanisation
Direction de la recherche,
des études, de l'évaluation et
des statistiques (DREES) &
Ministère des affaires
sociales, de la santé et des
droits des femmes (ADELI
database), data for 2008
Administrative region (22
regions)
National Institute of
Statistics and Economic
Studies (INSEE), data for
2008
National Institute of
Statistics and Economic
Studies (INSEE)
Hungary
Italy
South Korea
Mexico
New Zealand
Central Statistical Office
Hungary, data for 2008
Italian National Institute of
Statistics (Unit for Cause of
Death Statistics), data for
2008
Statistics Korea, data for
2008
Secretaria de Salud,
Sistema Nacional de
Informacion en Salud
(SINAIS)*, data for
2008
New Zealand Ministry of
Health, data for 2008
source of data
OECD, data for 2008
Ministry of Health (Italian
Health Care data), data for
2009
OECD, data for 2008
New Zealand Ministry of
Health, data for 2008
area level of linkage
Country-level
Macroregions (5 regions)
Country-level
Secretaria de Salud,
Sistema Nacional de
Informacion en Salud
(SINAIS)*, data for
2008
Municipality (2456
municipalities)
source of data
Hungarian Central Statistical
Office, data for 2008
OECD, data for 2008
n/a
New Zealand Ministry of
Health, data for 2008
area level of linkage
Country-level
Ministry of Health (Italian
Health Care data), data for
2009
Macroregions (5 regions)
Death certificate data
Hospital beds
District Health Board region
(21 regions)
LTC beds
Country-level
District Health Board region
(21 regions)
GPs
source of data
Hungarian Central Statistical
Office, data for 2008
Ministry of Health (Italian
Health Care data), data for
2009
OECD, data for 2008
area level of linkage
Country-level
Macroregions (5 regions)
Country-level
Population size per health
region
Hungarian Central Statistical
Office, data for 2008
National Institute of
Statistics (Istat)
OECD, data for 2008
Degree of urbanisation
/
National Institute of
Statistics (Istat)
Statistics Korea
Secretaria de Salud,
Sistema Nacional de
Informacion en Salud
(SINAIS), data for 2008
(physicians of all
specialties, including
GPs)
Municipality (2456
municipalities)
Instituto Nacional de
Estadistica Geografia
(INEGI)
n/a
New Zealand Ministry of
Health, data for 2008
District Health Board region
(21 regions)
New Zealand Ministry of
Health, data for 2008
n/a
Death certificate data
Spain (Andalusia)
Instituto de Estadística y
Cartografía de Andalucía, data for
2010. Place of death data were
not available on death certificates
for the whole of Spain at the time
of data collection.
The Netherlands
Netherlands: Statistics
Netherlands (CBS), data for
2008
United States
Center for Disease
Control and Prevention
(CDC) based on Data Use
Agreement, data for 2007
Wales
Office for National Statistics
(ONS), data for 2008
Andalusian Health Service
(Servicio Andaluz de Salud), data
for 2010
provinces (8 regions)
Netherlands institute for
health services research,
data for 2008
provinces (12 regions)
Statehealthfacts.org, data
for 2007
Statistics Wales, data for 2008
50 states + District of
Columbia
Primary Care Organizations (7
regions)
Andalusian Health Service
(Servicio Andaluz de Salud), data
for 2010
health districts (33 regions)
Netherlands institute for
health services research,
data for 2008
provinces (12 regions)
Statehealthfacts.org, data
for 2007
Care Standards Act data (care
homes active as at 31st March
2008) / CQC database
Country-level (no regional data
available)
Andalusian Health Service
(Servicio Andaluz de Salud), data
for 2010
health districts (33 regions)
Netherlands institute for
health services research,
data for 2008
provinces (12 regions)
Statistics Wales, data for 2008
Instituto Nacional de Estadisticas
(Spanish Statistical Office), data
for 2010
Instituto Nacional de Estadisticas
(Spanish Statistical Office)
Netherlands: Statistics
Netherlands (CBS), data for
2008
Statistics Netherlands
(CBS)
Statehealthfacts.org, data
for 2007 (primary care
physicians)
50 states + District of
Columbia
United States Census
Bureau, data for 2007
n/a
n/a
Hospital beds
source of data
area level of linkage
LTC beds
source of data
area level of linkage
50 states + District of
Columbia
GPs
source of data
area level of linkage
Population size per health
region
Degree of urbanisation
Primary Care Organizations (7
regions)
Statistics Wales, data for 2008
Table S2: Marital status of the persons dying from cancer in 2008a in 14 countries
Country
Abbreviations
Continental Western Europe
France
FR
Italy
IT
Spain (Andalusia)a
ES
Belgium
BE
Netherlands
NL
Central/Eastern Europe
Czech Republic
CZ
Hungary
HU
UK
England
ENG
Wales
WAL
New Zealand
NZ
North America
Canadab
CA
United Statesa
US
Latin America
Mexico
MX
East Asia
South Korea
KR
a
Marital status
Married
Widowed
Divorced/
Separated
Unmarried
54.3
59.6
61.2
53.9
56.6
25.0
28.8
24.9
28.8
/b
9.4
1.9
3.7
9.3
b
/
11.3
9.7
10.2
8.1
43.4b
53.1
49.3
28.7
29.3
12.6
14.2
5.5
7.2
51.8
52.6
30.6
30.6
9.4
9.2
8.1
7.6
56.0
50.8
26.9
27.2
9.0
14.1
8.1
7.9
59.8
22.2
2.1
15.9
65.1
25.2
5.3
4.4
Reference year is 2007 for USA and 2010 for Spain;
b
The Dutch data only distinguish between married and not married.
Table S3: healthcare supply in the studied countries
Country
Abbreviations
Continental Western Europe
France
FR
Italy
IT
Spain (Andalusia)
ES
Belgium
BE
Netherlands
NL
Central/Eastern Europe
Czech Republic
CZ
Hungary
HU
UK
England
ENG
Wales
WAL
New Zealand
NZ
North America
Canada
CA
United States
US
Latin America
Mexico
MX
East Asia
South Korea
KR
Total
Healthcare supply
hospital
beds per
10.000c
LTC beds
per 1.000
65+c
GPs per
10.000c
69.5
34.1
26.8
51.9
30.8
59.2
34.6
18.8
70.3
77.5
8.7
7.6
7.4
20.1
5.4
60.4
71.0
28.9
32.7
7.0
6.5
31.1
43.9
30.8
54.8
49.3
71.6
6.6
6.5
8.8
29.8
27.0
58.1
41.5
10.1
13.1
3.3
0.0
6.8
83.0
17.5
7.5
Table S4: Stepwise building up of the hierarchical binary logistic regression models with home vs all other places of death as dependent variable
FR
IT
ES
BE
NL
CZ
ENG
WAL
NZ
CA
US
MX
KR
a
Model 1
(Country)
(Model 1 in
Figure 1)
Model 2
(+cancer site, age,
sex)
(Model 2 in Figure
1)
Model 3
(+ marital
status)
Ref cat
Ref cat
Ref cat
3,52 (3,46-3,58)
1,96 (1,89-2,03)
1,73 (1,68-1,78)
3,65 (3,57-3,74)
0,94 (0,91-0,97)
1,50 (1,47-1,52)
1,50 (1,43-1,58)
1,69 (1,61-1,78)
0,81 (0,79-0,84)
2,72 (2,68-2,75)
5,71 (5,59-5,82)
0,57 (0,56-0,59)
3,52 (3,46-3,58)
1,94 (1,87-2,01)
1,73 (1,68-1,78)
3,64 (3,56-3,73)
0,92 (0,89-0,95)
1,50 (1,47-1,52)
1,50 (1,42-1,57)
1,70 (1,62-1,79)
0,82 (0,80-0,84)
2,77 (2,73-2,81)
5,96 (5,84-6,09)
0,54 (0,53-0,55)
3,44 (3,39-3,50)
1,89 (1,82-1,96)
1,73 (1,68-1,78)
4,04 (3,94-4,14)
0,92 (0,89-0,95)
1,50 (1,47-1,52)
1,49 (1,42-1,57)
/a
0,80 (0,77-0,82)
2,82 (2,78-2,86)
5,86 (5,74-5,98)
0,51 (0,50-0,53)
Model 4
(+ hospital beds)
model 5
(+long term care
beds)
Ref cat
1,53 (1,48-1,57)
0,70 (0,67-0,73)
1,15 (1,11-1,18)
1,65 (1,59-1,71)
0,75 (0,72-0,77)
0,61 (0,59-0,63)
0,81 (0,77-0,85)
/a
0,32 (0,30-0,33)
1,05 (1,02-1,09)
1,27 (1,20-1,33)
0,70 (0,68-0,72)
Ref cat
1,66 (1,61-1,71)
0,66 (0,63-0,69)
1,59 (1,54-1,64)
3,16 (3,04-3,30)
0,57 (0,55-0,59)
0,88 (0,85-0,92)
0,96 (0,91-1,01)
/a
0,52 (0,50-0,55)
1,35 (1,31-1,40)
1,24 (1,18-1,31)
0,36 (0,35-0,38)
Model 6
(+number of
GPs)
Ref cat
1,66 (1,61-1,71)
0,66 (0,63-0,69)
1,55 (1,49-1,61)
3,18 (3,05-3,31)
0,57 (0,55-0,59)
0,89 (0,86-0,92)
0,96 (0,91-1,02)
/a
0,52 (0,50-0,54)
1,33 (1,29-1,38)
1,24 (1,18-1,31)
0,37 (0,35-0,38)
Model 6 bis
(same as 6 but
without marital
status)
(Model 3 in
Figure 1)
Ref cat
1,70 (1,65-1,75)
0,68 (0,65-0,71)
1,60 (1,53-1,66)
2,83 (2,72-2,95)
0,57 (0,55-0,59)
0,88 (0,85-0,91)
0,96 (0,91-1,02)
1,10 (1,04-1,17)
0,54 (0,51-0,56)
1,33 (1,28-1,38)
1,27 (1,20-1,33)
0,39 (0,37-0,40)
: Because marital status is not available in the New Zealand data, this country is excluded from the analysis
cancer site includes 17 categories: head & neck; stomach; colon, rectum and anus; pancreas; other gastrointestinal; trachea, bronchus and lung; other respiratory; breast;
cervix uteri, corpus uteri and ovary; prostate; urinary tract; other genitourinary; central nervous system; Non-Hodgkin’s lymphoma; leukemia; other hematologic
malignancies
b:
Table S5: Contextual data for the countries studied and their correlation with the percentage of home deaths and
hospital deaths
France
Belgium
Spain
(Andalusia)
Italy
Hungary
Englanda
Walesa
Netherlands
New Zealand
Czech
Republic
Canada
United States
South Korea
Mexico
correlation
(r) with
home death
correlation
(r) with
hospital
death
Social
expenditure
as percentage
of GDPb
Health
expenditure
as percentage
of GDPb
Palliative care
services for
adults per
million
inhabitantsc
Opioid
consumption
(mg/person) per
year (morphine
equivalence)d
Percentage
of cancer
deaths that
occurred at
home
Percentage
of cancer
deaths that
occurred in
hospital
29.8
27.3
10.5
9.9
7.6
18.1
179.6
363.5
19.1
28.9
72.8
59.3
26.7
9.4
8.0
298.1
31.6
64.9
25.8
23.1
21.8
21.8
20.9
19.8
8.6
7.3
8.2
8.2
10.2
9.3
8.0
8.5
15.4
15.4
15.3
n.a.
126.7
63.3
225.3
225.3
167.6
198.7
45.3
/e
26.1
26.2
46.3
28.5
47.3
67.7
44.4
56.6
25.8
26.2
18.1
6.7
2.2
67.5
18.2
26.2
17.6
16.3
8.4
7.4
9.5
14.9
6.3
5.8
n.a.
n.a.
n.a
1.1
643.1
663.0
27.6
4.5
16.1
39.0
11.8
57.3
67.6
33.7
87.2
39.9
-0.12
0.10
-0.19
-0.13
/
/
0.10
-0.20
0.14
-0.01
/
/
The reference year is 2008, except for USA (2007) and Spain (2010)
a
Data for United Kingdom as a whole. No data separate for England and Wales were available
b
Source: OECD
c
Source: Atlas de Cuidados Paliativos en Latinoamérica - Edición Cartográfica 2013; EAPC Atlas of Palliative
Care in Europe.
d
Source: Pain & Policy Studies Group, University of Wisconsin-Madison. Includes Fentanyl, Hydromorphone,
Methadone, Morphine, Oxycodone, Pethidine.
e
‘Home’ not recorded as place of death in Hungary.