Density of dental practitioners and access to dental care for the elderly: A multilevel analysis with a view on socio-economic inequality

Abstract

To examine the relations between density of dental practitioners (DDP) and socio-economic and demographic factors shown to affect access to dental care for the elderly.
Data are taken from a cross-sectional survey - 2008 Disability Healthcare - Household section Survey (HSM). HSM is a representative random sample of French people living in their own domiciles. Our study focuses on the 9233 individuals aged 60 years and above. Multilevel models are employed to disentangle the relations between the determinants of dental care utilisation and DDP. Statistical analyses are conducted using SAS 9.2 and HLM 6.
Low-income and lack of complementary health insurance are associated with higher odds of not having visited a dentist, revealing a high unequal access to dental care. By using multilevel modelling, DDP appears to be a significant factor to access to dental services. When considering the intricate relations between income gradient and DDP, the latter lessens the income-related inequality to access dental services.
DDP seems favouring a more equitable access to dental care, mitigating under-caring of the poorest. This point is to be added in the debate about density of healthcare suppliers.

Full text

Health
Policy
103 (2011) 160–
167
Contents
lists
available
at
SciVerse
ScienceDirect
Health
Policy
j
ourna
l
ho
me
p
ag
e:
www.elsevier.com/locate/healthpol
Density
of
dental
practitioners
and
access
to
dental
care
for
the
elderly:
A
multilevel
analysis
with
a
view
on
socio-economic
inequality
Laurence
Lupi-Peguriera,e,f,∗,
Isabelle
Clerc-Urmesa,b,c,
Mohammad
Abu-Zaineha,c,
Alain
Paraponarisa,b,c,
Bruno
Venteloua,d
aINSERM,
U912
(SE4S),
Marseille,
France
bUniversité
Aix
Marseille,
IRD,
UMR-S912,
Marseille,
France
cORS
PACA,
Observatoire
Régional
de
la
Santé
Provence
Alpes
Côte
d’Azur,
Marseille,
France
dCNRS
GREQAM,
Marseille,
France
eLOM
(Laboratoire
de
Microbiologie
Orale)
URE01
Université
de
Nice
Sophia
Antipolis,
France
fPôle
d’Odontologie,
CHU
de
Nice,
France
a
r
t
i
c
l
e
i
n
f
o
Keywords:
Oral
health
services
Multilevel
Elderly
Social
inequality
Density
of
practitioners
a
b
s
t
r
a
c
t
Objectives:
To
examine
the
relations
between
density
of
dental
practitioners
(DDP)
and
socio-economic
and
demographic
factors
shown
to
affect
access
to
dental
care
for
the
elderly.
Methods:
Data
are
taken
from
a
cross-sectional
survey
–
2008
Disability
Healthcare
–
House-
hold
section
Survey
(HSM).
HSM
is
a
representative
random
sample
of
French
people
living
in
their
own
domiciles.
Our
study
focuses
on
the
9233
individuals
aged
60
years
and
above.
Multilevel
models
are
employed
to
disentangle
the
relations
between
the
determinants
of
dental
care
utilisation
and
DDP.
Statistical
analyses
are
conducted
using
SAS
9.2
and
HLM
6.
Results:
Low-income
and
lack
of
complementary
health
insurance
are
associated
with
higher
odds
of
not
having
visited
a
dentist,
revealing
a
high
unequal
access
to
dental
care.
By
using
multilevel
modelling,
DDP
appears
to
be
a
significant
factor
to
access
to
dental
services.
When
considering
the
intricate
relations
between
income
gradient
and
DDP,
the
latter
lessens
the
income-related
inequality
to
access
dental
services.
Conclusion:
DDP
seems
favouring
a
more
equitable
access
to
dental
care,
mitigating
under-
caring
of
the
poorest.
This
point
is
to
be
added
in
the
debate
about
density
of
healthcare
suppliers.
© 2011 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Thanks
to
advances
in
dental
prevention,
adults
are
now
able
to
keep
healthy
teeth
for
life.
Typically
how-
ever,
their
dental
needs
increase
at
a
time
when
they
may
also
be
undergoing
a
diminished
capacity
to
access
∗Corresponding
author
at:
ORS
PACA,
Unité
U912
SE4S,
23
rue
Stanislas
Torrents,
13006
Marseille,
France.
Tel.:
+33
616195088;
fax:
+33
491598918.
E-mail
address:
laurence.lupi-pegurier@inserm.fr
(L.
Lupi-Pegurier).
care
because
of
retirement,
which
often
implies
reduc-
ing
income
and
health
coverage.
Oral
health
is
essential
to
general
health
and
well-being
[1].
Poor
general
health
and
poor
dental
health
are
interrelated
primarily
because
of
common
risk
factors
that
lead
to
complex
relations
between
oral
infections
–
particularly
periodontitis
–
and
risk
of
chronic
disease
[1,2].
Therefore,
the
elderly
may
face
significant
hurdles
before
obtaining
the
necessary
den-
tal
care
[3].
Indeed,
previous
studies
conducted
elsewhere
have
already
demonstrated
that
not
only
the
physically
impaired,
but
also
the
economically
and
socially
disadvan-
taged
elderly,
are
more
likely
to
experience
tooth
loss
and
0168-8510/$
–
see
front
matter ©
2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.healthpol.2011.09.011

L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167 161
edentulism,
untreated
dental
decay
and
periodontal
dis-
eases
[4–10].
Inequalities
in
access
to
medical
or
dental
care
by
income
have
already
been
demonstrated
[11,12].
How-
ever,
such
analysis
has
never
been
carried
out
in
the
French
context,
where
density
of
practitioners
varies
a
lot
between
regions,
offering
an
occasion
to
refresh
the
issue
of
inequal-
ity
in
access
with
cross
information
on
the
geographical
organization
of
the
healthcare
providers.
In
2008,
there
were
37,810
dentists
in
France.
The
average
density
was
65
per
100,000
individuals
however
substantial
regional
disparities
were
marked,
with
val-
ues
ranging
from
single
to
double.
The
providers
usually
work
in
private
clinics
(almost
90%).
Dental
care
services
are
not
fully
covered
by
the
health
insurance
system
in
France:
while
conservative
and
surgical
care
services
are
rather
cheap
and
well
reimbursed
with
a
fixed-price
sched-
ule,
the
prosthetic
care
is
costly
and
only
covered
by
private
funding,
essentially
through
complementary
insur-
ance
schemes
or
direct
out-of-pocket
payments.
The
observed
discrepancy
between
conservative
and
prosthetic
treatments
has
culminated
in
unbalanced
load
activities:
while
conservative
and
surgical
cares
corre-
spond
to
more
than
two-thirds
of
the
total
dentists’
activities
they
only
represent
35%
of
their
total
sales.
Out-of-pocket
dental
expenditures
borne
by
patients
have
noticeably
increased
over
the
last
decade.
This
is
mainly
due
to
the
development
and
utilisation
of
new
techniques,
which
are
not
listed
in
the
French
“nomenclature”,
and
the
population
aging,
which
has
amplified
the
needs
for
prosthetic
care.
By
contrast,
non-dental
services
and
par-
ticularly
GPs
services
are
still
well
reimbursed
especially
in
the
aged
population
(benefiting
sometimes
of
a
100%
cov-
erage
rate),
justifying
a
reference
of
access
to
dental-care
with
for
ex.
GPs
services.
Although
equity
has
long
been
considered
as
a
key
pillar
of
the
public
healthcare
financing
and
of
the
exten-
sive
regulations
of
providers’
practices
in
France
[13]
regulations
regarding
the
practices
of
dental
providers
remain
significantly
limited
when
compared
with
other
healthcare
services
in
France
and
with
other
countries
of
similar
socio-economic
conditions
[14].
Such
limited
pub-
lic
involvement
in
this
vital
field
of
healthcare
services
is
likely
to
exacerbate
socio-economic
related
inequalities
in
health,
particularly
for
those
who
are
in
most
need;
the
elderly.
This
study
aims
to
examine
inequalities
in
den-
tal
care
utilisation
of
non-institutional
elderly
while
taking
into
account
the
density
of
dental
practitioners
(DDP),
using
national
data
from
the
2008
(HSM)
survey.
2.
Methods
2.1.
The
HSM
survey
and
study
population
The
Disability
Healthcare
Household
section
Survey
(Enquête
Handicap
Santé
–
Ménages,
HSM)
was
under-
taken
between
April
and
October
2008
by
the
French
National
Institute
of
Statistics
and
Economic
Studies
(INSEE)
and
the
French
Directorate
for
Research,
Stud-
ies,
Evaluation
and
Statistics
(DREES).
HSM
is
a
national
cross-sectional
survey
that
focuses
on
health
and
impair-
ment,
as
well
as
the
difficulties
encountered
by
individuals
in
their
daily
lives
[15].
The
survey
collected
information
from
28,500
individuals
residing
in
French
metropoli-
tans
and
overseas
departments.
Individuals
of
all
ages
were
asked
about
their
health
status
(diseases,
disabilities,
functional
limitations,
activity
restrictions,
need
help).
In
addition,
the
survey
contained
information
about
house-
hold’s
socio-demographic
(e.g.,
household
composition,
formal
and
informal
caregivers),
and
socio-economic
(e.g.,
income,
education)
characteristics.
All
information
was
gathered
directly
in
the
domiciles
by
special-trained
inter-
viewers
using
face-to-face
interviews.
In
case
individuals
were
unable
to
respond
to
the
questionnaire
themselves,
another
person
was
asked
to
offer
help.
Respondents
were
informed
about
the
objective
of
the
survey
and
asked
to
give
their
consent
before
the
interviews.
For
the
purpose
of
this
study,
the
unit
of
analysis
was
all
subjects
aged
60
and
above
who
were
able
to
provide
full
information
on
their
oral
health.
In
all,
the
study
included
9233
subjects.
Sample
weights
calculated
by
the
INSEE
were
used
to
adjust
for
missampling
and
to
assure
a
more
representative
sample
at
the
national
level.
To
ensure
sufficient
numbers
to
produce
reliable
national
estimates,
socio-demographic
variable
categories
were
combined
when
necessary.
Measurement
of
utilisation
of
dental
services
in
the
HSM
was
based
on
the
question
“Have
you
consulted
a
dentist
during
the
past
12
months?”.
Among
the
socio-
economic
characteristics
included
in
the
analysis
were:
age
grouped
in
six
categories,
starting
from
60–64
until
85
and
over),
gender,
level
of
education,
number
of
people
in
the
household,
income
per
consumption
unit,
using
a
3-level
scale:
less
than
999D
up
to
more
than
2000D
,
and
the
dwelling
place
(rural
or
urban).
Healthcare
variables
were
apprehended
through
a
set
of
indicators
including:
health
insurance
status,
a
general
indicator
of
morbidity,
assessed
by
both
the
self-reported
general
health
and
the
reported
degree
of
disability
(regrouped
in
three
classes).
Density
of
dental
practitioners
(DDP)
was
captured
by
the
number
of
dentists
per
100,000
individuals,
as
obtained
by
the
French
Ministry
of
Health.
Using
cen-
sus
data,
a
regional
socio-economic
level
was
also
defined
using
the
median
income
per
consumption
unit
of
French
administrative
departments.
Mainland
France
(excluding
overseas
territories)
was
subdivided
into
95
administrative
departments,
referred
to
below
as
areas
of
residence.
This
area-level
has
already
been
used
in
a
previous
study
focus-
ing
on
the
use
of
specialty
care
[16].
Each
area-level
variable
was
then
divided
into
four
categories
(low,
medium-low,
medium-high,
high)
with
the
25th,
50th
and
75th
per-
centiles
as
cut-off
points.
2.2.
Statistical
analysis
We
used
the
Pearson’s
Chi-square
to
compare
qual-
itative
variables
and
ANOVA
F-statistic
for
quantitative
variables
in
order
to
outline
the
characteristics
of
the
stud-
ied
population.
For
each
part
of
our
analysis,
we
started
by
using
logistic
models
to
select
potential
covariates
of
den-
tal
care
utilisation,
after
adjustments
for
some
individual
characteristics.
Then,
we
included
all
the
variables
show-
ing
a
univariate
association
with
our
dependant
variables

162 L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167
(use
of
dental
services
or
use
of
GPs
services)
with
a
p
value
<0.25,
in
the
final
multivariable
logistic
regression
model.
The
use
of
dental
services
is
assumed
to
be
influenced
by
external
factors
operating
at
the
departmental
level
and
not
by
the
elderly’s
own
characteristics.
They
were
investigated
using
a
two-level
hierarchical
logistic
model.
Multilevel
models
are
adapted
to
data
with
a
hierarchical
structure
as
this
is
the
case
with
the
HSM
survey
where
observa-
tions
are
nested
within
departments.
Correlation
between
individuals
of
a
same
group
may
then
bias
the
estimated
coefficients
if
standard
statistical
models
are
used
[17].
The
models
were
estimated
using
a
predictive
quasi-likelihood
method,
implemented
in
HLM®6.
The
relevance
of
using
these
models
was
confirmed
by
the
estimation
of
the
intra-
class
correlation
coefficient
(ICC)
obtained
in
empty
models
indicating
that
2.9%
of
the
total
variance
of
the
use
of
dental
services
was
explained
by
the
departmental
level.
The
like-
lihood
ratio
test
indicated
that
the
ICC
were
significantly
different
from
zero.
Logistic
regression
analyses
are
used
to
study
how
dental
attendance
(or
GP
visits)
varied
with
individual
characteristics
in
each
department.
The
logistic
regression
model
with
random
intercept
and
random
slope
at
level-1
(the
individual
level)
can
be
formally
expressed
as:
log ij
1
−
ij =
ˇ0j+
ˇijx1ij +
·
·
·
+
ˇkjxkij +
eij,
i
=
1,
2,
.
.
.
,
n
where
n
=
97
departmental
levels.
ij is
the
probability
of
the
ith
individual
in
the
jth
department
attending
a
dentist
during
the
previous
year
and
k
is
the
number
of
indepen-
dent
variables
in
the
model.
Therefore,
taking
into
account
departmental
level-2,
there
will
be
(k
+
1)
models:
ˇ0j=
00 +
u0j
ˇ1j=
10 +
u1j
.
.
.
ˇkj =
k0+
ukj
As
one
random
slope
ˇhof
variable
xh(income
per
consumption
unit),
was
significant
while
we
tested
all
possible
random
effects
on
ˇkin
the
list
of
the
k
indepen-
dent
variables
(1
<
h
<
k),
we
also
test
the
assumption
that
ˇhj =
h0+
ˇhZj+
uhj where
Zjis
the
independent
variable
(density
of
practitioners)
observed
at
the
departmental
level.
The
statistical
analyses
were
conducted
using
SAS
9.1
and
HLM
6
packages.
3.
Results
Table
1
reports
descriptive
statistics
on
the
distri-
bution
of
the
surveyed
population
according
to
socio-
demographic,
socio-economic
and
health
characteristics.
Our
sample
examined
9233
individuals
aged
60
years
or
older.
About
60%
of
whom
were
women
with
a
median
age
of
74
years
(range:
60–106
years).
The
majority
of
respondents
(90.3%)
had
a
complementary
public
or
private
health
insurance.
However,
some
differences
in
the
rates
Fig.
1.
Probability
(%)
of
visiting
a
dentist
during
the
previous
year,
by
income
and
DDP.
of
insurance
coverage
were
noticed
in
terms
of
income,
age,
education,
and
dwelling
place
(p
<
0.0001).
Besides,
it
is
worth
noting
that
gender
had
no
statistically
signifi-
cant
effect
on
the
probability
of
having
a
complementary
health
insurance
(p
=
0.23).
Overall,
more
than
one-third
of
the
sample
(40.3%)
reported
a
dental
visit
during
the
year
2007.
All
variables,
including
those
that
appeared
to
be
insignificant
in
the
exploratory
analyses,
are
listed
in
Table
1.
Results
of
the
hierarchical
analysis
on
the
utilisation
of
dental
services
are
presented
in
Table
2.
Multivariate
model
reveals
different
patterns
of
predictors
of
non-visiting
the
dentist
during
the
past
year.
Several
individual
factors
and
two
geographical
factors
are
associated
with
the
outcome.
Table
2
highlights
the
relationships
that
exist
between
den-
tal
services
utilisation
and
a
range
of
potential
confounders,
including
age,
gender,
income,
education
and
health
insur-
ance.
The
effect
of
median
departmental
income
emerges
significant
and
negative,
suggesting
that
the
elderly
liv-
ing
in
affluent
areas
are
more
likely
to
visit
a
dentist
than
their
counterparts
living
in
deprived
areas.
After
adjust-
ing
for
individual
factors,
the
odds
of
consulting
a
dentist
appears
to
be
even
higher
for
wealthy
people
compared
with
their
low-income
counterparts.
Nonetheless,
testing
for
level-2
random
effects
on
slopes
reveals
that
this
latter
relationship
with
income
is
geographically
dependent:
we
obtained
that
the
income-gradient
of
dental
care
utilisation
has
to
be
related
to
the
supply
of
dental
practitioners,
i.e.,
DDP
observed
at
the
departmental
level
(the
interaction
effect
between
the
level
of
income
and
the
DDP
is
demon-
strated
by
a
series
of
significant
odds
ratios
listed
in
the
upper
panel
of
Table
2).
Fig.
1
depicts
the
probabilities
of
visiting
a
dentist
at
least
once
during
the
previous
year,
for
three
levels
of
income
holding
other
independent
variables
constant.
Overall,
the
slope
of
the
regression
line
relating
the
DDP
to
the
probability
of
visiting
a
dentist
is
positive
and
sta-
tistically
significant,
indicating
that
the
probabilities
of
visiting
a
dentist
tend
to
increase
with
individuals’
income,
regardless
of
the
differences
in
the
degree
of
DDP.
Quite
interestingly,
income-related
differences
in
the
use
of
den-
tal
services
appear
to
be
even
more
marked
by
the
side
of
low
DDP.
Indeed,
when
DDP
is
low,
the
probability
of
using

L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167 163
Table
1
Use
of
dental
and
GPs
services
in
the
surveyed
population.
n
(%)
Number
of
subjects
who
consulted
a
dentist
during
the
past
year
(%)
p*Number
of
subjects
who
consulted
a
GP
during
the
past
year
(%)
p**
9233
(100.00) 3718
(40.3) 8866
(96.0)
Age
[60–64]
1834
(19.86)
949
(51.7)
0.000
1713
(93.4)
0.000
[65–69]
1420
(15.38)
668
(47.0)
1345
(94.7)
[70–74]
1704
(18.46)
736
(43.2)
1642
(96.4)
[75–79]
1824
(19.76)
680
(37.3)
1774
(97.3)
[80–84]
1343
(14.55)
425
(31.6)
1305
(97.2)
85
and
over 1108
(12.00)
260
(23.5)
1087
(98.1)
Gender
Men
3721
(40.30)
1484
(39.9)
3524
(94.7)
Women 5512 (59.70)
2234 (40.5)
5342
(95.8)
Education
Less
than
baccalaureat
3123
(33.82)
939
(30.1)
0.000
3011
(96.4)
0.000
Baccalaureat
4862
(52.66)
2020
(41.5)
4697
(96.6)
More
than
baccalaureat
1248
(13.52)
759
(60.8)
1158
(92.8)
Income
Unknown
904
(9.79)
360
(39.8)
0.000
862
(95.4)
0.000
<999
euros 2950 (31.95)
874
(29.6)
2843
(96.4)
[1000–1999
euros]
3968
(42.98)
1652
(41.6)
3836
(96.7)
>2000
euros
1411
(15.28)
832
(58.9)
1325
(93.9)
People
in
the
household
1 2979 (32.26)
1117 (37.5)
0.000
2865
(96.2)
0.450
2
4667
(50.55)
2116
(45.3)
4486
(96.1)
3
and
more 1587
(17.19)
485
(30.6)
1515
(95.5)
Dwelling
place
Rural
2338
(25.32)
851
(36.4)
2262
(96.7)
Urban
6895
(74.68)
2867
(41.6)
6604
(95.8)
Health
insurance
Complementary
insurance
7950
(86.10)
3369
(42.4)
0.000
7677
(96.6)
0.000
CMUC
389
(4.21)
112
(28.8)
363
(93.3)
No
complementary
insurance 894
(9.69)
237
(26.5)
826
(92.4)
Self-reported
general
health
Self-reported
general
health
good
or
very
good
2278
(24.67)
1100
(48.3)
0.000
2054
(90.2)
0.000
Self-reported
general
health
fair 3327
(36.03)
1347
(40.5)
3241
(97.4)
Self-reported
general
health
poor
or
very
poor
3628
(39.29)
1271
(35.0)
3571
(98.4)
Self-reported
oral
health
Self-reported
oral
health
good
or
very
good
3770
(40.83)
1665
(44.2)
0.000
2933
(94.9)
0.000
Self-reported
oral
health
fair
2970
(32.17)
1173
(39.5)
2545
(95.9)
Self-reported
oral
health
poor
or
very
poor
2493
(27.00)
880
(35.3)
3388
(97.1)
Reported
disability
Not
impaired
in
daily
life 3830
(41.48)
1309
(34.2)
0.000
3766
(98.3)
0.000
Slightly
impaired
in
daily
life
2720
(29.46)
1119
(41.1)
2646
(97.3)
Very
impaired
in
daily
life
(ref.)
2683
(29.06)
1290
(48.0)
2454
(91.5)
*Chi-square
test:
use
dental
services
versus
not.
** Chi-square
test:
use
GP
services
versus
not.
dental
services
is
almost
1.5
as
high
when
income
is
high,
suggesting
that
the
current
structure
of
the
supply-side
of
dental
care
play
a
central
role
in
generating
and
protracting
the
prevailing
income-related
inequalities
in
dental
care
utilisation.
In
order
to
allow
for
comparisons,
we
conducted
simi-
lar
analysis
for
the
case
of
general
practitioners
(GP).
In
our
sample,
96.03%
of
the
elderly
consulted
a
GP
during
the
last
year,
whereas
only
40.27%
visited
a
dentist.
Results,
which
are
presented
in
Appendix
A,
reveal
that
at
odds
with
the
case
of
dental
care,
both
income
and
density
vari-
ables
emerge
to
have
insignificant
effect
on
the
utilisation
of
GP
services.
In
effect,
the
latter
appear
to
depend
only
on
complementary
health
insurance
coverage.
4.
Discussion
This
paper
attempts
to
uncover
the
factors
that
shape
dental
care
utilisation
for
the
elderly
in
France,
with
a
particular
focus
on
the
contextual
factors
beyond
the

164 L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167
Table
2
Results
of
the
hierarchical
analysis
of
dental
services
utilisation
(n
=
9233).
Fixed
effects
Coefficient
Odds
ratio
(IC
95%)
p-Value
Intercept 3.74
42.25
(23.17–77.04)
0.000
•
Median
income
in
the
department
(<p
25%)
−0.35
0.70
(0.60–0.83)
0.000
•
Median
income
in
the
department
(<p
50%)
−0.17
0.84
(0.72–0.99)
0.033
•
Median
income
in
the
department
(<p
75%) −0.14
0.87
(0.76–0.99) 0.037
•
Median
income
in
the
department
(>p
75%)
(ref.) 1
Individual
income
per
consumption
unit
Unknown
−0.11
0.89
(0.70–1.13)
0.357
•
Very
low
density
of
dental
practitioners
(<p
25%)
−0.57
0.57
(0.41–0.79)
0.001
•
Low
density
of
dental
practitioners
(<p
50%) −0.40
0.67
(0.49–0.92)
0.014
•
High
density
of
dental
practitioners
(<p
75%)
−0.14
0.87
(0.64–1.17)
0.359
Less
than
999
euros
−0.47
0.63
(0.50–0.78)
0.000
•
Very
low
density
of
dental
practitioners
(<p
25%)
−0.48
0.62
(0.48–0.81)
0.001
•
Low
density
of
dental
practitioners
(<p
50%)
−0.21
0.81
(0.61–1.07)
0.139
•
High
density
of
dental
practitioners
(<p
75%) −0.14
0.87
(0.69–1.09) 0.221
Between
1000
and
1999
euros
−0.20
0.82
(0.67–0.99)
0.040
•
Very
low
density
of
dental
practitioners
(<p
25%)
−0.37
0.69
(0.55–0.86)
0.002
•
Low
density
of
dental
practitioners
(<p
50%)
−0.15
0.86
(0.70–1.05)
0.143
•
High
density
of
dental
practitioners
(<p
75%)
−0.23
0.79
(0.67–0.94)
0.008
More
than
2000
euros
(ref.) 1
Age −0.04
0.96
(0.95–0.97)
0.000
Gender
Men
−0.22
0.81
(0.74–0.87)
0.000
Women
(ref.)
1
Educational
level
No
degree
or
less
than
baccalaureat −0.64
0.53
(0.45–0.63) 0.000
Baccalaureat
−0.43
0.65
(0.56–0.76)
0.000
More
than
baccalaureat
(ref.) 1
Complementary
health
insurance
No
complementary
insurance
−0.23
0.79
(0.63–1.00)
0.052
CMUC
beneficiaries
−0.44
0.65
(0.55–0.75)
0.000
Complementary
insurance
(ref.) 1
Household
size
One
people
(alone)
(ref.)
1
Two
people
0.12
1.13
(1.00–1.28)
0.054
Three
people
or
more
−0.31
0.74
(0.63–0.86)
0.000
Dwelling
place
Rural
place
−0.12
0.89
(0.78–1.01)
0.062
Urban
place
(ref.) 1
Self-perceived
general
health
(SPGH)
Good
or
very
good
SPGH
−0.04
0.96
(0.85–1.09)
0.501
Fair
SPGH
0.01
1.01
(0.91–1.11)
0.868
Poor
or
very
poor
SPGH
(ref.) 1
Impairment
Not
impaired
in
daily
life
0.18
1.2
(1.03–1.40)
0.018
Slightly
impaired
in
daily
life
0.11
1.12
(1.00–1.25)
0.045
Very
impaired
in
daily
life
(ref.)
1
individual
factors.
The
complex
association
between
indi-
vidual
characteristics
such
as
income
and
health
insurance
coverage,
on
the
one
hand,
and
the
contextual
factors
such
as
the
density
of
dental
practitioners,
on
the
other
hand,
has
been
examined
using
a
multilevel
modelling
strategy.
The
analysis
conducted
in
this
paper
clearly
demon-
strates
unequal
utilisation
of
dental
care
across
income
groups
of
the
elderly
population:
the
low-income
elderly
group
appears
to
be
less
likely
to
use
dental
care
services
compared
with
their
wealthier
counterparts
(as
reflected
by
the
income
gradient
in
the
respective
odds
ratios
of
con-
sulting
a
dentist).
Generally,
such
finding
is
in
line
with
those
previously
reported
in
several
studies
conducted
in
developed
countries
[1,2,4,7,10,18–20].
However,
by
carefully
considering
the
intricate
relationship
between
individuals’
incomes
and
the
density
of
dental
practition-
ers,
results
indicate
that
the
higher
the
density
of
dental
practitioners
is
the
lower
the
role
of
income
would
be.
Such
results,
which
are
captured
by
the
variation
in
the
odds
ratios
of
dental
care
utilisation
as
per
different
degrees
of
density,
suggest
that
the
latter
can
rather
play
an
important
role
in
mitigating
the
prevailing
income-related
inequali-
ties
in
this
sector
of
healthcare.
The
few
studies
which
attempted
to
assess
inequali-
ties
in
dental
care
sector
have
produced
some
conflicting
results.
For
instance,
Bower
et
al.
[11]
found
no
significant

L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167 165
relationships
between
the
effect
of
area-based
measure
of
income
and
the
number
of
healthy
teeth
when
clustering
of
individuals
and
their
characteristics
are
taken
into
account.
Conversely,
the
area-based
measure
of
income
emerged
to
be
an
important
determinant
of
dental
care
utilisation
in
studies
conducted
by
Aggarwal
[21],
Locker
and
Ford
[22],
Locker
and
Ford
[23].
While
our
results
corroborate
the
latter
findings
on
the
role
of
area
of
living
(e.g.,
living
in
an
affluent
area
not
only
tend
to
increase
the
odds
of
consulting
a
dentist
but
also
the
overall
health
outcome),
they
also
confirm
the
importance
of
taking
into
account
not
only
individual
clustering
and
their
characteristics
but
also
the
contextual
factors
related
to
the
supply-side
of
dental
care
services,
density
of
practitioners
as
a
main
example.
Indeed,
our
results
regarding
the
specific
effect
of
sup-
plier
density
on
inequalities
in
dental
care
utilisation
while
strongly
indicate
the
presence
of
economic
bar-
riers
for
not
receiving
dental
services
(e.g.,
individuals’
incomes),
suggest
that
such
barriers
can
be
largely
miti-
gated
through
achieving
a
more
adequate
and
equitable
spatial
distribution
of
dental
care
supply.
Achieving
a
more
unbiased
distribution
of
dental
care
supply
shall
unleash
the
suppressed
demand
for
dental
care
services,
and
finally,
seems
to
improve
the
overall
dental
well-
being
of
individuals
regardless
of
their
socio-economic
status.
A
pending
research
question
would
be
to
verify
whether
the
extra
consultations
offered
in
high
DDP
areas
are
really
efficient
and
not
a
pure
phenomenon
of
“supplier-
induced
demand”
(SID)
[24,25].
Sintonen
and
Linnosmaa
[26],
Grytten
et
al.
[27]
argue
that
a
positive
association
between
the
dentist–population
ratio
and
utilisation
in
a
fixed
price
setting
can
be
posited
as
an
evidence
of
supplier-
induced
demand
in
dental
care.
Birch
[28]
suggests
that
a
positive
correlation
between
the
number
of
dentists
per
capita
and
treatment
content
per
visit
may
be
a
suffi-
cient,
but
not
necessary,
evidence
for
SID.
In
our
view,
the
fact
that
extra
consultations
benefit
to
the
poorest
people,
generally
under-deserved,
seems
to
give
an
argument
in
favour
of
“efficient”
extra
consultations
due
to
DDP.
But
we
recognise
that
this
is
difficult
to
conclude
in
the
lack
of
objective
measurement
of
individual’s
dental-health
outcomes
(the
survey
only
gives
a
subjective
self-report
assessment).
Lastly,
our
study
shows
that
the
probability
of
an
annual
dental
visit
in
France
is
lower
compared
with
the
prob-
ability
of
GPs
annual
encounters.
The
odds
of
consulting
a
dentist
during
the
previous
year
are
also
lower
com-
pared
with
other
countries
[29].
This
might
be
related
to
what
can
be
called
“unrecognised
need”
or
“dental
anx-
iety”.
An
advice
from
a
medical
professional
such
as
a
GP
could
help
remove
these
barriers.
Given
the
fact
that
results
from
the
hierarchical
model
of
the
utilisation
of
GP
care
does
not
demonstrate
the
same
social
gradient
as
that
of
the
dentist,
the
former
can
well
be
used
to
provide
dental
advice
to
individuals
who
confront
difficulties
in
accessing
dental
services.
This
strategy
to
involve
GP
in
promot-
ing
oral
health
has
already
been
advocated
by
the
United
States
Public
Health
Service
Office
of
the
Surgeon
Gen-
eral
[30]
and
by
Hale
[31].
However,
instead
of
trying
to
ask
GP
to
diagnose
oral
diseases
themselves,
which
seems
unlikely,
due
to
a
lack
of
time
and
specific
training
[32],
it
could
be
better
to
use
the
GP
as
an
advisor
for
dental
care
access.
Although
the
analysis
conducted
in
this
paper
has
shed
light
on
the
sources
of
inequalities
in
the
utilisa-
tion
of
dental
care
services
as
well
as
the
policy-relevant
measures
that
can
be
used
to
mitigate
them,
some
prac-
tical
limitations
of
our
study
are
worth
mentioning.
First,
the
analysis
was
conducted
using
self-reporting
data.
The
latter
is
arguably
considered
to
be
less
accurate
com-
pared
with
clinical
observation.
However,
the
HSM
survey
has
provided
us
for
the
first
time
in
France
with
reli-
able
estimates
that
are
nationally
representative.
Secondly,
although,
this
paper
has
considered
issues
related
to
access
to
dental
care
services,
our
analysis
was
confined
to
the
non-institutionalised
and
functionally
independent
seg-
ment
of
the
elderly
population,
which
constitute
according
to
the
last
census,
more
than
95%
of
elderly
(>60
years)
[33].
5.
Conclusion
Given
the
lack
of
previous
studies
in
this
specific
area
of
research
in
France,
this
paper
is
expected
to
contribute
to
the
ongoing
debate
surrounding
the
role
of
contextual
factors
affecting
dental
care
service
utilisation.
Despite
their
limitations,
results
presented
in
this
study
can
help
formulate
appropriate
policy
measures
to
facili-
tate
access
to
dental
care
services.
Among
these
measures
is
the
integration
of
dental
services
in
the
GP
practices,
which
can
then
offer
patients
better
information
about
oral
health.
A
need
is
there
to
restructure
the
supply-side
of
dental
care
market
through
creating
appropriate
measures
that
can
induce
dental
practitioners
to
be
also
installed
in
the
underserved
areas.
This
policy
is,
in
our
view,
comple-
mentary
to
other
mechanisms
generally
used
to
solve
the
shortages
in
patterns
of
service
provision
and
utilisation,
such
as
a
“mixed”
payment
system
(mix
fee-for-service
and
capitation),
or
the
existing
market
price
mechanism,
which
is
more
at
risk
to
create
undesirable
effects.
Conflicts
of
interests
None.
Fundings
This
work
was
supported
by
the
French
Directorate
for
Research,
Studies,
Evaluation
and
Statistics
(DREES,
Direction
de
la
Recherche,
des
Etudes,
de
l’Evaluation
et
des
Statistiques);
and
the
National
Solidarity
Fund
for
Autonomy
(CNSA
Caisse
Nationale
de
Solidarité
pour
l’Autonomie)
(permanent
invitation
to
tender
on
disability
and
loss
of
independence).
Acknowledgement
Thanks
to
Bérengère
DAVIN
for
her
helpful
participation
in
the
development
of
the
database.

166 L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167
Appendix
A.
Results
of
the
hierarchical
analysis
of
GP
services
utilisation
(n
=
9233)
Fixed
effects
Coefficient
Odds
ratio
(IC
95%)
p-Value
Intercept 2.24
9.41
(2.86–30.95)
0.001
•
Median
income
in
the
department
(<p
25%)
−0.16
0.85
(0.49–1.49)
0.574
•
Median
income
in
the
department
(<p
50%)
0.42
1.52
(0.97–2.38)
0.067
•
Median
income
in
the
department
(<p
75%)
0.26
1.30
(0.83–2.04)
0.253
•
Median
income
in
the
department
(>p
75%)
(ref.) 1
Individual
income
per
consumption
unit
Unknown
0.03
1.03
(0.64–1.66)
0.912
•
Very
low
density
of
GP
(<p
25%)
−0.43
0.65
(0.33–1.26)
0.202
•
Low
density
of
GP
(<p
50%) 0.14
1.15
(0.50–2.62)
0.743
•
High
density
of
GP
(<p
75%)
−0.11
0.89
(0.40–1.98)
0.784
Less
than
999
euros
0.21
1.24
(0.83–1.86)
0.304
•
Very
low
density
of
GP
(<p
25%)
−0.78
0.46
(0.26–0.80)
0.007
•
Low
density
of
GP
(<p
50%)
0.08
1.08
(0.65–1.80)
0.761
•
High
density
of
GP
(<p
75%) −0.06
0.94
(0.43–2.03) 0.871
Between
999
and
1999
euros
−0.04
0.96
(0.55–1.66)
0.877
•
Very
low
density
of
GP
(<p
25%) −0.00
1.00
(0.55–1.82)
0.994
•
Low
density
of
GP
(<p
50%)
0.20
1.22
(0.69–2.15)
0.502
•
High
density
of
GP
(<p
75%)
0.55
1.73
(1.03–2.91)
0.039
More
than
2000
euros
(ref.)
1
Age 0.03
1.03
(1.01–1.04) 0.001
Gender
Men
−0.43
0.65
(0.55–0.78)
0.000
Women
(ref.)
1
Educational
level
No
degree
or
less
than
baccalaureat 0.25
1.29
(0.96–1.73) 0.089
Baccalaureat
0.32
1.38
(1.03–1.85)
0.029
More
than
baccalaureat
(ref.) 1
Complementary
health
insurance
No
complementary
insurance
−0.91
0.40
(0.30–0.55)
0.000
CMUC
beneficiaries
−0.45
0.64
(0.43–0.93)
0.021
Complementary
insurance
(ref.) 1
Household
size
One
people
(alone)
(ref.)
1
Two
people
0.22
1.25
(0.98–1.60)
0.076
Three
people
or
more
0.07
1.07
(0.78–1.47)
0.663
Dwelling
place
Rural
place
−0.06
0.94
(0.67–1.32)
0.737
Urban
place
(ref.) 1
Self-perceived
general
health
(SPGH)
Good
or
very
good
SPGH
−1.42
0.24
(0.15–0.38)
0.000
Fair
SPGH
−0.37
0.69
(0.43–1.11)
0.123
Poor
or
very
poor
SPGH
(ref.) 1
Impairment
Not
impaired
in
daily
life
−0.08
0.92
(0.63–1.36)
0.682
Slightly
impaired
in
daily
life
−0.74
0.48
(0.30–0.76)
0.002
Very
impaired
in
daily
life
(ref.)
1
References
[1]
Petersen
PE,
Bourgeois
D,
Ogawa
H,
Estupinan-Day
S,
Ndiaye
C.
The
global
burden
of
oral
diseases
and
risks
to
oral
health.
Bulletin
of
the
World
Health
Organization
2005;83:661–9.
[2]
Dolan
TA,
Atchison
KA.
Implications
of
access,
utilization
and
need
for
oral
health
care
by
the
non-institutionalized
and
institutionalized
elderly
on
the
dental
delivery
system.
Journal
of
Dental
Education
1993;57:876–87.
[3]
Dolan
TA,
Atchison
K,
Huynh
TN.
Access
to
dental
care
among
older
adults
in
the
United
States.
Journal
of
Dental
Education
2005;69:961–74.
[4]
Slade
GD,
Spencer
AJ,
Gorkic
E,
Andrews
G.
Oral
health
status
and
treatment
needs
of
non-institutionalized
persons
aged
60+
in
Ade-
laide,
South
Australia.
Australian
Dental
Journal
1993;38:373–80.
[5]
Lundgren
M,
Osterberg
T,
Emilson
G,
Steen
B.
Oral
complaints
and
utilization
of
dental
services
in
relation
to
general
health
factors
in
a
88-year-old
Swedish
population.
Gerodontology
1995;12:81–8.
[6]
Avlund
K,
Holm-Pedersen
P,
Schroll
M.
Functional
ability
and
oral
health
among
older
people:
a
longitudinal
study
from
age
75
to
80.
Journal
of
the
American
Geriatrics
Society
2001;49:
954–62.
[7] Brothwell
DJ,
Jay
M,
Schönwetter
DJ.
Dental
service
utilization
by
independently
dwelling
older
adults
in
Manitoba,
Canada.
Journal
of
Canadian
Dental
Association
2008;74:161.
[8]
Holst
D.
Oral
health
equality
during
30
years
in
Norway.
Community
Dentistry
and
Oral
Epidemiology
2008;36:326–34.
[9]
Wamala
S,
Merlo
J,
Boström
G.
Inequity
in
access
to
dental
care
ser-
vices
explains
current
socioeconomic
disparities
in
oral
health:
the
Swedish
National
Surveys
of
Public
Health
2004–2005.
Journal
of
Epidemiology
and
Community
Health
2006;60:1027–33.
[10]
Manski
RJ,
Moeller
J,
Chen
H,
St
Clair
PA,
Schimmel
J,
Magder
L,
et
al.
Dental
care
utilization
and
retirement.
Journal
of
Public
Health
Den-
tistry
2010;70:67–75.

L.
Lupi-Pegurier
et
al.
/
Health
Policy
103 (2011) 160–
167 167
[11]
Bower
E,
Gulliford
M,
Steele
J,
Newton
T.
Area
deprivation
and
oral
health
in
Scottish
adults:
a
multilevel
study.
Community
Dentistry
and
Oral
Epidemiology
2007;35:118–29.
[12]
van
Doorslaer
E,
Masseria
C,
Koolman
X,
OECD
Health
Equity
Research
Group.
Inequalities
in
access
to
medical
care
by
income
in
developed
countries.
Canadian
Medical
Association
Journal
2006;174:177–83.
[13]
Huber
H.
Decomposing
the
causes
of
inequalities
in
health
care
use:
a
micro-simulations
approach.
Journal
of
Health
Economics
2008;27:1605–13.
[14] van
Doorslaer
E,
Wagstaff
A,
van
der
Burg
H,
Christiansen
T,
De
Graeve
D,
Duchesne
I,
et
al.
Equity
in
the
delivery
of
health
care
in
Europe
and
the
US.
Journal
of
Health
Economics
2000;19(5):553–83.
[15]
Bouvier
G.
L’approche
du
handicap
par
les
limitations
fonctionnelles
et
la
restriction
globale
d’activité
chez
les
adultes
de
20
à
59
ans.
France,
portrait
social.
INSEE
2009;12:5–142.
[16]
Chaix
B,
Boëlle
PY,
Guilbert
P,
Chauvin
P.
Area-level
determinants
of
specialty
care
utilization
in
France:
a
multilevel
analysis.
Public
Health
2005;119:97–104.
[17]
Snijders
TAB,
Bosker
AS.
Multilevel
analysis:
an
introduction
to
basic
and
advanced
multilevel
modeling.
London,
CA:
Sage
Thousand
Oaks;
1999.
[18]
Kiyak
HA,
Reichmuth
M.
Barriers
to
and
enablers
of
older
adults’
use
of
dental
services.
Journal
of
Dental
Education
2005;69:975–86.
[19]
Manski
RJ,
Goodman
HS,
Reid
BC,
Macek
MD.
Dental
insurance
visits
and
expenditures
among
older
adults.
American
Journal
of
Public
Health
2004;94:759–64.
[20]
Macek
MD,
Cohen
LA,
Reid
BC,
Manski
RJ.
Dental
visits
among
US
older
adults,
1999:
the
role
of
dentition
and
cost.
Journal
of
the
American
Dental
Association
2004;135:1154–62.
[21] Aggarwal
VR,
Macfarlane
TV,
Macfarlane
GJ.
Why
is
pain
more
common
amongst
people
living
in
areas
of
low
socio-economic
sta-
tus?
A
population-based
cross-sectional
study.
British
Dental
Journal
2003;194:383–7.
[22]
Locker
D,
Ford
J.
Using
area-based
measures
of
socioeconomic
status
in
dental
health
services
research.
Journal
of
Public
Health
Dentistry
1996;56:69–75.
[23]
Locker
D,
Ford
J.
Evaluation
of
an
area-based
measure
as
an
indi-
cator
of
inequalities
in
oral
health.
Community
Dentistry
and
Oral
Epidemiology
1994;22:80–5.
[24] Rice
N,
Smith
PC.
Ethics
and
geographical
equity
in
health
care.
Jour-
nal
of
Medical
Ethics
2001;27:256–61.
[25] Grytten
J.
Supplier
inducement—its
relative
effect
on
demand
and
utilization.
Community
Dentistry
and
Oral
Epidemiology
1992;20:6–9.
[26]
Sintonen
H,
Linnosmaa
I.
Economics
of
dental
services.
In:
Culyer
A,
Newhouse
JP,
editors.
Handbook
oh
health
economics.
Volume
1B.
Coll.
handbooks
in
economics
bk
17.
Amsterdam/NY:
Elsevier;
2000.
p.
1252–92
[Chapter
24].
[27] Grytten
J,
Holst
D,
Laake
P.
Supplier
inducement.
Its
effect
on
den-
tal
services
in
Norway.
Journal
of
Health
Economics
1990;9(4):
483–91.
[28]
Birch
S.
The
identification
of
supplier-inducement
in
a
fixed
price
system
of
health
care
provision.
The
case
of
dentistry
in
the
United
Kingdom.
Journal
of
Health
Economics
1988;7(2):129–50.
[29] van
Doorslaer
E,
Koolman
X.
Explaining
the
differences
in
income-
related
health
inequalities
across
European
countries.
Health
Economics
2004;13:609–28.
[30]
United
States
Public
Health
Service
Office
of
the
Surgeon
General.
Oral
health
in
America:
a
report
of
the
surgeon
general.
Rockville:
Dept
of
Health
and
Human
Services,
US
Public
Health
Service;
2000.
[31]
Hale
KJ.
Oral
health
risk
assessment
timing
and
establishment
of
the
dental
home.
Pediatrics
2003;111:1113–6.
[32]
Jones
TV,
Siegel
MJ,
Schneider
JR.
Recognition
and
management
of
oral
health
problems
in
older
adults
by
physicians.
A
pilot
study.
JABF
1998;11:474–7.
[33]
Delbès
C,
Gaymu
J.
Qui
vit
en
institution?
[Who
lives
in
an
institu-
tion?].
Gérontologie
et
société
2005;1(112):13.