Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Scientific Research Report
Sociodemographic factors impacting the spatial
distribution of private dental clinics in major cities of
Peoples Republic of China
Pengbo Liu
a,#
, Xuyuan Zhang
b,#
, Guoying Deng
c
**, Weihua Guo
a,d,e
*
a
State Key Laboratory of Oral Disease & National Clinical Research Center for Oral Diseases, West China Hospital of
Stomatology, Sichuan University, Chengdu, Sichuan, PR China
b
Department of Economics, University of Michigan, Ann Arbor, Michigan, USA
c
School of Economics, Sichuan University, Chengdu, Sichuan, PR China
d
Yunnan Key Laboratory of Stomatology, Kunming Medical University, Kunming, Yunnan, PR China
e
Department of Pediatric Dentistry, The Affiliated Stomatology Hospital of Kunming Medical University, Kunming,
Yunnan, PR China
ARTICLE INFO
Article history:
Received 30 November 2023
Received in revised form
4 March 2024
Accepted 15 March 2024
Available online 16 April 2024
ABSTRACT
Objectives: Investigate the geographical distribution of private dental practices in
major Chinese cities and analyze the variables influencing this distribution.
Methods: This study used Python to extract various types of Point of Interest (POI) data
spanning from 2016 to 2022 from the AutoNavi map. A 1km*1km grid was constructed to
establish the study sample. Additional spatial pattern data, including nighttime lighting,
population, and air quality data, were integrated into this grid. Global Moran’s I index was
used to analyze the spatial autocorrelation. The spatial lag model was used to explore the
influencing factors of private dental practice distribution.
Results: This study reveals a specific clustering pattern for private dental practices in
major Chinese cities. The primary influencing factors include nighttime lights, population
density, and housing prices, suggesting that dental practices are typically concentrated in
highly developed regions with dense populations and high housing costs. Additionally, we
discovered that patterns vary across different metropolises, with the most pronounced
clustering patterns and substantial inequalities found in the most developed areas.
Conclusions: This study establishes that factors such as regional development and popula-
tion density positively correlate with private dental practice. Additionally, it reveals a
strong mutual correlation in the clustering of dental practices, which does not show a sub-
stantial correlation with public resources. Finally, it suggests that the spatial heterogeneity
pattern implies a rising necessity to tackle inequality issues within urban areas as eco-
nomic development progresses.
Ó2024 The Authors. Published by Elsevier Inc. on behalf of FDI World Dental Federation.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Key words:
Dental practice
Spatial distribution
Regression analysis
Dental accessibility
China
Introduction
The World Health Organization’s “Global Oral Health Action
Plan” highlighted the influence of commercial and social
determinants.
1
The general oral health status of Chinese resi-
dents is a serious concern due to traditional ignorance about
oral health and socioeconomic reasons. Over the past 20 years,
shifts in Chinese economic status, demographics, societal and
environmental elements, disease trends, and scientific
advancements have catalyzed the demand for general
medical and oral health products.
2
Dental departments
*Corresponding author. Weihua Guo, West China Hospital of Sto-
matology, Sichuan University, No. 14, 3rd Section, Renmin nan
Road, Chengdu, Sichuan, 610041, PR China.
** Corresponding author. Guoying Deng, School of Economics,
Sichuan University, Chengdu, Sichuan, 610041, PR China.
E-mail addresses: dengguoying@scu.edu.cn (G. Deng),
guoweihua943019@163.com (W. Guo).
Weihua Guo: http://orcid.org/0000-0002-5839-7114
#
These authors contributed equally.
https://doi.org/10.1016/j.identj.2024.03.009
0020-6539/Ó2024 The Authors. Published by Elsevier Inc. on behalf of FDI World Dental Federation. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
international dental journal 74 (2024) 1089−1101
are burgeoning, especially in China’s major urban
agglomerations.
3
Despite noticeable growth in the number of dental clinics
becoming a key characteristic of development in the Chinese
dental care market, this rapidly expanding mode still faces
significant challenges. Given the significant disparities in
socioeconomic conditions and healthcare accessibility
among regions in China, a striking disparity in oral health sta-
tus persists among individuals differing in geographical
location.
4,5
In China, unequal access to oral healthcare is
influenced by the uneven distribution of dental institutions
and dental professionals, limited national programs, and
under-resourced public facilities lacking essential materials
and equipment.
6,
Chronologically, wealth distribution and sociodemo-
graphic development induced by urban development could
significantly influence the quality and provision of dental
care services. Urbanization-driven improvements in trans-
portation, public facilities, sociodemographic factors, and
increased human activity in medical services can signifi-
cantly alter the spatial distribution of dental clinics.
7
Conversely, a strategically planned distribution of dental
institutions can enhance access to dental services and pro-
mote equality in the allocation of health resources.
8
Furthermore, the relationship between public and private
medical care institutions needs to be carefully considered. In
China, public health institutions follow a rigid 3-tier struc-
ture, comprising primary, secondary, and tertiary hospitals,
based on the level and range of healthcare services they
offer.
9
Tertiary hospitals, situated in urban centres, provin-
ces, or at the national level, specialising in advanced medical
resources and expertise. Primary and secondary hospitals
deliver primary care services, often catering to rural or com-
munity settings with limited resources and specialist avail-
ability.
10
The dental departments of tertiary hospitals in
China commonly face overcrowding and struggle to meet the
high demand for services due to inadequate treatment
quality in many secondary and primary hospitals, where
highly technical procedures like cavity fillings may not be
available.
11
As a crucial part of the healthcare system, private dental
practices can alleviate the strain on public services caused by
overcrowding and fill this service gap by catering to diverse
needs and at various price points, providing accessible
options. For a long time, challenges such as insufficient
patient awareness, restricted coverage from the Bureau of
Medical Insurance, and insufficient backing from regional
administrations have impeded the growth of private dental
clinics in China thus sustaining the dominance of public ter-
tiary dental hospitals in China’s healthcare system.
12
How-
ever, following a healthcare reform phase in 2009, China’s
private health sector underwent significant development.
13
These reforms, in addition to a growing demand for health-
care, resulted in the number of private sector institutions
exceeding twice the share of public sector institutions.
14
Typ-
ically, China’s government leads the planning for the place-
ment of public hospitals. The National Health Commission’s
“Guiding Principles for the Planning of Medical Institutions
(2021-2025)” emphasizes conducting a comprehensive assess-
ment, which considers factors such as local population
demographics, resident influx, hospital admissions, bed
occupancy rates, and the corresponding service radius. Con-
versely, dental practitioners are afforded greater flexibility in
selecting sites for their dental clinics; however, the main
determinants remain unclear.
Current research mainly focuses on exploring the changes
caused by urban development on dental care services or qual-
ity,
15
as well as the spatial accessibility of dental care serv-
ices.
16
However, little is known about dental practice spatial
characteristics and the relationship between dental institu-
tions in China. As urbanization progresses, regional develop-
ment varies. This influences the spatial distribution, social
and productive activities of residents, and healthcare
institutions’ layouts.
5
Few evidence-based studies evaluated
the correlations between those changes from urban develop-
ment and dental practice spatial patterns.
Spatial autocorrelation can identify spatial characteristics
and transformation through clustering techniques, map com-
parisons, and other common spatial techniques like regres-
sion analysis to develop prediction models.
17
Medical
institutions located close to each other tend to share similar
spatial evolutionary patterns, because they share similar
socio-geographic environments, similar quality of dental
workforce and practitioners, and are also often connected by
the number and characteristics of competitors and customer
groups.
18
Private dental clinics can efficiently share similar
physical spaces and resources to streamline operations,
reduce operating costs, boost competitiveness, and pursue
opportunities for business expansion. Therefore, the distribu-
tion of factors influences private dental clinics, suggesting
potential relationships through spatial distribution.
Due to the unbalanced distribution of dental practices,
accurately identifying the spatial distribution of dental practi-
ces and their impacting factors, as well as figuring out the
possible relationship between dental institutions are crucial
for future dental service strategies and medical resource allo-
cation. In this study, we chose 11 major Chinese cities as our
study area. The proposed method combines spatial autocor-
relation and regression analysis to examine the evolution
process of spatial characteristics and the impacting factors of
private dental practices.
Methods
Study setting
This study is on 11 major cities in China, including Beijing,
Tianjin, Shanghai, Nanjing, Hangzhou, Guangzhou, Shenz-
hen, Chengdu, Chongqing, Wuhan, and Xi’an. Figure 1(A)
shows the scope of our research, encompassing eleven major
cities. Furthermore, our selected cities span across the east-
ern, middle, and western regions (the eastern cities are typi-
cally richer, while the western ones are poorer), as well as
different development levels of urban agglomerations in
China, which includes the Beijing-Tianjin-Hebei Urban
Agglomeration (BTH), the Yangtze River Delta (YRD), the Pearl
River Delta (PRD), the Chengdu-Chongqing City Group (CC),
and Central China. They are broadly representative of the
average population within China as key megalopolises as
1090 liu et al.
Fig. 1 –(A) Study area: eleven major cities in China. This figure contains the distribution of our study areas. Note that this
figure does not include a full map of China because our study area focuses on China’s major cities. (B) Proportion of public
and private dental clinics for 11 major cities in China. This figure demonstrates the proportion of public hospitals to the total
medical institutions in the city, evaluated for their developmental progress. The red bar symbolizes the percentage of tertiary
hospitals, the blue bar represents secondary hospitals, and the orange bar denotes primary hospitals. The data used to gen-
erate this figure is for the year 2022.
dental clinic distribution and impact factors 1091
political, economic, and cultural regional centres, drawing
significant attention amidst China’s urbanization. Major cit-
ies boast sophisticated 3-tier healthcare systems and well-
established dental care markets compared to rural or remote
urban areas. China’s ongoing healthcare reforms prioritize
major cities known for advanced development and higher
GDP as pilot sites. Given the dynamic nature of private dental
practices, which are sensitive to market fluctuations, territo-
rial economics, and urban agglomeration development, ana-
lyzing major cities facilitates the exploration of geographical
patterns, temporal trends, and industrial disparities within
China’s private dental sector. The selected time-period from
2016 to 2022 coincides with the thriving dental care market,
deepening medical reforms, and accelerated urban develop-
ment in these major Chinese cities.
Information extraction and indicator selection
To construct the research data, we first extracted the geo-
graphic information data and point-of-interest (POI) data
from the AutoNavi map. AutoNavi, a top mapping app in
China, boasts a user base exceeding 60 million. It stands out
for its detailed Points of Interest (POI) data and accurate pub-
lic transportation information, highlighting its superior navi-
gational accuracy and comprehensive urban mobility
solutions.
19
Our extracted AutoNavi map contains more than
1 million POI for each city each year. Based on the extracted
data, we categorized them into various types, described in
Table 1. In addition, we obtained 1,716,579 second-hand
house transactions from Lianjia, which is the largest real
estate brokerage in China, from 2016 to 2022 in our selected
cities. We constructed a grid based on the data of Lianjia’s
transactions in that city. We chose Lianjia as the representa-
tive of real estate transaction data due to its status as China’s
largest brokerage company, handling a significant percentage
of second-hand house transactions.
20
While Lianjia provides
accurate and transparent data, some cities (such as Beijing,
Chengdu, Xi’an, and Shenzhen) lack publicly available data
for 2021-2022 due to government controls. In these cases, we
utilized 2020 transaction prices as proxies for 2021 and 2022,
given the stability of house prices amidst COVID-19 and gov-
ernment price regulations from 2019 to 2022. Healthcare facil-
ities data for 2022 was obtained from the China Hospital
Information Management Association’s website, reflecting
stable hospital numbers during our study period. To focus on
economically active areas, we excluded grids without hous-
ing transactions, ensuring our sample accurately reflects
urban characteristics. We counted the number of different
types of POIs within that grid and then calculated the average
price and average number of houses sold in the area. Also, we
used the regional average atmospheric fine particulate matter
(PM2.5) concentration as a regional air quality indicator,
sourced from the Atmospheric Composition Analysis Group
at Washington University in St. Louis.
21
Table 1 – The statistical summary.
Without private dental clinics With private dental clinics
Variables Mean STD Mean STD Mean difference
Dental clinics 0 0 2.605 2.238 2.605***
Tertiary public hospitals 0.033 0.201 0.124 0.432 0.090***
Secondary public hospitals 0.053 0.305 0.220 0.644 0.167***
Primary public hospitals 0.020 0.179 0.096 0.404 0.076***
Housing price 3.301 2.152 4.299 2.592 0.998***
Population 0.700 0.950 1.696 1.753 0.996***
Night time light 25.09 12.83 35.29 12.93 10.20***
Pm2.5 45.82 13.05 43.67 13.27 2.151***
Electronic stores 0.136 0.806 0.571 1.994 0.435***
KTV number 0.468 1.355 1.953 3.719 1.485***
House age 10.97 9.892 16.75 10.17 5.777***
Hotel number 0.322 0.929 1.188 2.068 0.865***
Shopping mall 0.409 1.275 1.820 3.210 1.411***
Museum number 0.0653 0.356 0.210 0.684 0.145***
Elderly activity areas 0.110 0.403 0.281 0.725 0.170***
Kindergarten 1.389 1.741 3.580 2.749 2.191***
Middle school 0.295 0.669 0.775 1.141 0.479***
Primary school 0.384 0.737 1.050 1.227 0.666***
Restaurant 0.311 1.064 1.504 3.056 1.192***
Supermarket 0.382 0.834 1.226 1.529 0.844***
Green ratio 0.269 0.159 0.290 0.116 0.0200***
Number of buildings 33.91 73.66 23.33 41.76 10.58***
Elevator ratio 0.378 0.352 0.379 0.215 0.001***
Subway station 0.0980 0.309 0.273 0.493 0.175***
Floor ratio 1.768 1.443 2.702 12.48 0.934***
Number of residences 1220 1333 1365 1146 145.6***
Park number 0.541 1.335 1.196 1.707 0.655***
STD, standard deviation.
The statistical significance is under a 1% confidence interval.
Clustered standard errors are presented in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
1092 liu et al.
To accurately measure each grid’s level of economic devel-
opment, we collected nighttime lighting data as a proxy vari-
able. In addition, we directly measured the population
density of each region by utilizing data on population from
WorldPop,
22
which offered estimates of population per
(1000 m * 1000 m) grid. To obtain the addresses of public hos-
pitals in China more accurately, we further obtained the
detailed names and addresses of tertiary hospitals, secondary
hospitals, and primary hospitals from the government’s offi-
cial website, and queried the detailed latitude and longitude
of these hospitals by back-checking coordinates on the Auto-
Navi map. After performing the previously mentioned steps,
we finally constructed our research data, and a summary of
the statistics is shown in Table 1.
Spatial autocorrelation analysis
Spatial autocorrelation analysis is used to determine
whether the key variable is spatially correlated. Moran’s I
index quantifies spatial autocorrelation, ranging from “-1”
to “1”. Higher positive values indicate clustering of high
observations, while more negative values suggest cluster-
ing of low ones. Values near “0” imply negligible spatial
autocorrelation, with high and low values dispersed ran-
domly.
23
The Moran’s I index can be calculated using the
following equation:
I¼N
W
PiPjwij xixðÞxjx
PixixðÞxixðÞ
where N is the numeral of spatial one indicator by i and j; x is
the variable of the connector; x is the mean of x; wij is the spa-
tial weights matrix, and Wis the sum of all.
In our analysis, Python is initially used to compute the
Moran’s I spatial autocorrelation index, which evaluates the
qualities and locations of spatial features relative to their spa-
tial location.
24
A typical test statistic for spatial autocorrela-
tion is Moran’s I. Spatial autocorrelation can occur even
when the values are randomly distributed in space, even
though there is no correlation between the values. Global
Moran’s I index captures the correlation of each variable
across several locations by spatial autocorrelation.
25
Random effect model (RE)
Since the number of different level public hospitals remained
constant during our research period, we decided to use an RE
model to capture the general effect of different level public
hospitals on private dental clinics. RE model assumes that
the observed predictors in the model are not related to indi-
vidual specific random error.
26
RE model is a useful tool for
predicting time-invariant variables and serves as a bench-
mark model.
26
The equation can be listed as follows:
Yit ¼b0þbXit þɛit
where Yit is the number of dental practices in year tand grid i,
Xit is a set of control variables indicated in Table 1 in year t
and grid i,ɛit is the random unobserved random error and b0,
bare the coefficients. To make our estimation robust, we
cluster the standard deviation by each grid level to rule out
unobserved factors.
Fixed effect model (FE)
The RE model may not be sufficient to get a robust result
because it does not take the unobservable unit-specific and
time-specific confounders into account. Therefore, we should
adopt an FE model to control individually fixed effects allow-
ing us to identify more accurate and year-fixed effects to rule
out the time-specific factors.
27
In our study period, the num-
ber of public hospitals remained stable. Therefore, these vari-
ables will be automatically removed by the FE model because
these time-invariant effects are already included in the
model. The model can be described below:
Yit ¼b0þbXit þciþutþɛit
where ciis the unit-specific fixed effect and utis the time-spe-
cific fixed effect and the other remains the same.
Spatial lagged model (SLR)
The structure of the SLR model can be outlined as follows
28
:
Yit ¼b0þrWYit þbXit þciþutþɛit
where ris the coefficient, Wis the spatial weighted matrix
and Yit is the neighboring dependent variables, and other
notations remain the same with the FE model. The weighted
matrix is identical to the weighted matrix in calculating
Moran’s I index.
Result
Statistic summary
Figure 1(B) represents the percentage of different levels of
public hospitals, as well as private dental clinics. We can see
that the number of private dental clinics exceeds 50% in each
of the cities studied. The distribution of public hospitals is
heavily concentrated in Beijing and Shanghai, the two largest
cities in China, indicating uneven resource allocation. This
gap can potentially be addressed by private dental clinics.
Figures 2 (A-K) illustrate clinic dispersion based on 2022
data, depicting the distribution of dental clinics at the admin-
istrative district level. For comparison purposes, 2022’s POI
data are used for all other analyses.
To better understand the spread of these clinics, we
employed a standard deviational ellipse. In addition to
showcasing the distribution of dental clinics, we incorpo-
rated the placement of secondary hospitals (indicated by
green points) and tertiary hospitals (denoted by red points)
into the graph.
Upon evaluation, the graph shows a significant clustering
of dental clinics in city centers. There is also a noticeable dif-
fusion of clinics on the city outskirts. Considering the location
of the secondary and tertiary hospitals, it is apparent that
they are predominantly situated in urban and peri-urban
regions. The graphs reveal that Beijing and Shanghai, China’s
largest cities, have more medical resources compared to
other highly developed urban areas. In these areas, where
public resources are comparatively limited, private dental
clinics are more widespread.
dental clinic distribution and impact factors 1093
In Figure 3 (A-K), given our research samples were con-
structed using a 1 km grid, we subsequently created a corre-
sponding graph. The colour scheme in this graph represents
housing prices, which acts as an indicator of societal develop-
ment and affluence within these regions. Additionally, the
locations of dental clinics have been incorporated into the graph
as data points. In contrast to Beijing and Shanghai, which have
superior public resources, medical institutions in emerging cities
are mainly found in secondary hospitals, with limited presence
in primary and tertiary hospitals. This suggests that healthcare
Fig. 2 –Distribution of private dental practices in 11 major cities in China. The legend shows the number of dental clinics in
(A) Beijing, (B) Tianjin, (C) Shanghai, (D) Nanjing, (E) Hangzhou, (F) Guangzhou, (G) Shenzhen, (H) Chengdu, (I) Chongqing,
(J) Wuhan, and (K) Xi’an. Red points are tertiary hospitals and blue points are green points are secondary hospitals. The stan-
dard ellipses are calculated based on the distribution of private dental clinics in each administrative region. The data used to
generate this figure is for the year 2022.
1094 liu et al.
facility availability needs further improvement in these areas.
When examining the distribution of dental clinics against our
grid overlay, we observe that our grids cover most of the dental
clinics in the research areas. Dental clinics tend to cluster in
areas with high housing prices within each metropolis, due
largely to the affluent nature of these areas and their increased
ability to spend on dental services. Moreover, we note the lack
of dental clinics on the outskirts of these major cities, a factor
that could contribute to the discrepancy in accessing dental
resources across these metropolises.
Table 1 provides a statistical summary of the variables
used in our regression analysis. We initially subdivided the
Fig. 3 –Distribution of study sample based on 1km*1km grid. The legends indicate the housing price in each 1km*1km grid,
which is a good indication of the region’s development, and the points indicate the distribution of dental clinics in (A) Beijing,
(B) Tianjin, (C) Shanghai, (D) Nanjing, (E) Hangzhou, (F) Guangzhou, (G) Shenzhen, (H) Chengdu, (I) Chongqing, (J) Wuhan, and
(K) Xi’an, respectively. The data used to generate this figure is for the year 2022.
dental clinic distribution and impact factors 1095
sample into two categories: grids with dental clinics and grids
without dental clinics. Subsequently, we reported the mean
and standard deviation for these variables, respectively.
Finally, we computed the t-statistics and the corresponding
p-value to ascertain the significance of the difference
between these groups.
From the table, it is evident that many differences
between the two samples are statistically significant at a 1%
level. This suggests a tendency for dental clinics to cluster in
more developed areas. Furthermore, we note that in regions
with dental clinics, the ratio of the population to dental clin-
ics is 0.65. This implies that for every kilometer, approxi-
mately 6500 individuals have access to a dental clinic.
Regression results
Table 2 column 1 shows the result of the Random Effect
Model (RE). We find that the number of tertiary hospitals is
negatively correlated with the number of private dental prac-
tices. The reason is that tertiary hospitals are general hospi-
tals providing comprehensive, high-quality, and specialized
medical services, so residents in these areas do not have high
demands for private clinics, leading to a corresponding
decrease in the number of private clinics in these areas. Addi-
tionally, we have discerned that there is no significant corre-
lation between the number of secondary and primary
hospitals and the quantity of private dental clinics. This
observation implies that those basic healthcare units with a
smaller influence lack a good connection to dental clinics and
lack competition or opportunities for collaboration among
these healthcare institutions. Furthermore, we have observed
that dental clinics are often localized in areas with high popu-
lation density and steep housing prices. This aligns with the
understanding that these clinics generally show a preference
for highly developed areas.
Table 2 column 2 presents the Fixed Effects (FE) model
result. On introducing individual fixed effects, we can no lon-
ger observe public and private healthcare institutions’ rela-
tions, but additional controls could take more unobservable
factors into control. As is shown in the table, a larger impact
of housing prices and population density on private dental
clinics becomes apparent, suggesting that the RE model tends
to underestimate these clustering effects.
Furthermore, we use Moran’s I index to test if there are
geospatial clustering effects. We first constructed a K-Nearest
Neighbors spatial weighted matrix (k = 5), after which we cal-
culated Moran’s I index to test the spatial autocorrelation.
The test result was significant at the 1 percent (p = 0.000) level
and the index value was 0.72, indicating that the spatial auto-
correlation is significant, and the clustering effect is large.
Drawing from the insights provided by Moran’s I, such
increases in dental practices have manifested in a consider-
able spatial aggregation. This pattern indicates discernable
regional variations in this progression. A striking contrast is
evident between a city’s central area and its outer districts,
with the clustering effect becoming more pronounced across
regions over time. The spatial distribution of dental practices
warrants attention, necessitating the execution of regression
analysis due to the escalating and strengthening geographic
correlation. Moran’s I indices suggest potential bias in RE and
FE models, it is crucial to incorporate spatial elements when
further assessing the development.
When comparing the RE and FE models, we discover
numerous unobservable patterns that have an impact on our
model. Consequently, we should incorporate individual fixed
effects into our considerations. Furthermore, based on the
Table 2 – Regression results.
Dependent variable: number of private dental clinics
Variable RE FE SLR SLR-lagged by 1 y
Tertiary hospital 0.1277*
(0.0661)
Secondary hospital 0.0677
(0.0420)
Primary hospital 0.0530
(0.0681)
Housing price 0.0164* 0.0412*** 0.0439*** 0.0505***
(0.0097) (0.0123) (0.006) (0.006)
Population 0.1061*** 1.0286*** 0.8804*** 0.9429***
(0.0170) (0.1303) (0.041) (0.050)
Night time light 0.0113*** 0.0085*** 0.01090*** 0.0113***
(0.0012) (0.0016) (0.001) (0.001)
Wr0.3780*** 0.3919***
(0.005) (0.006)
POI control Y Y Y Y
Other Control Y Y Y Y
Year Fixed Effect Y Y Y Y
Individual Fixed Effect Y Y Y
All regressions are clustered at each grid fixed effect.
Clustered standard errors are presented in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
Y indicates that the control variables are included in the model.
In column 4, the regressors lagged one year, but the regressors included in the regression model are the same.
Note: Robust standard errors are in parentheses.
1096 liu et al.
computation of Moran’s I index, we deduce that private den-
tal practices tend to congregate—this clustering effect results
in spillovers, which can significantly influence our estima-
tions. To manage the confounding effect of the spatial auto-
correlation of the explanatory variables, we utilize the Spatial
Lagged Model (SLR), as proposed by Elhost.
28
The results are
reported in Table 2 column 3.
The result shows that ris significantly different from zero,
indicating the presence of autocorrelation in the SLR model.
The results suggest that given all other variables remain con-
stant, a population increase of ten thousand individuals per
kilometre corresponds to a rise of 0.88 in the number of den-
tal clinics. Additionally, with other factors unchanged, an
increment in the intensity of nighttime lights correlates with
a 0.01 increase in the count of dental clinics. Furthermore,
presuming all other controls remain the same, an increase of
1000 yuan in housing prices is associated with a 0.439
increase in the number of dental clinics.
To assess the robustness of our model, we incorporated a
lagged version of all independent variables and estimated the
outcomes using the SLR model. The findings are presented in
Table 2 column 4. The results are identical to those of our
original SLR model, demonstrating the stability and robust-
ness of our model over time.
Heterogeneous test results
To assess the heterogeneity in our regression more compre-
hensively, we bifurcate the sample into four distinct sections
based on city clusters. These urban agglomerations include
the BTH, YRD, PRD, and CC. The samples are divided into
these four components, employing the SLR to project the
model. The derived outcomes of this analysis are reported in
Table 3 columns 1 to 4.
When it comes to housing prices, there is a positive corre-
lation observed across all major cities. However, we find a
more pronounced influence in the PRD and CC, where regions
of higher value typically have superior clinical resources.
Additionally, population size significantly impacts dental
clinic distribution, particularly in cities like BTH and PRD.
Intriguingly, we find a negative correlation between popula-
tion density and this trend in CC. This phenomenon is in line
with the intuition that Chengdu and Chongqing are emerging
cities experiencing rapid population growth, a factor disrupt-
ing the preservation of property clustering. Besides, the influ-
ence of nighttime light, a measure of economic development,
demonstrates a nearly uniform impact across all these met-
ropolises, suggesting a constant relation with economic
development. Lastly, we observe that the spatial lag item is
significant, meaning that our SLR model is robust.
While we have controlled for the influence of public hospi-
tals on private dental clinic distribution, note that public hos-
pitals are strategically placed in densely populated areas.
This locational strategy could inadvertently affect the opera-
tional behaviour of private dental clinics. Due to the heteroge-
neity introduced by the selective establishment of public
institutions across regions, our sample needs to be subdi-
vided into areas with and without public hospitals. This sub-
division allows for a more nuanced analysis of economic
dynamics, particularly in the dental care industry. SLR model
results in Table 3 columns 7 to 8 indicate that in areas lacking
public hospitals, real estate price fluctuations significantly
impact dental clinic presence, reflecting private clinics’ sus-
ceptibility to market dynamics. Conversely, in areas with
public hospitals, population size has a stronger impact on
dental clinic distribution, validating the complementary role
of private clinics in public healthcare institutions.
Temporal effect test results
In Figure 4, the impact of the 2020 novel coronavirus (COVID-
19) pandemic on private clinics is depicted. These effects can
be attributed to a confluence of factors, including stringent
control measures implemented by the Chinese government,
Table 3 – Heterogeneous test results.
Dependent variable: number of private dental clinics
Variable BTH YRD PRD CC Before 2020 2020-2022 Without public With public
Housing price 0.0287** 0.0259*** 0.0671*** 0.2519*** 0.0431** 0.0037 0.0637*** 0.0438***
(0.0135) (0.0072) (0.0172) (0.0385) (0.0057) (0.0094) (0.0224) (0.0056)
Population 1.1675*** 0.5164*** 1.6507*** 0.3660** 0.3684*** 0.8137*** 0.7594*** 0.9591**
(0.0745) (0.0533) (0.1320) (0.1622) (0.0529) (0.0864) (0.1025) (0.0475)
Night time light 0.0107*** 0.0082*** 0.0116*** 0.0126*** 0.0025*** 0.0033*** 0.0125*** 0.0107***
(0.0016) (0.0012) (0.0033) (0.0021) (0.0009) (0.0011) (0.0028) (0.0008)
r0.2565*** 0.2620*** 0.3985*** 0.3095*** 0.3190*** 0.2338*** 0.3874*** 0.3573***
(0.0121) (0.0108) (0.0133) (0.0130) (0.0076) (0.0093) (0.0136) (0.0058)
POI control YYYY Y Y Y Y
Other Control YYYY Y Y Y Y
Year Fixed Effect YYYY Y Y Y Y
Individual Fixed Effect YYYY Y Y Y Y
All regressions are clustered at each grid fixed effect.
Clustered standard errors are presented in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
Y indicates that the control variables are included in the model.
The term ’Without public’ stands for the model not adjusted for the number of public hospitals, while ’With public’ represents the model adjusted
for the number of public hospitals.
Note: Robust standard errors are in parentheses.
dental clinic distribution and impact factors 1097
which restricted the operational capabilities of most private
clinics, and the economic downturn experienced by China’s
economy, resulting in the closure of several clinics. Given
this context, it is prudent to categorize the dataset into two
temporal cohorts: pre-2020 and post-2020. This division will
facilitate a more structured regression analysis, enabling an
examination of the pandemic’s impact on the private health-
care sector. The results are reported in Table 3 columns 5 to 6.
Before the onset of the pandemic, real estate pricing sig-
nificantly influenced the distribution of dental clinics, sug-
gesting that higher property values were positively correlated
with the presence of these healthcare services. However, the
post-pandemic landscape reveals a noticeable decline in the
role of real estate pricing in determining clinic distribution,
which may be indicative of a narrowing gap in medical acces-
sibility among different socioeconomic groups. Furthermore,
when examining the effect of population density on the dis-
tribution of dental clinics, it is observed that although its
influence remains statistically significant in both pre-and
post-pandemic periods, the magnitude of its impact was
comparatively lower before the pandemic. This trend sug-
gests a shift in China’s urban development dynamics, with
city facility allocation increasingly aligning with population
flow patterns rather than being predominantly driven by
economic development indicators. This evolution reflects a
nuanced understanding of the interplay between public
health infrastructure and urban planning in the context of
socio-economic transformations.
Discussion
To our knowledge, this study is the first to investigate the
spatial distribution and influencing factors of Chinese dental
clinics and the first to examine private dental practices in
major cities across the country.
One of the key findings is showing the factors associated
with private dental practice distribution during urban devel-
opment. Our results showed that regional development, the
population number, and the size and activity of the popula-
tion were significantly positively associated with the number
of dental practices in the city, resulting in a clustering pat-
tern. Regional variations in these factors can impact the spa-
tial distribution of private dental practices. For instance, as
regional economies grow and population activity increases,
the dental care market flourishes and plays a positive role in
the economies of scale. This, in turn, promotes the influenc-
ing factors, indicating a positive correlation between
Fig. 4 –The number of private dental clinics from 2016 to 2022. In this legend, Different colors (shapes) correspond to distinct
cities, with each color or shape representing a specific city. This data differs from the data from our regression and statistical
analyses because they remove some of the sample of suburban-distributed dental clinics. In contrast, here we utilize a city-
wide sample for our analysis, which helps us to look at overall trends. Moreover, the trend observed in this data aligns with
our constructed dataset, and the prevalence of private dental clinics in this dataset is more representative.
1098 liu et al.
economic growth and the expansion of dental care institu-
tions. As China’s health reforms deepen, notably with the
implementation of the 2030 Health Plan, the dental health
sector shows continuous improvement, potentially driving
the development of private dental clinics.
2
A previous study,
which analyzed the growth of Indian cities over three succes-
sive censuses, indicated an escalating need for improved
healthcare and infrastructure development strategies.
29
These strategies are essential for ensuring the long-term via-
bility of urban clusters.
Our study further revealed that clustering patterns among
metropolises at different stages of development are likely to
vary. We observed that this clustering effect becomes more
evident as a city undergoes substantial development. Such a
distribution pattern demonstrates a positive spatial external-
ity and produces a scale impact, potentially leading to
improved accessibility to dental health care which leads to
fewer absences from work due to oral diseases and conse-
quently improves economic efficiency but potentially exacer-
bates inequality.
30
Empirical findings highlight that the
spatial distribution of dental care facilities is influenced not
only by regional economic situations but also by additional
factors such as demographic distribution and population
activity. Typically, a variety of factors, including access to
healthcare, economic constraints, and needs for dental care
influence people’s utilization of dental services.
31
These ele-
ments, in turn, might differ between geographical regions
and population groups. Moreover, despite urbanization’s
progress, developmental differences persist between and
within cities. Our study discovered a positive association
between housing prices and population density after adjust-
ment for covariates, coupled with the business district loca-
tion, suggesting that dental clinic site selection preference of
highly developed areas.
The relative paucity of healthcare provision in low eco-
nomic level areas is already widely documented, and many
studies suggested that there is an imbalance of dental clinics
and medical resources when factoring in the local economic
development and urban development, as well as population
density imbalances.
32,33
The location of profitable dental
practices, the plans for government investment in medical
resources, and the influence of public medical institutions
are all affected by these factors. Regarding this concern, a
recent study demonstrated that the market-driven location
of dental clinics causes an imbalance in the distribution of
dental and medical resources between large and small
municipalities, resulting in better dental accessibility in Japa-
nese municipalities with larger populations than in munici-
palities with smaller populations.
34
In China, various studies
have shown that there is still a persistent unfair distribution
of medical resources. The fairness of medical resources in
each province is generally low due to the different economic
development levels in eastern, central, and western China as
well as the imbalance of urban and rural development.
32
Con-
siderations of these findings suggest that the significant spa-
tial deviation in the distribution of dental practices is leading
to inequity and low efficiency in the allocation of dental care
resources, as well as poor accessibility to dental care.
Another crucial finding that we explored is the relation-
ships and interactions between public and private dental
institutions, which showed the number of tertiary hospitals
has a negative correlation, while secondary and primary hos-
pitals have no significant effect on the number of private den-
tal practices. The relationship we observed is in line with this
previous evidence collected from many studies; for example,
people in underprivileged metropolitan regions frequently go
to private clinics for primary care because there are few pub-
lic primary care facilities as a consequence of tertiary care
hospitals being out of reach, expensive, and in great
demand.
35
In line with this, a previous study pointed out that
vulnerable groups prefer to choose public hospitals for their
cheaper fees.
36
While the fact that tertiary hospitals are rela-
tively more difficult to access than dental clinics, tertiary hos-
pitals still have a strong attraction to the surrounding
population when it comes to the need for dental treatment,
considering that they can provide comprehensive, high-qual-
ity, and specialized medical services.
37
Thus, our findings
provide a better understanding of the relationship between
tertiary hospitals and private dental clinics in China, which
shows clear implications for future site selection planning. A
study on resource distribution and influencing factors
between public and private healthcare institutions in China
suggests that while they compete in terms of numbers, they
also exhibit a symbiotic development in shared medical
resources.
38
This dual effect of public hospitals stems from
their shared resources and collaboration with private institu-
tions. In major cities, the hierarchical diagnosis and treat-
ment model is advancing. Current development of physician
practise in multiple sites would encourage many private clin-
ics to collaborate with renowned doctors from tertiary hospi-
tals to attract patients and enhance their reputation.
39
In the context of temporal dynamics, the COVID-19 pan-
demic served as an exogenous shock to the economic system.
The empirical evidence indicates that while the onset of the
pandemic exerted adverse effects on dental clinics, it para-
doxically facilitated their evolution over time.
40
This phe-
nomenon can be attributed to the economic recovery after
the initial downturn, which engendered a redistribution of
dental clinics favouring areas with higher population densi-
ties.
41
Such a redistribution augmented the synergistic rela-
tionship between private dental clinics and the public
healthcare system, thereby steering the market towards a
more robust developmental trajectory.
Given the constraints related to data accessibility and the
representativeness of cases, this analysis concentrates on 11
major cities within China. It must be acknowledged that
these cities do not entirely encapsulate the diverse patterns
of dental practice observed across the nation. Nonetheless, a
marked contrast exists between the comprehensive dental
services offered in these major urban centres and the basic
dental care available in rural and less urbanized areas. The
latter predominantly relies on primary and secondary public
healthcare facilities, which frequently encounter capacity
limitations. Moreover, China’s tiered dental healthcare sys-
tem remains underdeveloped. As a result, there is a growing
trend of patients gravitating towards tertiary hospitals or
superior dental clinics located in urban locales, leading to a
reduction in patient numbers at rural dental facilities and
potentially precipitating a dearth of dental professionals in
these regions.
dental clinic distribution and impact factors 1099
Disparities in economic development, urbanization, and
uneven healthcare reform implementation exacerbate
challenges in rural and remote areas. These factors hinder
the growth of local private dental practices and contribute
to the uneven spatial distribution of dental clinics. Conse-
quently, strategies for allocating private dental clinics in
rural and remote regions differ significantly from those in
major urban centres. Future research will explore the dis-
tribution and expansion of private dental clinics in rural
and remote urban areas, as well as assess nationwide dis-
parities and accessibility.
Furthermore, it is acknowledged that the current analy-
sis ignores existing access to public and private dental
services, which is influenced by variables such as hospital
numbers, accessibility, and patient preferences. Address-
ing these variables requires acquiring micro-level data on
patients’ choices and conducting further analysis in the
future.
We regard spatial autocorrelation and regression analy-
sis as broadly applicable approaches that can be used in
studying the distribution and influencing factors. Mapping
dental institution distribution and monitoring accessibility
can aid policymakers in reallocating resources and incen-
tivizing private sector involvement.
42
Practitioners and
stakeholders should consider the customer base and com-
petitive environment in practice site selection. In addition,
the study provides evidence-based guidance for the eco-
nomic development of the oral healthcare industry that
helps predict the location and market development trend
of dental practice with more potential and vitality.
Conclusion
This study examined the spatial distribution of private dental
practices in major Chinese cities, revealing distinct clustering
patterns. A positive correlation between regional develop-
ment and dental practice prevalence underscores the need to
address urban inequality as economic development pro-
gresses. Significant clustering and access inequalities were
observed in highly developed regions. Private practices are
increasingly complementing public services due to growing
demand, especially in these areas. Further research is needed
to explore causal mechanisms and validate findings across
broader regions in China.
Conflict of interest
All authors disclosed no relevant relationships.
Acknowledgements
This study was supported by the Education and Teaching
Research Project of Kunming Medical University (2023-JY-Z-
03). The authors gratefully acknowledge Dr. Hazem Abbas for
his valuable suggestions, Mr. Dashiell Lee for his language
editing support, and the constructive comments provided by
the anonymous referees.
Author contributions
Weihua Guo and Guoying Deng conceived and designed the
structure of this manuscript and revised the paper. Pengbo
Liu, Xuyuan Zhang: acquisition of data, analysis, and inter-
pretation, drafting this paper. All authors contributed to the
article and approved the submitted version.
Data availability statements
All data associated with this study are presented in the paper.
The supplementary materials that support the findings of
this study are openly available on GitHub at: https://github.
com/Sergio666zxy/Dental-Clinic-Distribution-and-Impact-
Factors
Supplementary materials
Supplementary material associated with this article can be
found in the online version at doi:10.1016/j.identj.2024.03.009.
REFERENCES
1. WHO discussion paper: draft global oral health action plan
(2023-2030) World Health Organization; Geneva, Switzerland;
2022. Available from: https://cdn.who.int/media/docs/
default-source/ncds/mnd/eb152-draft-global-oral-health-
action-plan.pdf?sfvrsn=ecce482e_4 Accessed September 16,
2023.
2. Zhou X, Xu X, Li J, Hu D, Hu T, Yin W, et al. Oral health in
China: from vision to action. Int J Oral Sci 2018;10:1. doi:
10.1038/s41368-017-0006-6.
3. Sun XY, Yuan C, Wang XZ, Wang X, Feng XP, Tai BJ, et al.
Report of the National Investigation of Resources for Oral
Health in China. Chin J Dent Res 2018;21:285–97. doi: 10.3290/
j.cjdr.a41087.
4. Kamil W, Kruger E, Tennant M. Utilisation of dental services
of older people in Australia: an economic explanatory model
based on cost and geographic location. Geriatrics (Basel)
2021;6:102. doi: 10.3390/geriatrics6040102.
5. Jakovljevic M, Chang H, Pan J, Guo C, Hui J, Hu H, et al. Suc-
cesses and challenges of China’s health care reform: a four-
decade perspective spanning 1985-2023. Cost Eff Resour. Alloc
2023;21:59. doi: 10.1186/s12962-023-00461-9.
6. Liu M, Yun Q, Zhao M, Chen W, Zhang H, Hou W, et al. Associ-
ation of siblings’ presence and oral health-related quality of
life among children: a cross-sectional study. BMC Oral Health
2021;21:153. doi: 10.1186/s12903-021-01526-y.
7. Liu X, Wei M, Zeng J. Simulating urban growth scenarios
based on ecological security pattern: a case study in Quanz-
hou, China. Int J Environ Res Public Health 2020;17:7282. doi:
10.3390/ijerph17197282.
8. Kempster C, Gray J, Burman K. Student-driven partnerships
enhancing oral health literacy. Health Promot J Austr 2021;32
(Suppl 1):98–103. doi: 10.1002/hpja.376.
9. Liang C, Zhao Y, Yu C, Sang P, Yang L. Hierarchical medical
system and local medical performance: a quasi-natural
experiment evaluation in Shanghai, China. Front Public
Health 2022;10:904384. doi: 10.3389/fpubh.2022.904384.
10. Chao J, Lu B, Zhang H, Zhu L, Jin H, Liu P. Healthcare system
responsiveness in Jiangsu Province, China. BMC Health Serv-
ices Res 2017;17:1–7.
1100 liu et al.
11. Zhu J, Li J, Zhang Z, Li H, Cai L. Exploring determinants of
health provider choice and heterogeneity in preference
among outpatients in Beijing: a labelled discrete choice exper-
iment. BMJ Open 2019;9:e023363. doi: 10.1136/bmjopen-2018-
023363.
12. Xue X. Annual report on China’s private hospitals develop-
ment. Beijing, China: Social Sciences Academic Press; 2017
2018.
13. Zhang X, Zimmerman A, Zhang Y, Ogbuoji O, Tang S. Rapid
growth of private hospitals in China: emerging challenges
and opportunities to health sector management. Lancet Reg
Health West Pac 2024;44:100991. doi: 10.1016/j.lanwpc.2023.
100991.
14. Deng C, Li X, Pan J. Private hospital expansion in China: a
global perspective. Glob Health J 2018;2:33–46. doi: 10.1016/
S2414-6447(19)30138-1.
15. Jo O, Kruger E, Tennant M Public transport access to NHS den-
tal care in Great Britain. Br Dent J. doi:10.1038/s41415-021-
3002-3.
16. Shah TI, Bell S, Wilson K. Spatial accessibility to health care
services: identifying under-serviced neighbourhoods in
Canadian Urban Areas. PLoS One 2016;11:e0168208. doi:
10.1371/journal.pone.0168208.
17. Gesler W. The uses of spatial analysis in medical geography: a
review. Soc Sci Med 1986;23:963–73. doi: 10.1016/0277-9536(86)
90253-4.
18. Crane FG. Professional services marketing: strategy and tac-
tics. Psychology Press; 1993.
19. Ma S, Ding J, Huang Z, Guo R. Evaluation of the industrial
cooperation of the Guangdong Hong Kong Macao Greater Bay
Area: Based on the Origin-Destination pairs of industrial
parks and coupling model efficiency. PLoS One 2023;18:
e0290572. doi: 10.1371/journal.pone.0290572.
20. Li H, Wei YD, Wu Y, Tian G. Analyzing housing prices in
Shanghai with open data: amenity, accessibility and urban
structure. Cities 2019;91:165–79. doi: 10.1016/j.cities.2018.11.
016.
21. van Donkelaar A, Hammer MS, Bindle L, Brauer M, Brook JR,
Garay MJ, et al. Monthly global estimates of fine particulate
matter and their uncertainty. Environ Sci Technol
2021;55:15287–300. doi: 10.1021/acs.est.1c05309.
22. WorldPop (www.worldpop.org - School of Geography and
Environmental Science, University of Southampton; Depart-
ment of Geography and Geosciences, University of Louisville;
Departement de Geographie, Universite de Namur) and Cen-
ter for International Earth Science Information Network (CIE-
SIN), Columbia University (2018), 2020.
23. Getis A. Reflections on spatial autocorrelation. Reg Sci Urban
Econ 2007;37:491–6.
24. Hosage CM, Goodchild MF. Discrete space location-allocation
solutions from genetic algorithms. Ann Operat Res 1986;6:35–
46.
25. Isazade V, Qasimi AB, Dong P, Kaplan G, Isazade E. Integration
of Moran’s I, geographically weighted regression (GWR), and
ordinary least square (OLS) models in spatiotemporal model-
ing of COVID-19 outbreak in Qom and Mazandaran Provinces,
Iran. Model Earth Syst Environ 2023:1–15. doi: 10.1007/s40808-
023-01729-y.
26. Raudenbush SW. Analyzing effect sizes: random-effects mod-
els. In: Cooper H, Hedges LV, Valentine JC, editors. The
handbook of research synthesis and meta-analysis. 2nd ed.
New York: Russell Sage Foundation; 2009. p. 295–315.
27. Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic
introduction to fixed-effect and random-effects models for
meta-analysis. Res Synth Methods 2010;1:97–111. doi:
10.1002/jrsm.12.
28. Elhorst JP. Specification and estimation of spatial panel data
models. Int Region Sci Rev 2003;26:244–68.
29. Abhishek N, Jenamani M, Mahanty B. Urban growth in Indian
cities: are the driving forces really changing? Habitat Int
2017;69:48–57. doi: 10.1016/j.habitatint.2017.08.002.
30. Listl S, Galloway J, Mossey PA, Marcenes W. Global economic
impact of dental diseases. J Dent Res 2015;94(10):1355–61. doi:
10.1177/0022034515602879.
31. Heaton LJ, Smith TA, Raybould TP Factors influencing use of
dental services in rural and urban communities: considera-
tions for practitioners in underserved areas. J Dent Educ
2004;68:1081−1089.
32. Li Q, Wei J, Jiang F, Zhou G, Jiang R, Chen M, et al. Equity and
efficiency of health care resource allocation in Jiangsu Prov-
ince, China. Int J Equity Health 2020;19:211. doi: 10.1186/
s12939-020-01320-2.
33. Zhang Y, Wang Q, Jiang T, Wang J. Equity and efficiency of pri-
mary health care resource allocation in mainland China. Int J
Equity Health 2018;17:140. doi: 10.1186/s12939-018-0851-8.
34. Huang Z, Kawamura K, Kitayama T, Li Q, Yang S, Miyake T.
GIS-based study of dental accessibility and caries in 3-year-
old Japanese children. Int Dent J 2023;73:550–7. doi: 10.1016/j.
identj.2022.11.002.
35. Khan MA, Owais SS, Maqbool S, Ishaq S, Khan HJ, Minhas FA,
et al. Is integrated private-clinic based early child develop-
ment care effective? A clustered randomised trial in Pakistan.
BJGP Open 2018;2(2). doi: 10.3399/bjgpopen18X101593.
36. Gil MR, Choi CG. Factors affecting the choice of national and
public hospitals among outpatient service users in South
Korea. Inquiry 2019;56:46958019833256. doi: 10.1177/
0046958019833256.
37. Yang S, Liu L, Wang C, Lo K, Wang D. Elderly people’s prefer-
ences for healthcare facilities in Shanghai: gender features
and influencing factor analysis. BMC Public Health
2023;23:356. doi: 10.1186/s12889-023-15279-6.
38. Zhang DE, Xie J, Wang YF, Tang WX. Influence of resource
level of public health institutions on the development of pri-
vate health institutions (2002-2017): Based on threshold
regression. Chinese J Health Policy 2019;12(9):66–71. doi:
10.3969/j.issn.1674-2982.2019.09.012.
39. Yu H, Yu S, He D, Lu Y. Equity analysis of Chinese physician
allocation based on Gini coefficient and Theil index. BMC
Health Services Res 2021;21:1–8.
40. Jiang L, Liu Y. China’s largest city-wide lockdown: how exten-
sively did Shanghai COVID-19 affect intensity of human activ-
ities in the Yangtze River Delta? Remote Sensing 2023;15:1989.
doi: 10.3390/rs15081989.
41. McCartney G, Pinto J, Liu M. City resilience and recovery from
COVID-19: the case of Macao. Cities 2021;112:103130. doi:
10.1016/j.cities.2021.103130.
42. Chung PC, Chan TC. Association between local spatial acces-
sibility of dental care services and dental care quality. BMC
Oral Health 2021;21:582. doi: 10.1186/s12903-021-01943-z.
dental clinic distribution and impact factors 1101
