Colombet et al.

BMC Palliative Care (2019) 18:35

https://doi.org/10.1186/s12904-019-0419-4

RESEARCH ARTICLE Open Access

Multicentre analysis of intensity of care at

the end-of-life in patients with advanced
cancer, combining health administrative
data with hospital records: variations in
practice call for routine quality evaluation
Isabelle Colombet1,2* , Carole Bouleuc3, Alain Piolot4, Aurélie Vilfaillot5, Hélène Jaulmes6, Sabine Voisin-Saltiel7,
François Goldwasser8, Pascale Vinant1 and the EFIQUAVIE study group

Abstract
Background: Accessible indicators of aggressiveness of care at the end-of-life are useful to monitor implementation of
early integrated palliative care practice. To determine the intensity of end-of-life care from exhaustive data combining
administrative databases and hospital clinical records, to evaluate its variability across hospital facilities and associations
with timely introduction of palliative care (PC).
Methods: For this study designed as a decedent series nested in multicentre cohort of advanced cancer patients, we
selected 997 decedents from a cohort of patients hospitalised in 2009–2010, with a diagnosis of metastatic cancer in 3
academic medical centres and 2 comprehensive cancer centres in the Paris area. Hospital data was combined with
nationwide mortality databases. Complete data were collected and checked from clinical records, including first referral
to PC, chemotherapy within 14 days of death, ≥1 intensive care unit (ICU) admission, ≥2 emergency department visits
(ED), and ≥ 2 hospitalizations, all within 30 days of death.
Results: Overall (min-max) indicator values as reported by facility providing care rather than the place of death, were:
16% (8–25%) patients received chemotherapy within 14 days of death, 16% (6–32%) had ≥2 admissions to acute care,
6% (0–15%) had ≥2 emergency visits and 18% (4–35%) had ≥1 intensive care unit admission(s). Only 53% of these
patients met the PC team, and the median (min-max) time between the first intervention of the PC team and death
was 41 (17–112) days. The introduction of PC > 30 days before death was independently associated with lower
intensity of care.
Conclusions: Aggressiveness of end-of-life cancer care is highly variable across centres. This validates the use of
indicators to monitor integrated PC in oncology. Disseminating a quality audit-feedback cycle should contribute to a
shared view of appropriate end-of-life care objectives, and foster action for improvement among care providers.
Keywords: (MesH heading or entry terms), End of life care, Quality of health care, Palliative care, Cancer care facilities,
Academic medical centers, Data collection methods

* Correspondence: isabelle.colombet@parisdescartes.fr;
isabelle.colombet@aphp.fr
1
Unité Fonctionnelle de Médecine Palliative, Hôpital Cochin, Assistance
Publique Hôpitaux de Paris, F-75014 Paris, France
2
Univ Paris Descartes, F-75006 Paris, France
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Colombet et al. BMC Palliative Care (2019) 18:35 Page 2 of 8

Background end-of-life cancer care from exhaustive data combining

Despite increased survival following advances in early administrative databases and hospital clinical records, to
detection and treatment, numbers of deaths from cancer evaluate their variability across hospital facilities and
are expected to increase as a result of the ageing popula- their association with timely integration of palliative
tion. Efforts to improve the quality of end-of-life cancer care.
care are therefore important. One of the challenges for
quality management is the efficient measurement of care
Methods
quality with rapid feedback to healthcare organizations
Design and setting
enabling them to undertake necessary actions for im-
We conducted a retrospective analysis of a nested series
provement. Earle et al. [1] developed indicators involving
of decedents in a cohort selected from the administrative
focus groups with patients, carers and health profes-
data of 2 comprehensive cancer centres and 3 academic
sionals, designed to be easily accessible and measurable
medical centres in the Paris region. All 2010 decedents
from health administrative data and to provide meaning-
were identified from these hospitals’ administrative data-
ful information on the quality of end-of-life cancer care.
bases and by linkage with national death certificates
These indicators describe high-intensity medical care de-
database, to analyse quality indicators for all inpatients
livered in the last month of life, such as overuse of
diagnosed with advanced cancer, whatever their place of
chemotherapy, underuse of hospice care, frequent hospi-
death. Quality indicators were then measured from hos-
talizations, emergency room visits, and intensive care
pital administrative database, completed by data col-
unit admissions. They set some achievable benchmarks
lected from health records. The REporting of studies
from the results of the 10% best-performing providers
Conducted using Observational Routinely collected
[2–4]. The methodology was also tested in Canada [5],
health Data (RECORD) guidelines were followed when
and the indicators were endorsed by the American Na-
relevant to report methods and results [22].
tional Quality Forum [3]. Since these original develop-
Organisation of the hospital-based palliative care con-
ments, and with the computerization of clinical activities
sultation team is similar in each centre. The team com-
and easier access to large health administrative data-
prises at least a palliative care physician and a palliative
bases, these indicators have been used in other countries
care nurse who collaborate systematically with social
[6–8], in child populations [8, 9], or for specific types of
workers and psychologists. They can be called on by at-
cancer [10, 11]. In France, a study described the chemo-
tending physicians to evaluate in- or outpatients, give
therapy indicator, using nation-wide hospital administra-
advice on symptom relief, and provide support for carers
tive data, primarily collected for hospital payment without
or healthcare professionals.
recording outpatient care [12]. In another study, all Earle’s
None of the participating centres has an inpatient pal-
indicators were used to evaluate the effect of integrated
liative care unit.
palliative care on the quality of end-of-life care from ex-
haustive data in an academic medical centre [13].
Based on evidence from several randomized clinical Data sources and study population
trials, American Society of Clinical Oncology (ASCO) as Patients over 18 years of age were selected from each
well as European Society of Medical Society (ESMO) hospital administrative database (PMSI database, the
published guidelines to recommend patients and outpa- French equivalent of DRG database), based on a hospital
tients with advanced cancer should receive dedicated stay coded under metastatic cancer ICD-10 diagnosis
palliative care services, early in the course of disease, (C76_, C78_, C79_, C80_), between October 1, 2009 and
concurrent with active treatment [14, 15]. A Cochrane December 31, 2010. Patients recorded as deceased in
meta-analysis confirmed that early palliative care could 2010 the database of one of the facilities were identified.
improve quality of life and reduce symptom intensity For patients whose death was not found in the hospital
with no effect reaching statistical significance on survival administrative database on the date of the request, a vital
[16]. Some studies found that early palliative care also status search was performed on the National Vital Sta-
had a favorable impact on end-of-life care aggressive- tistics (RNIPP, for National Register for the Identification
ness, suggesting that such indicators as chemotherapy of Private Individuals) database, by application to the
administration or intensive care resource use can be national death certificate database to obtain the cause
considered as interesting to monitor implementation of and place of death [23, 24]. All 2010 decedents in the
early palliative care practice [17–21]. initial cohort selected from hospital administrative data-
In this multicentre study, we selected a cohort of pa- bases were thus identified, whatever their place of death.
tients diagnosed with metastatic cancer in 5 academic Two hundred patients per centre were randomly se-
medical centres or comprehensive cancer centres. The lected from this decedent series, stratified according to
study aimed to determine the intensity and trajectory of age, gender and place of death (in hospital or elsewhere),
Colombet et al. BMC Palliative Care (2019) 18:35 Page 3 of 8

to form a sample of 1000 patients, equally spread across difficult. In addition, we conducted some sensitivity ana-
the 5 participating centres and representative of each. lysis only for the place of death outcome, in order to test
the question of palliative care versus no palliative care
Measures of intensity of care (using other representation of “No intervention” versus
For the 200 patients selected in each centre, a second re- “Intervention of palliative care”), separately from the
quest to each hospital administrative database was made question of early versus late palliative care.
to enable the reconstruction of clinical trajectories in the All statistical tests were two sided and a p-value under
last month of life: the number of visits to emergency 0.05 was considered statistically significant. All analyses
room or oncology clinic, the number of admissions to were performed using SAS software version 9.4.
intensive care unit or to acute care.
Results
Additional data collected from clinical records Study population and patient characteristics per Centre
The data extracted from hospital administrative database A total of 7858 patients were hospitalized with a diagno-
was systematically checked and completed individually sis of metastatic cancer between October 2009 and De-
for each patient by a search of hospital clinical records cember 2010 in the 5 participating centres, among
in each participating centre. Additional data described whom 2063 were identified as decedents in 2010 (see
the exact date of the first intervention of the palliative Fig. 1); 724 (35%) patients who died outside hospital
care team, and, when available, the modalities of last ad- were identified thanks to the request to RNIPP. From
ministered chemotherapy (route and exact date of pre- these 2063 patients, a random sample of 1000 patients
scribing). We also checked clinical records to appreciate was drawn to pursue data collection from clinical re-
whether the centre where the patient was identified was cords. Three patients were excluded from the analysis
the patient’s reference centre for cancer treatment, and due to identity or primary diagnosis errors.
the length of the patient’s follow-up in that centre. The Men accounted for 54%, mean age 66 (±14.2) years at
study protocol was approved by the CEERB (N° the time of death (Table 1). The most frequent primary
IRB00006477). tumor sites were breast (18%), lung (17%), urogenital
(prostate, bladder, kidney, 14%) and colon/rectum (11%).
Statistical analyses In the study population, 599 (66%) patients had ≥2
Quantitative and qualitative variables were described by metastatic sites, most frequently liver (48%), bone (43%),
means (SD) and frequencies (%). For results per centre, lung (42%), peritoneal (20%) and brain (16%).
each patient was reported in the centre where he/she
was identified by request to the hospital administrative Intensity of end-of-life care
database. First, Chi-square and Student tests were per- All indicators were highly variable across centres
formed to assess the associations between outcomes (i.e. (Table 2).
measures of intensity of care) and the following covari- Of the 738 patients for whom the data could be found
ates: age at death, gender, disease incurable at initial in hospital records, 16% received chemotherapy in their
diagnosis, number of metastasis sites, study centre, inter- last 14 days of life (Table 2). This proportion varied across
vention of palliative care team, group of primary tumor centres, from 8.1 to 13.2% in the three academic medical
sites (defined in 3 categories of expected survival accord- centres and reached 16 and 25% in Comprehensive Can-
ing to published French epidemiological data) at least cer Centres 1 and 2. The last line of chemotherapy was
one admission in an intensive care unit, at least one ad- started at a median of 42 days preceding death. It was pre-
mission in acute care, at least one emergency visit. Then, scribed by oral route for 54/236 (23%) patients.
logistic regressions were used to predict the logit of the Concerning clinical trajectories in the last month before
probability of experiencing each outcome. Since the out- death, in the overall study population, 16.4% were hospi-
come indicators of quality of end of life care include the talized twice or more (reaching 32% for Comprehensive
timeframe of the last 30 days of life (e.g. emergency visit, Cancer Centre 2) and 90% of these admissions were moti-
intensive care unit or acute care admission in the last vated by needs for palliative or supportive care.
30 days of life), we represented the timely intervention Sixty-one (6%) patients visited emergency room twice
of palliative care team variable as “Early intervention of or more. Only 10 (3.4%) of these visits led to hospitalisa-
palliative care team (> 30 days before death)” versus “No tion. In all 17.5% were admitted at least once to inten-
or late intervention of palliative care team (< 30 days be- sive care unit, with a median length of stay of 4 days.
fore death)”. Variables with a p-value under 0.05 in the Between-centre variations for these indicators should be
simple analysis were included in the multivariable ana- interpreted bearing in mind that Comprehensive Cancer
lysis, making analysis of the effect of palliative care Centre 1 has no emergency room, and is organized to
within the same timeframe of the last 30 days of life receive patients in need of urgent care in unplanned
Colombet et al. BMC Palliative Care (2019) 18:35 Page 4 of 8

Fig. 1 Flow chart of study

consultations, or addresses them to the emergency room team more than 30 days before death was associated
in a nearby public hospital. with lower likelihood of receiving chemotherapy near
death (OR 0.50 [IC95% 0.30–0.82]), of being admitted to
Impact of palliative care on intensity of end-of-life care acute care in the last month (OR 0.64 [IC95% 0.46–
The median anteriority of follow up differed across cen- 0.89]), and of dying in acute care unit (OR 0.33 [IC95%
tres and the timing of referral to palliative care should 0.23–0.47]). According to sensitivity analyses for the
be read in this context (Table 3). Both comprehensive place of death outcome, the intervention of palliative
cancer centres were considered as the referent centre for care team, whenever its timing, as compared with no
more than 90% of the patients (respectively 93 and 98%) intervention, was still significantly associated with more
and a large majority had been followed for over 6 frequent dying in acute care unit (OR 0.40 [IC95% 0.28–
months. In the overall population, the palliative care 0.57], p < 0.0001) .
team was mobilized for 492 (53%) patients, the propor- No other covariable was significantly associated with
tion ranging from 30% in academic medical centre 2 to intensity of care, except the centre and older age. The
70% in comprehensive cancer centre 1. centre was significantly associated with all indicators,
Table 4 shows the results of the multivariable logistic and older age only associated with less chemotherapy
regression analyses. The intervention of a palliative care near death (OR 0.97 [IC95% 0.96–0.98]).

Table 1 Patient characteristics by centre

TOTAL UH 1 CCC 1 UH 2 CCC 2 UH 3
n = 997 n = 200 n = 200 n = 199 n = 200 n = 198
Age at death, mean (SD) 66 (14) 68 (13) 64 (14) 70 (13) 60 (14) 70 (13)
Men, n (%) 535 (54) 130 (65) 42 (21) 126 (63) 111 (56) 126 (64)
Primary tumor site, n (%) n = 976 n = 198 n = 200 n = 197 n = 200 n = 181
Breast 173 (18) 12 (6.1) 106 (53) 15 (7.6) 24 (12) 16 (8.8)
Lung 165 (17) 32 (16) 27 (14) 48 (24) 40 (20) 18 (9.9)
Urinary tract and kidney 136 (14) 31 (16) 9 (4.5) 40 (20) 19 (9.5) 37 (20)
Colorectal 104 (11) 21 (11) 9 (4.5) 24 (12) 23 (12) 27 (15)
Liver, Pancreas, Biliary tract 89 (9.1) 31 (16) 4 (2.0) 15 (7.6) 10 (5.0) 29 (16)
Other 309 (32) 71 (36) 45 (23) 55 (28) 84 (42) 54 (30)
Abbreviations: CCC Comprehensive Cancer Centre, UH University Hospital
Colombet et al. BMC Palliative Care (2019) 18:35 Page 5 of 8

Table 2 Intensity in end-of-life care, per centre

TOTAL UH 1 CCC 1 UH 2 CCC 2 UH 3
n (%) n (%) n (%) n (%) n (%) n (%)
Chemotherapy in last 14 days of life 116/738 (15.7) 15/126 (11.9) 30/185 (16.2) 21/159 (13.2) 42/169 (24.9) 8/99 (8.1)
Trajectory of care in last month of life n = 997 n = 200 n = 200 n = 199 n = 199 n= 198
1 admission in acute care 520 (52.2) 101 (50.5) 92 (46.0) 113 (56.8) 103 (51.5) 111 (56.1)
≥ 2 admissions in acute care 164 (16.4) 27 (13.5) 35 (17.5) 27 (13.6) 64 (32.0) 11 (5.6)
1 emergency visit 197 (19.8) 57 (28.5) 11 (5.5) 60 (30.2) 89 (44.5) 41 (20.7)
≥ 2 emergency visits 61 (6.1) 9 (4.5) 0 . 13 (6.5) 29 (14.5) 10 (5.1)
≥ 1 admission in Intensive Care Unit 174 (17.5) 35 (17.5) 8 (4.0) 38 (19.1) 23 (11.5) 70 (35.4)
Patients transferred in palliative care unit n = 157 n = 33 n = 58 n = 29 n = 10 n = 19
≤ 3 days before death 12 (7.6) 0 – 3 (5.2) 6 (20.7) 0 – 3 (15.8)
Place of Death n = 978 n = 199 n = 189 n = 199 n = 198 n = 193
Acute care hospital 672 (68.7) 129 (64.8) 95 (50.3) 139 (69.8) 182 (91.9) 127 (65.8)
Acute care ward 583 (59.6) 113 (56.8) 93 (49.2) 115 (57.8) 146 (73.7) 116 (60.1)
Intensive Care Unit 62 (6.3) 15 (7.5) 2 (1.1) 22 (11.1) 17 (8.6) 6 (3.1)
Emergency room 27 (2.8) 1 (0.5) 0 . 2 (1.0) 19 (9.6) 5 (2.6)
Palliative care unit 189 (19.3) 41 (20.6) 60 (31.7) 32 (16.1) 10 (5.1) 46 (23.8)
Home 82 (8.4) 24 (12.1) 13 (6.9) 24 (12.1) 4 (2.0) 17 (8.8)
Other 35 (3.6) 5 (2.5) 21 (11.1) 4 (2.0) 2 (1.0) 3 (1.6)
Abbreviations: CCC Comprehensive Cancer Centre, UH University Hospital

Discussion of patients receiving chemotherapy near death and

This study provides exhaustive baseline data on the in- greater likelihood of dying in palliative care unit or at
tensity and trajectory of end-of-life care delivered to home.
adult populations with metastatic cancer in two compre- Results obtained from large nationwide health admin-
hensive cancer centres and three academic hospitals in istrative data are interesting to give a broad view of prac-
the Paris area. Its results highlight some inter-centre tice. Studies of these indicators from other countries
variability of practice, with max/min ratios between 3 reveal large variations which can be explained by differ-
and 9, depending on indicators. Hardly more than half ences in healthcare systems and public health policies at
patients met the hospital-based palliative care consult- the national level, and by heterogeneity in measurement
ation team. Adjusting for centre effect, the intervention methods and data sources [4, 6, 12, 25–27]. Our findings
of the palliative care team more than a month before remain in the broad range of results published. It is ne-
death was significantly associated with lesser likelihood cessary for a good performance measure to detect

Table 3 Clinical trajectory and context of referral to palliative care

TOTAL UH 1 CCC 1 UH 2 CCC 2 UH 3
Study centre is referent for the patient’s 842/932 (90) 181/200 (91) 196/200 (98) 163/199 (82) 187/200 (94) 115/133 (87)
cancer, n/total (%)
Anteriority of follow up in the centre, 12 (4–37) 10 (4–24) 33 (9–106) 8 (3–21) 17 (6–43) 8 (2–21)
Median time in months (Q1 - Q3)
Intervention by the Palliative Care Team, n/total (%) 492/926 (53) 101/196 (52) 140/199 (70) 59/196 (30) 112/199 (56) 81/136 (60)
ECOG PS at 1st intervention ≤2 95/349 (27) 23/67 (34) 43/83 (52) 5/21 (24) 12/101 (12) 12/77 (16)
Time between first intervention and date of death n = 475 n = 98 n = 138 n = 55 n = 107 n = 77
≤ 7 days 81 (17) 12 (12) 10 (2, 7) 10 (18) 32 (30) 17 (22)
]7–30] days 117 (25) 15 (15) 19 (14) 22 (40) 35 (33) 26 (34)
]30–90] days 128 (27) 39 (40) 31 (23) 14 (26) 24 (22) 20 (26)
> 90 days 149 (31) 32 (33) 78 (57) 9 (16) 16 (15) 14 (18)
median (Q1 - Q3) 41 (13–122) 63 (25–115) 112 (38–281) 25 (11–50) 17 (7–54) 21 (8–64)
Abbreviations: CCC Comprehensive Cancer Centre, UH University Hospital
Colombet et al. BMC Palliative Care (2019) 18:35 Page 6 of 8

Table 4 Unadjusted frequencies of each indicator by delivery of palliative care and multivariable logistic regression predicting
intensity of care near death
Early intervention of PCT No or late intervention of PCT Multivariable analysisa
(> 30 days before death) (< 30 days before death)
Indicators n (%) n (%) OR IC95% p-value
Chemotherapy in last 14 days of life 28/240 (11.7) 89/487 (18.3) 0.50 [0.30–0.82] 0.006
≥ 1 emergency visits 62/282 (22.0) 188/644 (29.2) 1.04 [0.72–1.49] 0.844
≥ 1 admission in ICU 49/282 (17.4) 114/644 (17.7) 1.45 [0.95–2.21] 0.082
≥ 1 admission in acute care 180/282 (63.8) 481/644 (74.7) 0.64 [0.46–0.89] 0.009
Place of death in acute care hospital 153/273 (56.0) 511/637 (80.2) 0.33b [0.23-0.47] < 0.0001
Abbreviations: PCT Palliative Care Team, ICU Intensive Care Unit
a
Odds Ratio of indicator, according to the timing of intervention of PCT, adjusted on age at death, gender, disease incurable at initial diagnosis, number of
metastasis sites, group of primary tumor sites (defined in 3 categories of expected survival according to published French epidemiological data), and study centre
b
also adjusted on previous indicators of intensity of care

differences in quality of care. We obtained a significant reported nationally for cancer deaths (8% versus 19%)
variability of practice between the 5 participating cen- [31]. In their European study from death certificates, Co-
tres, similar to the results of Earle et al. whose ratios be- hen et al. give some insights into between-country varia-
tween the 5% best and 5% worst performing health care tions concerning place of death and quality indicator
geographic areas ranged from 2.2 to 5, according to out- results. French healthcare organization is characterized
come [2]. by a high proportion of people dying in hospital, espe-
We found high frequencies of chemotherapy adminis- cially from cancer (> 70%), alongside one of the highest
tration in the 14 days before death, with overuse more ratios per 10,000 of both acute care hospital beds and
frequent in comprehensive cancer centres than in aca- long term care beds. Among the other 8 European coun-
demic medical centres and higher intensity of care tries participating in the study, France also has the high-
among young patients, as previously reported in the est healthcare and social welfare expenditure, with the
French national hospital administrative database [12]. lowest rates of palliative care services for adults per mil-
Another strength of this study is the collection of data lion inhabitants.
from clinical records giving access to the accurate timing
of the first intervention of the palliative care team which
is not recorded in hospital administrative database [28]. Conclusion
This allowed to describe large variations in the timing of This study brings a first multicentre measure of all Ear-
referral to palliative care teams across centres and to le’s indicators in French setting, from health administra-
analyse the accurate association between this interven- tive data completed by hospital clinical records. It
tion and the indicators of intensity of care. These varia- supports a prospective approach to quality of care,
tions suggest that early palliative care, known to reporting indicators of practice from the point of view of
improve quality-of-life [16] and internationally recom- cancer care providers, by facility providing care rather
mended for patients with advanced cancer [14, 15], is than the place of death. Unlike results obtained from
unequally put into practice. More systematic monitoring large health administrative databases which bring a
of the median time between first referral to palliative macroscopic view of practice at the national level, our
care and death as an indicator of this practice could pro- approach provide practitioners with the opportunity to
vide interesting leverage for change [29, 30]. reflect on their practice knowing their own specificities
However, the study cohort was recruited from aca- and organization, and to engage in a quality
demic medical centres and comprehensive cancer cen- evaluation-improvement cycle at each centre level [32].
tres only, making overall results on indicators not Our results also add to those other practice evaluation
representative of all patients with advanced cancer. As studies which found that early palliative care could have
one centre (comprehensive cancer centre 1) is special- a favourable impact on end-of-life care aggressiveness
ized in breast cancer treatment, and another (academic [20, 21]. This suggests that routine measure of such indi-
medical centre 1) is an expert centre for the diagnosis cators as chemotherapy administration, acute care re-
and treatment of sarcoma, women and young patients source use in the last month of life and timing of first
were over-represented. We found a particularly high rate referral to specialized palliative care should be recom-
of patients admitted at least once in intensive care unit mended to monitor actual implementation of early pal-
during their last month (18%) and the proportion of liative care practice and end-of-life care quality at health
deaths at home observed in our study is lower than that care facility level.
Colombet et al. BMC Palliative Care (2019) 18:35 Page 7 of 8

Acknowledgements France. 8Oncologie, Hôpital Cochin, Hôpitaux Universitaire Paris Centre,

We also thank all other collaborators of the EFIQUAVIE study group: Assistance Publique Hôpitaux de Paris, F-75014 Paris, France.
Dr. Jérôme Alexandre (oncologist, for Cochin centre), Dr. Muriel Mons (IGR),
Dr. François Hemery (Hôp H Mondor), Dr. Samir Bouam (Hôp Cochin), Ilhem Received: 5 January 2019 Accepted: 27 March 2019
Cherrak (HEGP), Pr Gilles Chatellier (HEGP), as collaborators facilitating access
to administrative databases in each centre and providing data management
counseling.
We also thank Dr. Pierre Durieux for his critical review and advice on manuscript. References
1. Earle CC. Identifying potential indicators of the quality of end-of-life Cancer
care from administrative data. J Clin Oncol. 2003;21(6):1133–8.
Funding
2. Earle CC. Evaluating claims-based indicators of the intensity of end-of-life
The EFIQUAVIE study was funded by the French Ministry of Health (“Programme
cancer care. Int J Qual Health Care. 2005;17(6):505–9.
Hospitalier de Recherche Clinique”, from the Direction Générale de l’Offre de Soins;
3. Earle CC, Landrum MB, Souza JM, Neville BA, Weeks JC, Ayanian JZ.
PHRC 2010, grant number AOM 10 249). The funding source had no role in the
Aggressiveness of Cancer care near the end of life: is it a quality-of-care
design and conduct of the study; collection, management, analysis, and
issue? J Clin Oncol. 2008;26(23):3860–6.
interpretation of the data; preparation, review, or approval of the manuscript;
4. Setoguchi S, Earle CC, Glynn R, Stedman M, Polinski JM, Corcoran CP, et al.
and the decision to submit the manuscript for publication.
Comparison of prospective and retrospective indicators of the quality of
end-of-life Cancer care. J Clin Oncol. 2008;26(35):5671–8.
Availability of data and materials 5. Grunfeld E, Lethbridge L, Dewar R, Lawson B, Paszat LF, Johnston G, et al.
Dataset can be made available for review process, upon request to Dr. I Colombet Towards using administrative databases to measure population-based
(corresponding author) or to Aurélie Vilfaillot (aurelie.vilfaillot@aphp.fr). indicators of quality of end-of-life care: testing the methodology. Palliat
Med. 2006;20(8):769–77.
Authors’ contributions 6. Ho TH, Barbera L, Saskin R, Lu H, Neville BA, Earle CC. Trends in the
IC made substantial contributions to the conception AND design of the work aggressiveness of end-of-life Cancer Care in the Universal Health Care
AND the acquisition, analysis AND interpretation of data AND drafted the work System of Ontario, Canada. J Clin Oncol. 2011;29(12):1587–91.
and substantively revised it. CB and PV made substantial contributions to the 7. Hung Y-N, Liu T-W, Lin D-T, Chen Y-C, Chen J-S, Tang ST. Receipt of life-
conception AND design of the work AND interpretation of data AND drafted sustaining treatments for Taiwanese pediatric patients who died of Cancer
the work and substantively revised it. AP, HJ, SV-S made substantial in 2001 to 2010: a retrospective cohort study. Medicine (Baltimore). 2016;
contributions to the acquisition AND interpretation of data. AV made 95(16):e3461.
substantial contributions to the analysis AND interpretation of data AND 8. Park JD, Kang HJ, Kim YA, Jo M, Lee ES, Shin HY, et al. Trends in the
drafted the work and substantively revised it. FG made substantial contributions aggressiveness of end-of-life care for Korean pediatric cancer patients who
to the conception AND interpretation of data. AND ALL authors have read and died in 2007-2010. PLoS One. 2014;9(6):e99888.
approved approved the submitted version AND have agreed both to be 9. Kassam A, Sutradhar R, Widger K, Rapoport A, Pole JD, Nelson K, et al.
personally accountable for their own contributions and to ensure that questions Predictors of and trends in high-intensity end-of-life care among children
related to the accuracy or integrity of any part of the work, even ones in which with Cancer: a population-based study using health services data. J Clin
the author was not personally involved, are appropriately investigated, resolved, Oncol Off J Am Soc Clin Oncol. 2017;35(2):236–42.
and the resolution documented in the literature. 10. Jang RW, Krzyzanowska MK, Zimmermann C, Taback N, Alibhai SMH.
Palliative care and the aggressiveness of end-of-life care in patients with
Ethics approval and consent to participate advanced pancreatic cancer. J Natl Cancer Inst. 2015;107(3):dju424. Available
The study analyses a retrospective series of human decedents and is not from: https://doi.org/10.1093/jnci/dju424.
considered by French law as an interventional research involving human 11. Kao Y-H, Chiang J-K. Effect of hospice care on quality indicators of end-of-
subjects. However, the use of health administrative database planned for life care among patients with liver cancer: a national longitudinal
decedent’s identification and data collection, was not yet clearly regulated population-based study in Taiwan 2000–2011. BMC Palliat Care. 2015;14(1)
by French Law at the time of the protocol elaboration. Therefore, the study Available from: http://bmcpalliatcare.biomedcentral.com/articles/10.1186/
protocol has been submitted to the Institutional Review Board (N° IRB00006477) s12904-015-0036-9. [cited 2018 Jun 18].
of Paris North Hospitals, Paris 7 University, AP-HP (Comité d’Evaluation de l’Ethique 12. Rochigneux P, Raoul JL, Beaussant Y, Aubry R, Goldwasser F, Tournigand C,
des projets de Recherche Biomédicale (CEERB) du GHU Nord) for advice and methods et al. Use of chemotherapy near the end of life: what factors matter? Ann
validation and it has been approved. Oncol Off J Eur Soc Med Oncol. 2017;28(4):809–17.
13. Colombet I, Montheil V, Durand J-P, Gillaizeau F, Niarra R, Jaeger C, et al.
Consent for publication Effect of integrated palliative care on the quality of end-of-life care:
Not applicable retrospective analysis of 521 cancer patients. BMJ Support Palliat Care. 2012;
2(3):239–47.
Competing interests 14. Ferrell BR, Temel JS, Temin S, Alesi ER, Balboni TA, Basch EM, et al.
The authors declare that they have no competing interests. Integration of palliative care into standard oncology care: American Society
of Clinical Oncology clinical practice guideline update. J Clin Oncol. 2017;
35(1):96–112.
Publisher’s Note 15. Kaasa S, Loge JH, Aapro M, Albreht T, Anderson R, Bruera E, et al. Integration
Springer Nature remains neutral with regard to jurisdictional claims in of oncology and palliative care: a lancet oncology commission. Lancet
published maps and institutional affiliations. Oncol. 2018. Available at https://doi.org/10.1016/s1470-2045(18)30415-7.
16. Haun MW, Estel S, Rücker G, Friederich H-C, Villalobos M, Thomas M, et al.
Author details Early palliative care for adults with advanced cancer. Cochrane Database
1
Unité Fonctionnelle de Médecine Palliative, Hôpital Cochin, Assistance Syst Rev. 2017;6:CD011129.
Publique Hôpitaux de Paris, F-75014 Paris, France. 2Univ Paris Descartes, 17. Temel JS, Greer JA, Muzikansky A, Gallagher ER, Admane S, Jackson VA, et al.
F-75006 Paris, France. 3Département de Soins de Support, Institut Curie, Paris, Early palliative care for patients with metastatic non-small-cell lung cancer.
France. 4Unité Mobile d’Accompagnement et de Soins Palliatifs, Hôpital Henri N Engl J Med. 2010;363(8):733–42.
Mondor, Assistance Publique Hôpitaux de Paris, F-94000 Créteil, France. 18. Maltoni M, Scarpi E, Dall’Agata M, Schiavon S, Biasini C, Codecà C, et al.
5
Unité de Recherche Clinique, Hôpital européen G Pompidou, Hôpitaux Systematic versus on-demand early palliative care: a randomised clinical trial
Universitaire Paris Ouest, Assistance Publique Hôpitaux de Paris, F-75015 assessing quality of care and treatment aggressiveness near the end of life.
Paris, France. 6Unité Mobile d’Accompagnement et de Soins Palliatifs, Hôpital Eur J Cancer. 2016;69:110–8.
européen G Pompidou, Hôpitaux Universitaire Paris Ouest, Assistance 19. Temel JS, Greer JA, El-Jawahri A, Pirl WF, Park ER, Jackson VA, et al. Effects of
Publique Hôpitaux de Paris, F-75015 Paris, France. 7Unité Mobile early integrated palliative Care in Patients with Lung and GI Cancer: a
d’Accompagnement et de Soins Palliatifs, Institut Gustave Roussy, Villejuif, randomized clinical trial. J Clin Oncol. 2017;35(8):834–41.
Colombet et al. BMC Palliative Care (2019) 18:35 Page 8 of 8

20. Triplett DP, LeBrett WG, Bryant AK, Bruggeman AR, Matsuno RK, Hwang L, et
al. Effect of palliative care on aggressiveness of end-of-life care among
patients with advanced Cancer. J Oncol Pract. 2017;13(9):e760–9.
21. Scibetta C, Kerr K, Mcguire J, Rabow MW. The costs of waiting: implications
of the timing of palliative care consultation among a cohort of decedents
at a Comprehensive Cancer Center. J Palliat Med. 2016;19(1):69–75.
22. Benchimol EI, Smeeth L, Guttmann A, Harron K, Moher D, Petersen I, et al.
The REporting of studies conducted using observational routinely-collected
health data (RECORD) statement. PLoS Med. 2015;12(10):e1001885.
23. PROCEDURE_ACCES_RNIPP_UTILISATEURS_V2_19_05_2014.pdf. Available
from: http://cesp.vjf.inserm.fr/~webifr/pdf/PROCEDURE_ACCES_RNIPP_
UTILISATEURS_V2_19_05_2014.pdf. [cited 2018 Jun 18]
24. CépiDc - causes médicales de décès. Available from: http://www.cepidc.
inserm.fr/index.php?p=accueil. [cited 2018 Jun 18]
25. Keam B, Oh D-Y, Lee S-H, Kim D-W, Kim MR, Im S-A, et al. Aggressiveness of
Cancer-care near the end-of-life in Korea. Jpn J Clin Oncol. 2008;38(5):381–6.
26. Yun YH, Kwak M, Park SM, Kim S, Choi JS, Lim H-Y, et al. Chemotherapy use
and associated factors among Cancer patients near the end of life.
Oncology. 2007;72(3–4):164–71.
27. Bekelman JE, Halpern SD, Blankart CR, Bynum JP, Cohen J, Fowler R, et al.
Comparison of site of death, health care utilization, and hospital
expenditures for patients dying with Cancer in 7 developed countries.
JAMA. 2016;315(3):272–83.
28. Goldwasser F, Vinant P, Aubry R, Rochigneux P, Beaussant Y, Huillard O, et
al. Timing of palliative care needs reporting and aggressiveness of care near
the end of life in metastatic lung cancer: a national registry-based study.
Cancer. 2018;124(14):3044–51.
29. Vinant P, Joffin I, Serresse L, Grabar S, Jaulmes H, Daoud M, et al. Integration
and activity of hospital-based palliative care consultation teams: the
INSIGHT multicentric cohort study. BMC Palliat Care. 2017;16(1):36.
30. Colombet I, Vinant P, Joffin I, Weiler F, Chaillot N, Moreau N, et al. Suivi
d’indicateurs dans le bilan d’activité d’une équipe mobile de soins palliatifs :
un levier pour l’amélioration des pratiques. Presse Med. 2015;44:e1–11.
31. 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.
32. Cowall DE, Yu BW, Heineken SL, Lewis EN, Chaudhry V, Daugherty JM. End-
of-life Care at a Community Cancer Center. J Oncol Pract. 2012;8(4):e40–4.